Berkeley Law
Berkeley Law Scholarship Repository
Faculty Scholarship
1-1-1985
From Informed Consent to Patient Choice: A New
Protected Interest
Marjorie Maguire Shultz
Berkeley Law
Follow this and additional works at: htp://scholarship.law.berkeley.edu/facpubs
Part of the Law Commons
Recommended Citation
Marjorie Maguire Shultz, From Informed Consent to Patient Choice: A New Protected Interest, 95 Yale L.J. 219 (1985),
Available at: htp://scholarship.law.berkeley.edu/facpubs/1737
his Article is brought to you for free and open access by Berkeley Law Scholarship Repository. It has been accepted for inclusion in Faculty
Scholarship by an authorized administrator of Berkeley Law Scholarship Repository. For more information, please contact jcera@law.berkeley.edu.
The Yale Law Journal
Volume 95, Number 2, December 1985
Articles
From Informed Consent to Patient
Choice: A New Protected Interest
Marjorie Maguire Shultz-
INTRODUCTION
Judges and legal scholars have long asserted the importance of patient
autonomy in medical decisionmaking. Yet autonomy has never been recognized as a legally protectable interest. It has been vindicated only as a byproduct of protection for two other interests-bodily security as protected
by rules against unconsented contact, and bodily well-being as protected
by rules governing professional competence. Neither bodily security nor
bodily well-being, however, is an adequate surrogate; they do not coincide
with autonomy. Nor is autonomy merely a formal issue. Decisionmaking
by competent professionals does not provide an adequate substitute for
patient choice. Injuries that arise from invasion of patients' interest in
medical choice are both substantial and distinct.
Part I of this Article explains the importance of patient autonomy and
describes how existing doctrines protect that value. Part II examines gaps
and flaws in that current scheme of protection. Part III analyzes clusters
of cases in which greater vindication of patient autonomy has begun to
t" Lecturer, Boalt Hall School of Law, University of California, Berkeley. B.A. 1962, College
of
Wooster; M.A.T. 1964, University of Chicago; J.D. 1976, University of California, Berkeley.
I am
deeply grateful to William Fletcher, Robert Cole, Meir Dan-Cohen and Jane Staw; each
gave invaluable assistance during the development of this Article. I am also much indebted to my research
assistants and seminar students: Susan Raffanti, Dena Belinkoff, Catherine Fisk, Tina Stevens,
Jon Polland and Allyn Taylor.
HeinOnline -- 95 Yale L.J. 219 1985-1986
Vol. 95: 219, 1985
The Yale Law Journal
emerge, and urges that these developments should be generalized. Part IV
recommends the creation of a distinct and independently protected interest
in patient autonomy.
I.
A.
STARTING POINTS
The Importance of Patient Autonomy
Individuality and autonomy have long been central values in AngloAmerican society and law. In general, the more intense and personal the
consequences of a choice and the less direct or significant the impact of
that choice upon others, the more compelling the claim to autonomy in the
making of a given decision." Under this criterion, the case for respecting
patient autonomy in decisions about health and bodily fate is very strong.
The very fact that health care choices are extremely important, however, generates fear that individuals will make mistakes. The complex and
esoteric nature of modern medicine necessitates advice from experts.
Needed perspective and emotional support can be provided by family and
friends. Given that medical choices affect the quality and even the length
of life itself, individuals making such choices may well be urged to seek all
the help, in terms of both love and knowledge, that they can find. Ultiintense
mately, however, the stake of both experts and loved ones is less
3 Patients'
than that of the patient whose well-being is directly affected.
preferences, therefore, ought generally to be controlling.
1. Perhaps the most articulate advocate of this view was John Stuart Mill:
IT]he sole end for which mankind are warranted . . . in interfering with the liberty of
one,
action of any of their number, is self-protection. . . . The only part of the conduct of any
merely
for which he is amenable to society, is that which concerns others. In the part which
body and
concerns himself, his independence is, of right, absolute. Over himself, over his own
mind, the individual is sovereign.
J. MILL, ON LIBERTY 6 (1873).
in
2. The classic statement of this value in the medical context is that of Judge Cardozo
has a right
Schloendorff v. Society of N.Y. Hosp.: "Every human being of adult years and sound mind
withto determine what shall be done with his own body; and a surgeon who performs an operation
(1914).
93
92,
N.E.
105
129-30,
125,
N.Y.
211
.
."
.
.
assault
an
commits
consent,
patient's
out his
A more recent statement of the importance of patient autonomy is found in 1 PRESIDENT'S COMM'N
AND BEHAVIORAL REFOR THE STUDY OF ETHICAL PROBLEMS IN MEDICINE AND BIOMEDICAL
cited as MAKING DECISIONS].
SEARCH, MAKING HEALTH CARE DECISIONS 2-4 (1982) [hereinafter
See generally Goldstein, For Harold Lasswell: Some Reflections on Human Dignity, Entrapment,
protection
Informed Consent, and the Plea Bargain, 84 YALE L.J. 683 (1975) (critiquing inadequate
of self-determination in three disparate processes, including informed consent to medical intervention).
tradition.
3. In Western medicine, deference to the patient's interest is rooted in the Hippocratic
deference to the
See R. VEATCH, A THEORY OF MEDICAL ETHICS 21-25 (1981). In legal terms, the
464 F.2d
patient's interest is rooted in the doctor's status as a fiduciary. See Canterbury v. Spence,
inter772, 782 (D.C. Cir.), cert. denied, 409 U.S. 1064 (1972). Under these principles, the doctor's
ests in income, prestige and convenience, as well as in her own professional opinions and preferences,
the paconstitute a less immediate and compelling claim to authority than that which derives from
religious
ethical,
have
may
doctors
however,
occasion,
On
consequences.
of
bearer
the
as
tient's status
349.
or professional convictions that lead them to wish to refuse some services. See infra note
HeinOnline -- 95 Yale L.J. 220 1985-1986
Protecting Patient Choice
B. Implicationsfor the Doctor-PatientRelationship
Although the principle of individual autonomy is widely endorsed in
theory, its practical implications for the doctor-patient relationship are
controversial. Individuals exercise their autonomy in medical decisionmaking by arranging for needed professional services. Presumably, these individuals remain the source of authority and can choose to delegate all or
only some of their control to professionals.4 Yet, ironically, the most significant threat to patient autonomy comes from the very doctors whom
patients hire. Because of their knowledge and traditional role, doctors
often preempt patient authority.
Although scholars have proposed various models to describe or prescribe the distribution of power within the doctor-patient relationship, 5 for
a number of years one view dominated professional ideology and customary practice. Under that view, the patient was seen as making only one
key decision, to place herself in a given doctor's care, thereby delegating
all subsequent authority to the doctor." Such a model assumed that the
patient lacked the technical ability to make medical decisions, and that
expertise justified the doctor's making decisions on the patient's behalf.
In the past several decades, however, new developments have strengthened the argument that patient autonomy should receive more than pro
forma respect. Advancing medical technology has greatly expanded the
options available to the patient. Increased knowledge has heightened
4. Such a right flows from the contractual nature of the doctor-patient relationship. A patient's
decision to forego information and consent procedures is typically called a waiver under informed
consent doctrine. See Meisel, The "Exceptions" to the Informed Consent Doctrine: Striking a Balance Between Competing Values in Medical Decisionmaking, 1979 Wis. L. REv. 413, 453-60 (discussing waiver exception) [hereinafter cited as Meisel, Exceptions]. The difficult issue is determining
when such a waiver has been made by the patient.
5. See, e.g., R. BURT, TAKING CARE OF STRANGERS: THE RULE OF LAW IN DOCTOR-PATIENT
RELATIONS (1979) (proposing extensive dialogue rather than black and white rules of thumb); Branson, The Secularization of American Medicine, HASTINGS CENrER STun., Nov. 1973, at 17 (advocating that science no longer be regarded as "religion," and that doctors assume more modest, less
"priestly" role); Childress, Metaphors and Models of Medical Relationships, 8 Soc. REsp.: JouRNALISM, L. MED. 47 (1982) (analyzing various models of medical relationships); May, Code, Covenant, Contract, or Philanthropy, HASTINGS CENTrstrt REP., Dec. 1975, at 29 (espousing "covenant"
as most inclusive and satisfying model of obligation); Szasz & Hollender, A Contribution to the Philosophy of Medicine: The Basic Models of the Doctor-PatientRelationship, in 97 A.M.A. ARCHIVES
OF INTERNAL MED. 585 (1956); Veatch, Models for Ethical Medicine in a Revolutionary Age, HASTINGS CENTER REP., June 1972, at 5 (pointing out significance of choice of relationship model).
6. See 1 MAKING DECISIONS, supra note 2, at 19; Meisel, The Expansion of Liabilityfor Medical Accidents: From Negligence to Strict Liability By Way of Informed Consent, 56 NEB. L. REV. 51,
77-80 (1977) (tracing legal requirements imposed upon doctors from early days of minimalism to
more demanding contemporary times) [hereinafter cited as Meisel, Expansion]. Under earlier models,
detailed disclosure to and consent by the patient played a role only insofar as they were thought to be
essential to the therapeutic outcome. Pernick, The Patient's Role in Medical Decisionmaking:A Social History of Informed Consent in Medical Therapy, in 3 MAKING DECISIONS, supra note 2, at 1,
14.
HeinOnline -- 95 Yale L.J. 221 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
7
awareness of how much remains unknown. Debate and conflict within
the medical community are widespread and public." Differences in experts' advice can often be resolved only on the basis of risk and value
preferences. This medical uncertainty accentuates the need for professional advice, but it also strengthens the case for ultimate decision by the
person whose life is directly involved.
Medical choice increasingly depends on factors that transcend professional training and knowledge. As medicine has become able to extend
life, delay and redefine death, harvest and transplant organs, correct abnormality within the womb, enable artificial reproduction, and trace genetic defect, questions about values have come to the fore in medical decisionmaking. Health care choices involve profound questions that are not
9
finally referable to professional expertise.
In the face of value pluralism, factual indeterminacy, and increasing
options, patient autonomy has become a central principle of both popu1
lar ° and philosophical analysis of medical decisionmaking. Self-care and
consumer movements have applied that principle, seeking to shift the balance away from professional dominance and toward individual knowledge
and control.11 Although medical traditions historically have downgraded
THE SILENT WORLD OF DOCTOR AND PATIENT 183-84 passim (1984) (discussmedical uncertainty).
of
ing pervasiveness
Brooks,
8. Newspaper articles describing medical differences of opinion are common. See, e.g.,
at 1, col.
1984,
16,
July
J.,
St.
Wall
Victims,
Potential
and
Stroke
of
Victims
to
Offered
is
New Hope
D2, col. 3.
1; Top Surgeon Sees No Futurefor Artificial Heart, N.Y. Times, Feb. 14, 1984, at
at 29.
9. See Veatch, Generalization of Expertise, HASTINGS CENTER STUD., Nov. 1973,
MED.
Disclosure,
More
Doctors?:
from
Want
Really
Society
10. See, e.g., Enright, What Does
Peck, What Does
ECON., May 29, 1978, at 77 (reporting public preferences for greater information);
1978, at 93,
Society Really Want from Doctors?: Greater Accountability, MED. ECON., May 29,
Stick up for
Hospital?
to
Going
consumers);
to
doctors
of
105-11 (discussing growing accountability
Pres.,
Your Rights, U.S. NEWS & WORLD REP., Oct. 1, 1984, at 61 (interview with Norma Calhoun,
decisionmakin
role
greater
gain
to
how
patients
(advising
Representatives)
Patient
of
Soc'y
National
discussing); The PatientHas a Right to Know, N.Y. Times, June 15, 1984, at A26, col. 1 (editorial
Commission
ing need for disclosure about health care provider costs and results). The Presidential
in
found that the vast majority of respondents to its survey wanted information and participation
17.
at
2,
note
supra
medical decisions. 1 MAKING DECISIONS,
(1982)
11. See, e.g., J. CHILDRESS, WHO SHOULD DECIDE? PATERNALISM IN HEALTH CARE
AuDworkin,
contexts);
medical
in
justification
their
analyzing
and
(classifying types of paternalism
role
tonomy and Informed Consent, 3 BIOETHICS REP.: LITERATURE 309 (1983) (clarifying special
8 J.
of autonomy in justifying medical consent norms); Gordon, The Doctor-PatientRelationship,
need for physician to
MED. & PHIL. 243 (1983) (from perspective of religious humanist, emphasizing
respect autonomy of patient).
391-92 (1983)
12. See P. STARR, THE SOCIAL TRANSFORMATION OF AMERICAN MEDICINE
of self-care in
theorist
leading
A
dominance).
professional
of
(describing movements exhibiting distrust
HEALTH (1976).
medicine is Ivan Illich. See I. ILLICH, MEDICAL NEMESIS: THE EXPROPRIATION OF
WOMEN'S
Both alternative care and self-care have become major influences. See, e.g., BOSTON
BY AND FOR WOMEN
HEALTH BOOK COLLECTIVE, THE NEW OUR BODIES, OURSELVES: A BOOK
tools which will
xiii (1984) (one stated goal of book is "to reach as many women as possible with the
care").
health
own
their
of
charge
greater
take
to
enable them
7. See J. KATZ,
HeinOnline -- 95 Yale L.J. 222 1985-1986
Protecting Patient Choice
patient autonomy, 3 doctors, too, have begun to recognize and accept patient demands for more information and control. 14
The law's response to pressures for greater recognition of patient autonomy has been ambivalent.1 5 Existing rules repudiate the view that the
mere hiring of a doctor transfers all authority from patient to doctor. Yet
full vindication of patient autonomy interests would necessitate placing
final authority regarding important decisions in the hands of any patient
having the capacity and the desire to exercise it." I shall argue that precisely such a model for the allocation of authority is appropriate, but, as
Part II will demonstrate, no such guarantee of patient autonomy is currently mandated by the law.
C. Existing Legal Protection of PatientAutonomy
Although the doctor-patient relationship ordinarily arises through contract, 7 courts have deemed patients incapable of bargaining with doctors
over the quality of medical services. Doctors' performance has therefore
been monitored under standardized tort rules that govern professional
13. Jay Katz traces the history of nondisclosure and failure to share decisions with patients
throughout the history of Western medicine. J. KATz, supra note 7, at 1-29. He points out the direct
conflict between the medical norm of "custody" and the legal norm of personal liberty. Id. at 2.
14. See, e.g., Novack, Plumer, Smith, Ochitill, Morrow & Bennett, Changes in Physicians' Attitudes Toward Telling the Cancer Patient,241 J. A.M.A. 897, 897 (1979) (reporting dramatic reversal in attitude among doctors from 1961 to 1979 over whether to tell patients they have cancer); see
also Harris & Associates, Views of Informed Consent and Decisionmaking: ParallelSurveys of Physicians and the Public, in 2 MAKING DECISIONS, supra note 2, at 17, 18, passim (physicians reported positive attitudes toward disclosure).
15. This point is developed particularly by Jay Katz. See J. KATZ, supra note 7; Katz, Informed
Consent-A Fairy Tale? Law's Vision, 39 U. Prrr. L. REv. 137, 138 & nn.1-39 (1977) [hereinafter
cited as Katz, Informed Consent]. The point is also explored in White, Informed Consent: Ambiguity
in Theory and Practice, 8 J. HEALTH PoL., PoL'Y & L. 99 (1983).
16. The province of this discussion is decisionmaking by the autonomous patient who is able to
and wants to make her own choices. Traditional exceptions to informed consent doctrine-situations
where an emergency exists, where the patient waives the authority to make choices, or where the
patient lacks decisionmaking capacity-are not affected by this analysis. Waiver is technically not an
exception to respect for autonomy, because it is itself an expression of choice. The other exceptions
track the most commonly recognized circumstances under which beneficence, even in the form of
paternalism, may justifiably displace autonomy. See Dworkin, Paternalism, 56 MoNISr 64, 76-84
(1972). The most extensive treatment of exceptions is in Meisel, Exceptions, supra note 4. Determining patients' decisionmaking capacity can be very problematic. See Roth, Meisel & Lidz, Tests of
Competency to Consent to Treatment, 134 AM. J. PSYCHIATRY 279 (1977) (describing five possible
standards for determining competency). Moreover, issues may arise regarding temporary or situationspecific incapacity. See, e.g., In re Yetter, 62 Pa. D. & C.2d 619 (Northampton County Ct. 1973)
(patient's confinement in mental hospital did not mean she lacked capacity to refuse medical treatment); Nishi v. Hartwell, 52 Hawaii 188, 473 P.2d 116 (1970) (disclosure of side effect of treatment
might have had adverse effect on patient).
17. See, e.g., Gray v. Grunnage, 423 Pa. 144, 166, 223 A.2d 663, 674 (1966) ("IT]he agreement
between the physician and his patient is contractual in nature . . ."). The consensual origins of the
relationship are especially emphasized in Epstein, Medical Malpractice:The Casefor Contract, 1976
Am. BAR. FOUND. RESEARCH J. 87, 119. Entry into a contract for services is typically, though not
always, the basis for imposition of physician duties of care.
HeinOnline -- 95 Yale L.J. 223 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
malpractice rather than under contractual criteria of individual expectain selftion. 18 Mainly as a derivative matter, the patient's interest
19
theories.
tort
determination has also been analyzed under
1. Battery
Patient autonomy was initially identified with and subsumed under an
interest in physical security, protected by rules proscribing unconsented
touch. 20 Medical care often involves touching, and may be considered bat21
tery if the touching is unconsented. By mandating patient consent to
specific procedures, battery doctrine counters the implication that doctors
acquire authority to make decisions simply by virtue of the contract for
professional services. Moreover, professional competence is no defense to a
medical battery action. 22 Under battery analysis, the patient's wishes take
priority over even the fully competent recommendation of a doctor, unless
an exception applies. 3 Apart from traditional defenses,2 the right to be
secure against unconsented touching is close to absolute. ' Application of
battery doctrine to medical care thus establishes an uncompromising baseline of protection for patients' self-determination.
Despite the capacity of battery doctrine to protect a degree of physical
principles in
18. See Epstein, supra note 17, at 91-96 (discussing interplay of tort and contract
of lay judgment-here
doctor-patient relationship). Under tort law, the same fear about inadequacy
of professional
the judgment of jurors rather than patients-leads courts to concede most regulation
KEETON, D. DOBBS, R. KEErON & D.
W.
itself.
profession
the
of
standards
the
to
competence
(5th ed. 1984) [hereinafter
OWEN, PROSSER & KEETON ON THE LAW OF TORTS § 32, at 188-89
contract term by
cited as PROSSER & KEETON]. Epstein conceives of this as supplying an omitted
incorporating custom of the trade. Epstein, supra note 17, at 110.
19. The question of self-determination might be analyzed as a problem of the scope of an agency,
have urged a more
to be determined by contract between principal and agent. Some commentators
17.
note
supra
Epstein,
e.g.,
See,
relations.
medical
contractually oriented analysis of
PER20. See Pernick, supra note 6, at 29-30; see also C. FRIED, MEDICAL EXPERIMENTATION:
evolution of these doctrines).
SONAL INTEGRITY AND SOCIAL POLICY 14-18 (1974) (discussing
21. PROSSER & KEETON, supra note 18, § 9, at 39.
22. W. PROSSER, HANDBOOK OF THE LAW OF TORTS § 18, at 104 (4th ed. 1971).
Legal restric23. Under battery doctrine, the competent patient's choices are largely unfettered.
may compel
tions may arise from the protectable rights of other individuals. The state, for example,
on individual
restrictions
policy
Public
(1905).
11
U.S.
197
Massachusetts,
v.
Jacobson
vaccination.
(Cal. Super.
choice are, however, hotly debated. See, e.g., Bouvia v. County of Riverside, No. 159780
Moreover, constituCt. Dec. 16, 1983) (order denying preliminary injunction against forced feeding).
tional privacy decisions have contracted the scope of permissible intervention in medical choice. See,
Andrews v.
e.g., Roe v. Wade, 410 U.S. 113 (1973) (privacy right to choice concerning abortion);
treatment
of
form
choose
to
patients
of
right
(privacy
1980)
Tex.
(S.D.
Ballard, 498 F. Supp. 1038
362 So. 2d 160 (Fla.
violated by state restrictions on performance of acupuncture); Satz v. Perlmutter,
life does not override
Dist. Ct. App. 1978), aff d, 379 So. 2d 359 (1980) (state interest in preserving
treatment).
refuse
to
right
patient's
sick
mortally
but
competent
for purely digni24. C. FRIED, supra note 20, at 16. Damages range from symbolic recompense
touching (no
tary injury, to compensation of physical or economic injury resulting from unconsented
particularly culpable
matter how well intentioned or expertly conducted), to punitive awards for
see also PROSSER
touchings. RESTATEMENT (SECOND) OF TORTS §§ 901, 903, 905, 907, 908 (1979);
of intentions, it
kind
worst
the
of
a
matter
is
battery
("Since
40-41
at
9,
§
18,
note
& KEETON, supra
damages.").
punitive
is a tort which frequently justifies
HeinOnline -- 95 Yale L.J. 224 1985-1986
Protecting Patient Choice
autonomy in patients' relations with doctors, many aspects of the medical
care relationship do not fit comfortably within the battery model. Doctors
lack the antisocial motivation usually associated with intentional torts such
as battery. 25 Further, unlike in the typical battery case, the patient usually
has given a degree of consent to the doctor's treatment, if only in the
broad sense that the patient has sought medical care from the doctor.
Once courts began more thoroughly to examine the subtleties of the
doctor-patient relationship, the difficulties inherent in applying battery
analysis to problems of medical consent became impossible to ignore. On
the one hand, a general consent to treatment given without awareness of
risks, prognoses, and options was seen as an insufficient basis upon which
to authorize treatment, even medically defensible treatment. 26 Yet to hold
that such uninformed consent was invalid, thereby subjecting doctors to
actions for battery, threatened to yield unacceptably harsh results. Given
the absolute nature of battery, the narrowness of its defenses, and the
breadth of its remedies, doctors could end up paying significant damages
after providing faultless medical treatment, simply because some minor
informational aspect of the consent process was questioned."
25. This point is stressed in many cases, e.g., Trogun v. Fruchtman, 58 Wis. 2d 596, 599, 207
N.W.2d 297, 313 (1973).
26. See Canterbury v. Spence, 464 F.2d 772, 782 (D.C. Cir.), cert. denied, 409 U.S. 1064 (1972).
27. See, e.g., Berkey v. Anderson, I Cal. App. 3d 790, 804, 82 Cal. Rptr. 67, 77 (1969) (issue of
technical battery raised by failure to disclose risks of myelogram); Fogal v. Genesee Hospital, 41
A.D.2d 468, 344 N.Y.S.2d 552, 559 (1973) (failure to disclose risks of using hypothermia blanket
cognizable under theory of assault and battery rather than negligence); Congrove v. Holmes, 37 Ohio
Misc. 95, 308 N.E.2d 765 (C.P. Ross County 1973) (summary judgment for plaintiff for damages for
paralysis of vocal cords resulting from thyroidectomy, where physician did not warn of risks); Cooper
v. Roberts, 220 Pa. Super. 260, 286 A.2d 647 (1971) (assault and battery theory applies where doctor
failed to inform patient of collateral risks of perforation of stomach during gastroscopic examination).
These results are possible because battery doctrine employs a very simple analysis of causation. See
Plant, An Analysis of "Informed Consent," 36 FORDHAM L. REv. 639, 666 (1968) [hereinafter cited
as Plant, Analysis]. Presumably this is because ordinarily a battery defendant's conduct is assumed to
be antisocial and not deserving of the protection of more thorough causal analysis. PROSSER & KEETON, supra note 18, § 9, at 35, § 43, at 263; Riskin, Informed Consent: Looking for the Action, 1975
U. ILL. L.F. 580, 583-84. Thus, battery analysis does not inquire whether the patient would have
consented if the doctor had acted properly. Battery doctrine treats the consent, if flawed, as completely
invalid. See W. PROSSER, HANDBOOK OF THE LAW OF TORTS, § 32, at 165 (4th ed. 1971). The
likely result would be recovery of all damages, even if the injury was likely to have occurred in any
case because the patient would have consented. See Riskin, supra, at 583-84, 601; cf King, Causation, Valuation and Chance in PersonalInjury Torts Involving Preexisting Conditions and Future
Consequences, 90 YALE L.J. 1353 (1981) (critiquing tendency in present tort law to confuse issues of
causation and valuation).
If battery doctrine had continued to be applied to medical consent, some more sophisticated (and
more fair) fashion of handling causation and damages might have developed. Such a development
might parallel earlier decisions in cases of medical battery that allowed a set-off for benefits derived
from unauthorized treatment to reduce damages for injury to a dignitary interest. See, e.g., Mohr v.
Williams, 95 Minn. 261, 104 N.W. 12 (1905) (taking into account benefits from unauthorized operation on plaintiff's ear); McCandless v. State, 3 A.D.2d 600, 162 N.Y.S.2d 570 (1957) (award for
unauthorized abortion reduced by offset for improvement in condition), af/'d, 4 N.Y.2d 797, 149
N.E.2d 530, 173 N.Y.S.2d 30, (1958). But see D. DOBBS, LAW OF REMEDIES 182 (1973) (questioning wisdom of offsetting for benefits received in context of intentional tort). However, the transfer of
HeinOnline -- 95 Yale L.J. 225 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
Discomfort with treating doctors under a doctrine aimed at antisocial
conduct has prompted most jurisdictions to limit the battery action to
those relatively unusual situations where a medical procedure has been
carried out without any consent, rather than where the consent has merely
been insufficiently informed.2 The modern allegation of battery typically
arises when consent to a particular procedure is given and a different or
additional procedure is carried out.2 9 The relative infrequency with which
battery claims arise today should not, however, obscure the fact that battery doctrine retains a critical philosophical and practical function in protecting patient self-determination.
2. Informed Consent"
Most litigation about patient autonomy now occurs over doctors' nondisclosure of information, analyzed as an issue of professional negligence.-" Doctors' responsibility for professional care of patients' physical
most disclosure issues to analysis under negligence doctrine removed the necessity for such a
development.
28. See, e.g., Cobbs v. Grant, 8 Cal. 3d 229, 240, 502 P.2d 1, 8, 104 Cal. Rptr. 505, 512 (1972)
("We agree with the majority trend. The battery theory should be reserved for those circumstances
when a doctor performs an operation to which the patient has not consented."); Trogun v.
Fruchtman, 58 Wis. 2d 596, 598-600, 207 N.W.2d 297, 312-13 (1973) (explaining why battery
action is appropriate where operation is unauthorized but not where only issue is nondisclosure). At
least one state has abolished any action for medical battery. See ARIz.REV. STAT. ANN. § 12-562(B)
(1982).
29. See, e.g., Reddington v. Clayman, 334 Mass. 244, 134 N.E.2d 920 (1956) (question of battery
presented where surgeon allegedly removed child's uvula during operation for removal of tonsils and
adenoids); Hively v. Higgs, 120 Or. 588, 253 P. 363 (1927) (where consent given to operate on
septum of nose, removal of tonsils constitutes battery).
The most difficult decisions concern the scope of a consent. See, e.g., Kinikin v. Heupel, 305
N.W.2d 589 (Minn. 1981) (upheld jury verdict on battery theory regarding removal of more of breast
than patient intended); Gray v. Grunnagle, 423 Pa. 144, 223 A.2d 663 (1966) (patient's claim that he
consented only to exploratory surgery, not to laminectomy, should have gone to jury on battery ciaim);
see also A. ROSOFF, INFORMED CONSENT: A GUIDE FOR HEALTH CARE PROVIDERS 8-13 (1981)
(discussing factors leading courts to accept or refuse battery theory in disputes over scope of consent).
30. The literature on informed consent is voluminous. Some of the articles that emphasize the
importance of patient autonomy include: J. KATz,supra note 7, at 104-64; J. KArz & A. CAPRON,
CATASTROPHIC DISEASES: WHO DECIDES WHAT? 82-85 (1975); Epstein, supra note 17, at 87, 119;
Goldstein, supra note 2, at 695 nn.27-29; Katz, Informed Consent, supra note 15; Meisel, Exceptions, supra note 4, at 413-22; Riskin, supra note 27; White, supra note 15. For additional sources
on the doctrine, see Capron, Informed Consent in CatastrophicDisease Research and Treatment,
123 U. PA. L. REV. 340 (1974); Hagman, The Medical Patient'sRight to Know: Report on a Medical-Legal-Ethical,Empirical Study, 17 UCLA L. REv. 758 (1970); Meisel, Expansion, supra note
6; Meisel & Kabnick, Informed Consent to Medical Treatment: An Analysis of Recent Legislation,
41 U. Pirr. L. REV. 407 (1980); Plant, Analysis, supra note 27; Plant, The Decline of "Informed
Consent," 35 WASH. & LFR L. REv. 91 (1978); Waltz & Scheuneman, Informed Consent to Therapy, 64 Nw. U.L. REV. 628 (1969); Note, Informed Consent and the Dying Patient, 83 YALE L.J.
1632 (1974); Note, Informed Consent in Medical Malpractice,55 CAL. L. REV. 1396 (1967); Note,
RestructuringInformed Consent: Legal Therapy for the Doctor-PatientRelationship, 79 YALE L. J.
1533 (1970) [hereinafter cited as Note, Restructuring].For a review and synopsis of the vast array of
empirical studies in the area, see Meisel & Roth, Toward an Informed Discussion of Informed Consent: A Review and Critique of the Empirical Studies, 25 ARiz L. REv. 265 (1983).
31. PROSSER & KEETON, supra note 18, §§ 15, 30, at 106, 165.
226
HeinOnline -- 95 Yale L.J. 226 1985-1986
Protecting Patient Choice
well-being gives rise to various specific duties, one of which is the provision of sufficient information to allow a patient's decision to be intelligently informed. To recover for nondisclosure under the rules of professional malpractice, the patient must first show a violation of the duty to
inform, defined in many states by the standard of expert professionals. 32
Second, the nondisclosure must be shown to have caused a harm cognizable under negligence doctrine. Most states have adopted an objective standard of causation in medical informed consent cases.33 This standard requires the patient to show that the undisclosed information would have
induced not just this patient, but a reasonable patient, to withhold consent
to the treatment in question. 34
3.
The Justifying Prototype
The shift to negligence analysis made apparent analytic and practical
sense. 35 Although some critics decried losses to patient autonomy that
would result from emerging negligence rules,3" current legal protection of
patient autonomy has generally been deemed adequate. That judgment,
however, rests upon assumptions that are insufficiently examined and ultimately erroneous.
Assumptions that have implicitly governed the debate over patient autonomy are exposed in a prototypical example of informed consent that is
frequently used in discussion and reflected in litigation. 7 The doctor pro32.
For a state by state summary, see 3 MAKING DECISIONS, supra note 2, at 206-45. See also J.
AREEN, P. KING, S. GOLDBERG & A. CAPRON, LAW, SCIENCE AND MEDICINE 384 n.4 (1984) (interpreting evidence gathered in MAKING DECISIONS as follows: "[A]s of 1982, 26 states that had de-
clared law on informed consent had adopted a professional standard of disclosure, 19 a patientoriented standard, and 6 had no law on the subject.").
33. See 3 MAKING DECISIONS, supra note 2, at 197, 206-45. Although commentators have urged
the subjective standard, e.g., Katz, Informed Consent, supra note 15, at 163-64, only one state at
present seems to have explicitly embraced it, see Scott v. Bradford, 606 P.2d 554, 559 (Okla. 1980).
34. See, e.g., Canterbury v. Spence, 464 F.2d 772, 791 (D.C. Cir.), cert. denied, 409 U.S. 1064
(1972).
35. Protection of patient choice was thought to contribute to the "malpractice crisis" and health
care cost escalation. See Bly v. Rhoads, 216 Va. 645, 222 S.E.2d 783 (1976); J. LUDLUM, INFORMED
CONSET 41-42 (1978); Adams & Zuckerman, Variation in the Growth and Incidence of Medical
MalpracticeClaims, 9 J. HFALTH POL, PoL'Y & L. 475, 485 (1984) (principle of informed consent
is significantly associated with higher annual rate of claims after 1972); Miller, Informed Consent: I,
244 J. A.M.A. 2100, 2102 (1980).
36. See, e.g., Goldstein, supra note 2, at 691; Katz, Informed Consent, supra note 15, at 139;
Meisel, Expansion, supra note 6, at 112.
37. Not all elements of the prototype are evident in every case or every discussion, but they recur
often enough to be a legitimate composite. For instance, surgical cases overwhelmingly dominate the
exhaustive list of cases summarized by Rosoff. See A. RosoFF, supra note 29, at 471-520 (case index). A number of the leading and most frequently discussed cases involve such a prototype. See, e.g.,
Canterbury v. Spence, 464 F.2d 772 (D.C. Cir. 1972) (undisclosed risks of surgery), cert. denied, 409
U.S. 1064 (1972); Cobbs v. Grant, 8 Cal. 3d 229, 502 P.2d 1, 104 Cal. Rptr. 505 (1972) (same);
Salgo v. Leland Stanford Jr. Univ. Bd. of Trustees, 154 Cal. App. 2d 560, 317 P.2d 170 (1957)
(same).
HeinOnline -- 95 Yale L.J. 227 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
poses crucial surgery to a seriously ill patient. The surgery offers the patient's best hope for recovery; no viable alternatives exist. The surgical
intervention is competently recommended and competently carried out.
The doctor's only failing is nondisclosure of a low-probability risk of complication, such as the possibility of allergic reaction to the anesthetic. The
allergic reaction occurs, the patient suffers injury, and the doctor is sued
for failure to secure informed consent.
The surgery in the prototype is an invasive touching that requires consent; if there is a conflict between doctor and patient over the advisability
of doing the surgery, the competent patient's choice will prevail, as required by battery doctrine. Under modern analysis, the consent, if given,
remains valid, 8 and issues of nondisclosure will be analyzed under negligence doctrines governing duties to inform. There will be recovery only if
the disclosure is one that would have been made by competent doctors, if a
reasonable person in the patient's position would have refused the surgery
had the disclosure been made, and if the injury is cognizable under standard negligence principles.
In the context of this example, those limitations seem reasonable; the
potential for injury appears to be rather slight. As a reasonable person,
the patient would presumably have agreed to the surgery even had the
omitted disclosure been made. After all, the information that is undisclosed concerns a remote risk, the patient is seriously ill, and the surgery
is the only viable treatment. Thus, although individualists will stress that
in not disclosing all the facts the doctor has injured the patient's dignity
.and integrity, 9 many-perhaps most-will on these facts construe such
an injury to be largely symbolic. Moreover, a patient who would not have
consented to the surgery on these facts is aberrational. The idiosyncratic
and the symbolic do receive some protection under the rules of battery:
Basic, if not fully and exhaustively informed, consent is required in surgical cases. Where other important interests such as fairness to doctordefendants or medical cost escalation are involved, it may plausibly be
argued that such absolute protection need not be extended to the relatively
less crucial disclosure of information about remote risks.
On the other hand, the argument continues, in a situation where the
doctor's recommendation of surgery was one that a reasonable person
would not have accepted had disclosure been made (as might be the case if
the illness were minor and the undisclosed risk severe or likely to occur),
in all likelihood the recommendation itself was probably "wrong." If the
38. Under earlier analyses, evidence of the nondisclosure regarding a possible allergic reaction
would have invalidated the consent, potentially yielding recovery under a battery theory for all consequences of the surgery no matter how faultlessly performed. See supra note 27.
39. See, e.g., Goldstein, supra note 2.
HeinOnline -- 95 Yale L.J. 228 1985-1986
Protecting Patient Choice
recommendation was wrong, it would presumably be sanctioned under the
mainstream rules governing professional competence. Thus, if the informed consent action involved nondisclosures that led to reasonably
avoided and significant harms, it would seem to be largely duplicative of
an action in professional negligence.
Consequently, although autonomy remains an important value in theory, the prototypical example makes strong protection of patient choice
seem largely unnecessary in fact. The patient's interest in autonomy is
conceded basic protection under battery rules. Beyond that, under the prototype, the autonomy interest tends to be seen as either mainly symbolic or
highly aberrational on the one hand, or as largely redundant to protections under competence-regulating negligence rules on the other. Thus, in
the prototypical case, present doctrines may be argued to provide adequate
protection for patient autonomy.
But the prototypical case is not representative of the full range of cases
in which the autonomy interest is implicated. The conclusion regarding
the adequacy of current protection of patient autonomy derives from the
prototype's unwarranted assumptions, first, about the respective roles of
physical contact and information in medical choice, and second, about the
relationship between injuries to autonomy and injuries to physical wellbeing. As a result, the prototype both underestimates the degree to which
professional preemption of patient autonomy can occur and overestimates
the degree to which regulation of medical expertise provides an adequate
backup for doctrines safeguarding patient choice. Part II challenges these
assumptions and demonstrates how they have led to unacceptable flaws in
present legal protection for patient autonomy.
II.
A.
FLAWS IN EXISTING PROTECTION OF PATIENT AUTONOMY
Choices That Involve No Physical Touching Receive No Protection
Under Battery Doctrine
The doctrinal prototype described above assumes that important medical decisions are implemented through actual physical touching. Defining
the scope of an autonomy interest in terms of physical contact with the
body has intuitive appeal and offers a certain simplicity of administration.
But ultimately, physical contact is too literal a demarcation for what is a
much broader, non-tangible interest in patient choice. 0
40. Just as electronic sophistication made earlier notions of search and seizure insufficient and
required the expansion of privacy protection beyond limits on purely physical invasions, see Katz v.
United States, 389 U.S. 347 (1967), modern medical care requires a more flexible, less literal definition of the interest in patient autonomy than can be achieved through analysis based on physical
contact.
HeinOnline -- 95 Yale L.J. 229 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
Health care choices of vast consequence can be made and implemented
without such bodily contact as predictably triggers battery analysis. Most
notably, this occurs when a doctor makes a decision not to act. For instance, a doctor's judgment that a given level of diagnostic clarity is sufficient, that an acceptable outcome of treatment has been achieved, or that
no medical treatment can be administered, each has potentially grave
ramifications for the patient. As much as any that involve literal touching,
such judgments implicate important autonomy interests; yet under touchoriented rules they need not receive the patient's consent.
For example, in initial testing to determine the cause of eye problems,
the patient in Gates v. Jensen4 1 showed results consistent with borderline
glaucoma. The doctor undertook further diagnostic testing, and ultimately
concluded that the patient's problems derived not from glaucoma but from
contact lens irritation. He chose not to perform further tests that could
with greater certainty have determined the presence or absence of
glaucoma. Instead, the doctor prescribed treatment for what he thought to
be the problem. After Ms. Gates became legally blind, it was determined
that she did have glaucoma, and that it could have been treated and controlled had it been identified earlier. 42 The doctor's decision not to undertake further diagnostic testing for glaucoma was not expressed as a recommendation to the patient. Rather, the doctor exercised his own judgment
unilaterally, on the patient's behalf. Yet, because the doctor's judgment
involved no touching of Ms. Gates' body, there was no potential battery:
No consent to this "non-action" was required under that doctrine.
The choice of which Ms. Gates was deprived-whether to undergo further testing for glaucoma-was certainly as important as the one she was
given-whether to undergo treatment for the contact lens problem. Analyzed in terms of her interest in autonomy rather than her literal physical
security, this patient's opportunity to adopt, reject or modify the doctor's
unvocalized "recommendation" of inaction should have received as much
protection as the choice about the contact lens treatment.
Because the prototypical touching over which the vast bulk of doctorpatient litigation has taken place is surgery, 43 not even all actions by doctors will trigger battery analysis of patient consent. Where medical proposals involve less discrete or less invasive conduct than surgery, protection for the patient's interest in choice becomes correspondingly
attenuated." For instance, only one court seems to have applied battery
41.
42.
43.
44.
doctors
neither
92 Wash. 2d 246, 595 P.2d 919 (1979) (en banc).
Id. at 250, 595 P.2d at 924.
See A. RosoFF, supra note 29, at 471-520 (index of cases litigated, by category).
Thus in the survey conducted for the President's Commission by Harris Associates, 53% of
indicated that they sought neither written nor oral consent for prescriptions, and 43% sought
type of consent for blood tests. 2 MAKING DECISIONS, supra note 2, at 168. Outside the realm
HeinOnline -- 95 Yale L.J. 230 1985-1986
Protecting Patient Choice
rules to the prescription of medication, and that case involved atypical
factors.45 Writing a "No-Code" order in a patient's chart, like a choice
not to write such an order, will not require patient choice under touchtriggered rules. 46 Consent requirements also may not be triggered when
there is a continuing course of action, even when new information makes
new choices available. Thus, when a patient enters a hospital, a consent
form for hospitalization, including routine physical touching, is normally
signed. If, however, after some period of treatment it becomes clear that a
given patient's condition is medically hopeless, the patient may not get a
renewed opportunity to consent to the doctor's recommended course of
treatment. Where the doctor in effect recommends continued hospitalization by failing to recommend possibilities such as going home to die or
going to a hospice, her recommendations may well be imposed by default.
Thus, although courts have explained their concern over patient consent
in terms of autonomy values, 47 by confounding autonomy with control
of surgery or major and invasive diagnostic tests, there is little clarity about when specific consent is
required, as has been demonstrated by the empirical research of Meisel and Lidz. Id. at 328-35.
45. The aberrant case is Mink v. University of Chicago, 460 F. Supp. 713 (N.D. Ill. 1978) (for
discussion of case, see infra Part III(A)(1)). Most prescriptions are not analyzed as potential battery.
See 2 MAKING DECISIONS, supra note 2, at 333-34; 3 MAKING DECISIONS, supra note 2, at 16-17.
Two possible reasons not to apply battery analysis to prescriptions can be suggested. First, although
physical consequences may be significant, the doctor does not touch the patient at all. The doctor's
behavior in prescribing drugs is essentially judgmental and intellectual. Second, battery analysis may
be rejected because patients are deemed to have consented through their voluntary use of the drugs.
Such implied consents are often suspect, however. See A. RosOFF, supra note 29, at 5-6. Because
courts do not in these instances analyze whether the implied consent is valid, it seems likely that no
potential battery is perceived.
Compare the partly analogous problem of patient control over forced feeding. See, e.g., Bouvia v.
County of Riverside, No. 159780 (Cal. Super. Ct. Dec. 16, 1983) (refusing to restrain hospital from
force-feeding a patient who wished to die); see also Dresser, Feeding the HungerArtists: Legal Issues
in Treating Anorexia Nervosa, 1984 WIs. L. Rav. 297 (analyzing legal grounds for force-feeding
anorexics).
46. If mentally competent and sufficiently aware, patients could initiate such decisions. However,
it is hard for patients to focus on available options unless someone isolates these options from business
as usual and makes them visible. The tendency toward passivity and inertia among the ill is widely
acknowledged. See Parsons, Epilogue to THE DocTOR-PATIENT RELATIONSHIP IN THE CHANGING
HEALTH SCENE 445-46 (E. Gallagher ed. 1978).
Although some doctors may voluntarily consult with patients, little in their professional training
will encourage them to do so. Professor Katz analyzes doctors' disinclination to initiate consultation as
"a systematic and intentional omission based upon deeply held professional beliefs that silence is in
the patient's best interest." J. KATZ, supra note 7, at 58.
The President's Commission for the Study of Ethical Problems in Medicine and Behavioral Research found that only 52% of physicians say they would initiate discussion of resuscitation with a
patient in the last stages of a degenerative disease; 38% said they would not do so. 2 MAKING DECIStONS, supra note 2, at 226. By contrast, 79% of the public feels that a decision between aggressive
and supportive therapy should be made by the patient. Only 12% feel the physician should make the
decision unilaterally and 8% think it should be made jointly; 24% believe the patient should control. J.
KATZ, supra note 7, at 224-25. Professor Katz notes the special difficulties faced in communicating
honestly about death. Id. at 215-25. See also Comment, A Structural Analysis of the PhysicianPatient Relationship in No-Code Decisionmaking, 93 YALE L.J. 362 (1983) (patient should control
decisions whether or not to resuscitate through system of informed consent).
47. E.g., Natanson v. Kline, 186 Kan. 393, 406, 350 P.2d 1093, 1104 (1960) ("Anglo-American
HeinOnline -- 95 Yale L.J. 231 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
over physical contact they have left significant medical choices insulated
from patient control. Although safeguarding the body's physical perimeter
is important, medical care is affirmatively sought, and contact with the
body is a relatively unexceptional aspect of that care. The maintenance of
personal autonomy in a situation of dependence on expert knowledge and
skill is both more subtle and more significant. If the key issue is knowledge and choice regarding the fate of one's body, there is no meaningful
difference between a decision that will be implemented by touching the
body and one that will be implemented without doing so. Physical invasions have symbolic importance, and they constitute one important class of
situations in which autonomy interests are involved. To treat that subcategory as co-extensive with the autonomy interest as a whole, however,
creates grave deficiencies in the protection of the broader interest.
B.
Negligence Doctrine Embeds Protection of Patient Choice Within the
Interest in Physical Well-Being
Informed consent is a subcategory of professional negligence doctrine.",
Standard negligence analysis protects an interest in physical well-being.
The doctrine of informed consent injects into the established framework of
negligence a concern with patient choice that would otherwise be absent.
It recognizes that one way that actionable physical injury may occur is
through the failure to disclose information that would have resulted in
non-consent to treatment. The concern with choice does not, however, rise
to the level of a fully protected interest under negligence doctrine. Rather,
choice remains encapsulated within the dominant interest in physical
well-being.
The subordination of choice that results from its submergence within
the negligence analysis of physical injury reflects a pervasive fear that
plaintiffs making such claims will recover when they have not "really"
been injured, or that doctors will be held liable when they have not "really" done anything wrong. Moreover, the tendency to assume that reasonable people choose as their doctors tell them to leads to a conclusion
that patients' choices will in any event mimic professionals' competent
choices. Any "real" and deserving injuries would then be remedied under
standard (i.e. non-choice oriented) professional negligence analyses. The
importance of separately protecting patient autonomy would greatly
diminish.
The following sections trace the distortion of the interest in choice that
law starts with the premise of thorough-going self-determination."); Pratt v. Davis, 118 Ill.
App. 161,
166 (1905), affd, 224 Ill.
300, 79 N.E. 562 (1906) (consent prior to surgery is essential to maintenance of right to inviolability of individual).
48. PROSSER & KEETON, supra note 18, § 32, at 189-93.
HeinOnline -- 95 Yale L.J. 232 1985-1986
Protecting Patient Choice
takes place at each stage of the standard negligence analysis and demonstrate that although the chain of assumptions that informs that analysis
may be accurate with reference to the prototype, these assumptions are
false as applied to nonprototypical cases.
1. Duty to Disclose: Physical Contact Revisited
Because informed consent is the doctrinal category that introduces a
concern about patient autonomy into the standard negligence analysis,
whether or not a case is classified as one of informed consent often determines whether patient autonomy will receive any protection. Although
many aspects of informed consent have been exhaustively discussed, little
attention has been paid to when the doctrine is applied. The physical parameters of battery analysis are definitional and unsurprising. More unexpected is the fact that, as this section will demonstrate, negligence analysis also uses physical contact to determine when to impose a duty to
disclose for purposes of securing informed consent. Negligence analysis
may employ physical contact as a limiting device on professional duty because habits of thought have been carried over from battery. Or, as discussed above, the limitation may reflect suspicion about the legitimacy of
the informed consent action.4 9 Whatever the reason, adoption of physical
contact as the triggering concept for negligence duties creates additional
gaps in the protection afforded to patient choice.
a. The Duty to Disclose Will Be Abandoned
A New York decision, Karlsons v. Guerinot,50 illustrates how a duty of
disclosure may be abandoned where facts do not fit the standard prototype
of a proposal to touch. The plaintiff, a thirty-seven-year-old woman, alleged negligent care based on her doctor's failure to perform amniocentesis, a procedure that would have identified Down's Syndrome in her
fetus early enough to perform an abortion. She also alleged denial of informed consent based on her doctor's failure to inform her of the existence
of the amniocentesis procedure. Allegedly as a result of these failings, she
bore a defective child whom she would rather have aborted. She and her
husband sought damages for pain and suffering, mental anguish, and for
medical and other expenses of rearing the child."'
In Karlsons the court held that, in the absence of any proposed touch49. Such an approach would be comparable to special limiting requirements on actions for emotional distress. See RESTATEMENT (SECOND) OF TORTS § 46(2)(b) (1965).
50. 57 A.D.2d 73, 394 N.Y.S.2d 933 (1977).
51. Although several other causes of action are alleged in the case, id. at 75, 394 N.Y.S.2d at 934,
these are the ones crucial to the analysis here.
HeinOnline -- 95 Yale L.J. 233 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
ing or invasion of the body, no issue of informed consent could properly be
raised:52
Although [our earlier] pronouncement of the scope of the [informed consent] doctrine seems broad on its face, its application has
consistently been limited to those situations where the harm suffered
arose from some affirmative violation of the patient's physical integrity such as surgical procedures, injections or invasive diagnostic
tests.53
Resting its conclusion on both common law and statutory grounds," the
court affirmed the lower court's dismissal of the informed consent cause of
action. 5
How serious a problem was the dismissal of the informed consent
count? 56 In instances of invasive touching, battery doctrine dictates that
patients are to make decisions about their own care. A doctor's responsibilities for professionally competent care must then be to know, to advise
and recommend, and to implement, but not to decide. 57 1 have argued that
autonomy interests are as compelling where non-invasive decisions are involved as they are where touching is proposed. However, where no touching is proposed, patient decisionmaking may be ignored not only under
52. Wrongful birth actions have surfaced in the wake of new genetic technology and altered public policy. Although they were originally treated as presenting standard medical malpractice and consent issues, special rules have now emerged in these cases. See generally Capron, Tort Liability in
Genetic Counseling, 79 COLUM. L. REv. 618 (1979) (examining liability for birth of child with
condition that if diagnosed early enough would have led parents to avoid such birth); Collins, An
Overview and Analysis: PrenatalTorts, Preconception Torts, Wrongful Life, Wrongful Death and
Wrongful Birth: Time for a New Framework, 22 J. FAM. L. 677, 695 (1983-84) (courts agree parents may state some cause of action based on wrong information, but disagree on proper measure of
recovery); Robertson, Civil Liability Arising from "Wrongful Birth" Following an Unsuccessful
Sterilization Operation, 4 AM. J.L. & MFD. 131 (1978) (analyzing actions in both tort and contract
for wrongful birth). In Part III, infra, I will argue that the need for different rules is not as situationspecific as analysis in these cases has suggested.
53. 57 A.D.2d at 87, 394 N.Y.S.2d at 939 (citations omitted); accord Malloy v. Shanahan, 280
Pa. Super. 440, 443, 421 A.2d 803, 804 (1980) ("The doctrine of informed consent has been applied
only to suits involving surgical operations or procedure, wherein 'an operation without the patient's
consent is a technical assault.' ") (quoting trial court, in turn quoting Gray v. Grunnagle, 423 Pa.
144, 155, 223 A.2d 663, 669 (1966)).
54. The court acknowledged that the statute it cited was not in effect at the time these facts took
place. Karlsons, 57 A.D.2d at 82 n.4, 394 N.Y.S.2d at 939 n.4. Nevertheless, the court viewed the
statute as providing an indication of legislative inclinations.
55. The Karlsons court also rejected the plaintiff's theory that disclosure of the procedure was
required under the branch of informed consent dealing with disclosure of alternatives. See infra Part
II(C).
56. Some patients may know about amniocentesis, unlike other medical information, independently of medical sources. The possibility of independent knowledge, however, should not reduce the
scope of the doctor's duty. Rather, it should constitute an affirmative defense, with the doctor bearing
the burden of proving the plaintiff's actual knowledge or the applicability of a common knowledge
exception.
57. For a discussion of battery rules, see supra Part I(C).
234
HeinOnline -- 95 Yale L.J. 234 1985-1986
Protecting Patient Choice
battery analysis, but also under negligence doctrine. Where choice is not
identified as an interest, courts will impose no duty to disclose in order to
inform the patient's consent. Yet, in these instances, receiving information
from the doctor is the only way a patient can become aware of a pending
choice. Nondisclosure here is tantamount to loss of the choice interest itself; only issues of professionally competent care remain.
Karlsons illustrates precisely that outcome. Having denied the existence
of any issue of informed consent because of the absence of a proposal to
touch, the court observed:
[T]he alleged undisclosed risks did not relate to any affirmative
treatment but rather to the condition of pregnancy itself. Allegations
such as these have traditionallyformed the basis of actions in medical malpractice and not informed consent.5 8
Using the framework of malpractice, the court only inquired into
whether the treatment that the doctor recommended and actually implemented constituted professionally competent care of physical well-being.
Thus, the sole issue regarding amniocentesis that the court allowed to be
considered was: Did defendant doctors breach their duty of professional
care "by not properly diagnosing the condition of the child"?"
The doctor's failure to recommend amniocentesis (and to implement it,
thereby discovering those facts that the test would have revealed) might
have been argued to be a failure of competent care. However, it is quite
likely that what the court formulated as "not properly diagnosing the condition of the child" 60 would not have constituted professionally incompetent care, particularly in 1973.1 Amniocentesis involves risks to the fetus.
The test was not uniformly used in 1973, and the doctors could competently have held differing opinions about whether such a test should have
been recommended.6 2 Moreover, non-medical values about the ultimate
issue, abortion, are so disproportionately important in the chain of decisions that the medical risks of amniocentesis pale by comparison. Thus,
unless such recommendations are made purely on the basis of such medical matters as risk to the fetus, there is significant danger that a doctor's
advice regarding amniocentesis might improperly be influenced by his
personal values with regard to the issue of abortion. An expression of such
58. 57 A.D.2d at 82, 394 N.Y.S.2d at 939 (citations omitted and emphasis added).
59. Id. at 78, 394 N.Y.S.2d at 936.
60. Id.
61. See Capron, supra note 52, at 670-71. Professor Capron discusses the difficulty of establishing a standard of practice regarding new fields or procedures like genetic counseling. Id. at 620-25.
Precisely the same facts that make difference of professional judgment acceptable make the case for
disclosure and patient decision more compelling. See infra Part III.
62. See discussion of the limits of competence regulation infra Part III.
HeinOnline -- 95 Yale L.J. 235 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
an opinion might be appropriate if the patient requested it, but absent
such a request, the doctor might well claim that he properly made no
recommendation regarding amniocentesis as a prelude to possible
abortion.
For these reasons, it might have been both easy and legitimate for the
doctor in Karlsons to defeat any allegation that failure to diagnose the
fetal abnormality constituted professionally incompetent care.6 3 Yet there
should have been an additional issue derived from the concern with patient choice: Did the defendant doctors breach their duty to protect the
patient's autonomy interest by not informing her of the existence of a test,
amniocentesis, which could have been used, if the patient so chose, to detect fetal abnormalities? Protection of that interest was lost when the absence of a proposal to touch prompted the court to eliminate issues of
64
disclosure from the case.
Injuries arising from invasion of the interest in choice may be factually
similar to injuries arising from failures of competent care. In Karlsons, if
the doctor's failure to recommend amniocentesis had been judged to constitute professionally incompetent care, the resulting harm would have
been the same as if injuries resulted from an invasion of the interest in
choice: The parents would have had a child they would not otherwise
have had. The possibility of such overlap may easily be construed as
meaning that informed consent is nothing but a second, easier bite at the
same apple of malpractice recovery.6 5 Yet, although the injuries overlap,
the analysis differs, because the interests at stake are different. Liability
may legitimately be found under one analytic theory, although there is
none under the other. If both concerns are to be vindicated, it is crucial to
keep the two issues analytically distinct.
Where Karlsons illustrates loss of the choice interest in a pre-care diagnostic setting, Kelton v. District of Columbia"6 reflects an analogous gap
in analysis of post-care nondisclosure. Six years after the delivery of her
63. The result would likely be tougher on the doctor if the issue were failure to know of amniocentesis. Such a failure to know of the procedure probably would be found professionally incompetent. Presently, issues of knowledge are often left unexamined because rules do not require inquiry
beyond what the doctor recommended or did. The result may be that both incompetence (of knowledge) and inappropriate substitution of judgment go unremedied. For a discussion of the limits of
competence regulation, see infra Part III. In any case, no such issue of the doctor's knowledge was
formulated in Karlsons.
64. Professor Capron alludes to this problem but does not pursue it except to suggest that the
goals of informed consent are applicable even though there is a lack of touching. Capron, supra note
52, at 629-30 n.36.
65. There is a suggestion of such attitudes in Meisel's description of how informed consent functions, Meisel, Expansion, supra note 6, at 74-77, and in Epstein's critique of the doctrine in which
he states that there are "simply no principled limits to a doctrine of informed consent," Epstein, supra
note 17, at 124.
66. 413 A.2d 919 (D.C. 1980).
HeinOnline -- 95 Yale L.J. 236 1985-1986
Protecting Patient Choice
second child by Caesarian section, the plaintiff underwent exploratory
surgery to determine why she had been unable to become pregnant a third
time. Her surgeon discovered that there were scars on her fallopian tubes
that were consistent with the performance of a tubal ligation or some surgical trauma. The plaintiff sued the doctor who delivered her second child.
She alleged that she had not consented to any tubal ligation, nor had she
been told of any intentional or accidental surgical intervention affecting
her fallopian tubes. Her complaint alleging an unconsented sterilization
(battery) was barred by the statute of limitations. However, the plaintiff
also claimed that the doctor had breached a separate duty to tell her of the
damage to her fallopian tubes, and that that claim regarding nondisclosure
was timely under negligence limits.
If the plaintiffs allegations were true, she had been injured in several
ways. She had been deprived of fertility without her knowledge or consent. The actions that produced this physical harm may have constituted
either battery (if intentional) or negligence (if accidental). But the failure
to disclose both the physical facts and the doctor's knowledge regarding
how the damage came about constituted a different and additional harm.
Even after the surgical trauma took place, Ms. Kelton still had a prospective interest in choice. Had she known about the scarring of her tubes, she
might have sought surgical repair. At the very least, she would not have
had to undergo additional surgery to determine why she was unable to
conceive. Disclosure of the scarring would also have allowed her to exercise choice regarding non-medical consequences of the medical facts. Properly informed, Ms. Kelton might have sought to adopt further children, or
she might have elected to file suit for the initial injury (the unconsented
67
ligation).
The Kelton court, like the Karlsons court, allowed a touch-based definition of disclosure requirements to eliminate the patient's interest in
these choices. Distinguishing a leading informed consent case, Canterbury
v. Spence,68 as involving a failure to disclose risks of prospective surgery,
the Kelton court upheld a dismissal under timeliness provisions on the
ground that there was no cause of action other than battery.6 The court
67. Loss of opportunity to file the battery action may not have been attributable to the doctor's
failure to disclose the injury that had occurred. The plaintiff did not file a timely battery action when
she did later discover the injury. It is not necessarily the case, however, that a similar delay would
have occurred had she been informed earlier. In any event, the other injuries to her interest in choice
could not be so easily dismissed.
68. 464 F.2d 772 (D.C. Cir.), cert. denied, 409 U.S. 1064 (1972).
69. The court may have seen the disclosure cause of action as a subterfuge to get around the
statute of limitations in the "real" battery action. If the court was hostile to the disclosure action for
this reason, it underscores the tendency to "see" choice-based disclosure issues within a narrow, precut pattern. For discussion of the duty to disclose malpractice, see Vogel & Delgado, To Tell the
Truth: Physicians' Duty to Disclose Medical Mistakes, 28 UCLA L. REv. 52 (1980) and infra Part
HeinOnline -- 95 Yale L.J. 237 1985-1986
Vol. 95: 219, 1985
The Yale Law Journal
stated: "Thus a breach of duty to disclose [lack of consent] is not actionable in negligence unless it induces a patient's uninformed consent to a
risky operation from which damages actually result."" The court's statement implies the following: Consent and consent-oriented disclosure duties
arise only in surgery cases and only in advance of a proposed intervention,
and such duties extend only to disclosure of the risks posed by the proposed intervention itself. This formulation of the rule ignores various
71
types of disclosure essential to protection of patient autonomy. Thus,
while the court recognized that the patient's choice interest may have been
violated if a tubal ligation was intentionally performed without the patient's consent, it failed to understand that a choice interest was also implicated in the nondisclosure of the facts after the intervention had occurred. As in Karlsons, analysis of disclosures essential to the protection
of autonomy was foreclosed because the issue arose outside the touchbased parameters of the informed consent prototype.
b.
The Duty to Disclose Will Be Transposed
Where a case does not involve physical touching, the issue of disclosure
may not wholly disappear, as it did in Karlsons and Kelton. Instead, a
duty to disclose may be considered, but analysis of that duty will shift
when a case is not identified as one of informed consent. Where nothing
signals the involvement of patient autonomy, duties to disclose will be
transposed into issues about professional care of physical well-being. Roark v. Allen,72 a 1982 case decided by the Texas Supreme Court, illustrates this pattern.
The plaintiffs were the parents of a baby allegedly injured during a
breech birth. After the delivery, the defendant doctors noticed forceps indentations on the infant's head. They considered the possibility that the
indentations might indicate a fractured skull, but after further examination decided that there was no fracture. No one informed the parents of
the possibility of a skull fracture, nor were x-rays, which would have revealed the fracture, ordered. Weeks later, bilateral fractures of the infant's
skull were detected and the problem was then corrected. The parents sued
the obstetrician for negligent use of forceps and their family doctor for
III.
70. 413 A.2d at 922.
71. The court's statement also ignores the fact that failure to disclose information may in some
situations actually be a failure to provide competent care. See, e.g., Crosby v. Grandview Nursing
Home, 290 A.2d 375 (Me. 1972) (failure of physician to advise regarding care of injured foot). The
court probably makes this error because it assumes a usage whereby verbal communication necessary
to competent care of physical well-being is called "instruction" or "advice."
72. 633 S.W.2d 804 (Tex. 1982), affg in part, rev'g in part, 625 S.W.2d 411 (Tex. Ct. App.
1981).
HeinOnline -- 95 Yale L.J. 238 1985-1986
Protecting Patient Choice
failure to inform them of the possibility of skull fracture. The jury found
for the plaintiffs against both doctors."3
The failure of the family doctor to inform the parents of the possibility
of skull fracture is the primary concern here. The Texas Supreme Court
objected to the lower court's use of informed consent theories to analyze
this issue. According to the court, informed consent applies "only to medical procedures which have yet to be performed and . . . it is inapplicable
• . . where the patient has already undergone the proposed treatment and
been injured."7 4 Characterizing this type of nondisclosure as "a totally
different cause of action" from informed consent,7 5 the court analyzed it as
a problem of ordinary negligence and reversed the trial court's judgment
for the plaintiff. The Roark court's method of arriving at that decision
betrays the transposition of interests that occurs when informed consent
analysis is deemed inapplicable.
The plaintiffs and the lower courts had posed this nondisclosure as a
breach of a duty to protect the parents' interest in choice; they called it an
action for lack of informed consent. Once the supreme court decided that
informed consent doctrine did not apply, it abandoned the interest in
choice and analyzed disclosure in light of the traditional negligence interest in professionally competent care of physical well-being. The court observed that expert testimony was essential to evaluate the duty because
"diagnosis of skull fractures is not within the experience of the ordinary
layman. 17 6 This characterization shifted the issue from one of informing
about possible courses of action to one of diagnosing skull fractures. The
73. Id. at 807-08. The Texas Supreme Court affirmed the court of appeals' reversal of the verdict
against the obstetrician, albeit on a different ground (that although he was negligent in his use of the
forceps, the causal connection between this negligence and the baby's injuries was not sufficiently
established). Id. at 811. Roark illustrates that, even where there is a finding of incompetent care that
produces a harm factually similar to the harm resulting from invasion of autonomy, analysis of the
autonomy interest would not be redundant.
74. Id. at 808. The court seems unwilling to acknowledge that the distinction between pre-care
and pest-care is largely semantic. This case could be described as one involving a need for further
diagnosis and care as a result of birth injuries. Indeed the trial court used just such a phrasing. Id.
75. Id.
76. Id. at 809. At the time the facts occurred, Texas also required expert testimony to establish
the standard of care in informed consent actions. See Wilson v. Scott, 412 S.W.2d 299, 302 (Tex.
1967). The way the issues are posed in these two actions, however, differs. In an informed consent
action, the law itself establishes that there is a choice to be made by the patient; only then does it
require medical evidence regarding what a competent doctor would have disclosed to the patient making that choice. Thus Wilson, the leading Texas informed consent case, states: "Physicians. . . have a
duty to make a reasonable disclosure to a patient. . . based upon the patient's right. . . to exercise
an informed consent. . . . [That duty is measured by what a reasonable doctor] would have disclosed
to his patient about the risks incident to a proposed diagnosis or treatment. . .. " Id. at 301-02. If
the law had required the patient's consent (here to a recommendation that the indentations need not
be x-rayed), some disclosure, at least of the existence of a choice to be made, would have been necessary. By contrast, where no issue of informed consent is recognized, unless expert testimony is offered
regarding the need to disclose, there may be no information given about the very existence of a choice
to be made.
HeinOnline -- 95 Yale L.J. 239 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
defendant's own testimony had been used to establish the standard of care.
Asked whether the correct medical procedure would be to advise of the
possibility that such a fracture existed, he replied, "Yes, if the fracture
was assumed to be there." 77 The supreme court reversed a jury verdict for
the plaintiffs, saying there was no evidence to support it because "the
is uncontroverted that Dr. Allen assumed a fracture did not exevidence
ist."7 s8 What began as a question of informing patient choice ended as an
issue of the doctor's diagnostic accuracy. 9
Similarly, in Sinkey v. Surgical Associates,"0 the court began with an
alleged duty to inform the parents of a child patient (regarding the opinions of a consulting radiologist) and transposed it into a different issue:
Was the reading of the x-ray professionally competent? Again, the information was unconnected to any prototypical proposal to touch; the case
was not identified as one of informed consent. Yet an interest in choice
was probably the basis on which the plaintiff parents alleged that the
81
information should have been communicated to them. The Iowa Supreme Court affirmed a directed verdict for the defendant, saying:
The second allegation of negligence is based on an asserted duty to
advise the patient that the x-ray finding was consistent with appendicitis and that the radiologist's impression was appendicitis ...
We find no evidence to support the allegation of an incorrect
interpretationof the x-ray [by the attending doctor who diagnosed
tonsilitis] and hold that under these circumstances the doctor had
no duty to advise the patient of the fact that the82condition shown in
the x-ray was also consistent with appendicitis.
If the allegation were that the doctor should have diagnosed the appendicitis, the competence of the x-ray reading would be critical. Where patient
knowledge and choice are the concern, however, the competence of the
reading is not the issue.
77. Allen v. Roark, 625 S.W.2d 411, 415 (Tex. Ct. App. 1981), affd in part,rev'd in part,633
S.W.2d 804 (Tex. 1982).
78. Allen v. Roark, 633 S.W.2d at 809.
79. This decision may have been technically acceptable. The court, however, could easily have
taken the "yes" part of the defendant's answer as sufficient, particularly where it reversed the decision
under a "no evidence" standard. Even if failure to make such a recommendation was not shown to be
incompetent, as the court concluded here, the issue of disclosure for choice should still have remained
live. As between the doctor and the parents, it ought to have been the parents' choice whether to incur
risks and costs in order to determine with greater accuracy whether or not fractures were present.
80. 186 N.W.2d 658 (Iowa 1971).
81. Where professionally competent care is the interest which requires verbal communication, that
communication tends to be denominated "diagnosis," "advice," or "recommendation." Where the term
"disclosure" is used, it ordinarily conveys a concern about patient autonomy.
82. 186 N.W.2d at 661 (emphasis added).
HeinOnline -- 95 Yale L.J. 240 1985-1986
Protecting Patient Choice
Policies for evaluating the competence of doctors' diagnoses differ significantly from those involved in patient choice. The uncertainty of medical judgment legitimately excuses a failure to diagnose correctly, if the
judgment was carefully and reasonably made. Such uncertainty should
not, however, excuse a failure to disclose information that would have
permitted a patient to exercise a different choice. Indeed, the more uncertain the medical judgment involved, the more reason there is to excuse a
wrong diagnosis. But, the more uncertain the medical judgment, the less
acceptable is the doctor's substitution of her judgment for the patient's.
The courts in Roark and Sinkey applied notions of uncertainty appropriate to analysis of competent care rather than to analysis of disclosure to
protect choice. Had the absence of a prototypical proposal to touch not
prevented these issues from being recognized as questions of informed consent, the courts would, I think, have been less likely to make this
mistake.8 3
2.
Duty To Disclose-PhysicalContact Transcended?
A requirement to disclose alternatives is sometimes mentioned in informed consent cases. 84 If broadly applied and interpreted, such a requirement could close some of the identified gaps in protection of choice, but
the requirement to disclose alternatives is itself constricted by concepts
centered on touching. Many jurisdictions impose no such disclosure requirements.8 5 When they do, the mandate is typically to disclose only al83. The central tenet of classical informed consent doctrine is disclosure of risks, i.e., possible but
uncertain occurrences. Roark is a relatively pure type under which the interest in choice is wholly
abandoned. Sometimes when a case not falling within the prototypical boundaries of informed consent
is treated as an instance of ordinary negligence, it is unclear whether the court analyzes disclosure as a
duty derived from an interest in choice or from one in professionally competent care. Stills v. Gratton,
55 Cal. App. 3d 698, 127 Cal. Rptr. 652 (1976), provides an interesting example. The court's selection and discussion of facts, duty, standard of care, and damages reflect alternating and somewhat
inconsistent concerns first with choice and then with competent care. Although, as a case treating
disclosure under ordinary negligence doctrine, Stills provided a greater than average degree of protection for patient choice, that result was shaped by several atypical factors. The case involved the completeness of an abortion, a subject about which sensitivity to choice has been heightened. See infra
Part III. Further, because two different doctors were involved, disclosure was factually separated from
professional care. The same result might not be forthcoming in other circumstances.
84. E.g., Scott v. Bradford, 606 P.2d 554, 557 (Okla. 1979); Siegel v. Mount Sinai Hosp., 62
Ohio App. 2d 12, 21, 403 N.E.2d 202, 209 (1978); Natanson v. Kline, 186 Kan. 393, 410, 350 P.2d
1093, 1106 (1960).
85. According to Andrews, only 10 of 23 statutes governing informed consent specifically require
that the doctor disclose alternatives to treatment. Andrews, Informed Consent Statutes and the Decisionmaking Process, 5 J. LEGAL MED. 163, 197 (1984). Only a few cases squarely hold a doctor
liable for failure to disclose alternatives. One such case is Archer v. Galbraith, 18 Wash. App. 369,
379, 567 P.2d 1155, 1161 (1977). But cf. Dunham v. Wright, 423 F.2d 940, 946 (3d Cir. 1970)
(disclosure of alternatives not actionable when doctor determines patient has "no alternative"). Even
when the issue is raised, courts often leave it to the jury, making no requirement as a matter of law.
See, e.g., Masquat v. Maguire, 638 P.2d 1105 (Okla. 1981) (for discussion of case, see infra note
133).
HeinOnline -- 95 Yale L.J. 241 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
ternative treatments, not alternative diagnostic data, theories of the case,
or courses for case management.8 6 Disclosure requirements are rarely ap87
plied to situations of aftermath disclosure, to situations in which no
change in continuing treatment is proposed, or to situations where either
no treatment is possible or none is proposed. Thus, the alternatives rule
does not extend far enough to encompass many situations identified here
as gaps in autonomy protection.
Developments in two states suggest potentially broader changes in the
limitations that derive from negligence requirements for physical contact.
In Gates v. Jensen,88 the Washington Supreme Court selected the doctor's
possession of knowledge, rather than his proposal to touch, as the occasion
for a duty to inform the patient's choice. The court held that an ophthalmologist had an obligation to inform the patient of test results showing
possible glaucoma even though he did not propose either to test further for
that condition, or to treat it.
The jury found for the doctor on the plaintiff's claim that the ophthalmologist's failure to diagnose her glaucoma was professionally incompetent. Yet the doctor's failure to tell the patient that her tests showed borderline symptoms of the disease, and that further testing for it was
Andrews notes that only 14 percent of physicians surveyed by the President's Commission considered information about alternatives to be integral to informed consent. Andrews, supra, at 197. The
enactment of statutes specifying exactly what alternatives are to be disclosed in particularly controversial circumstances provides further evidence of the inadequacy of general statutory or common law
requirements. E.g., CAL. HEALTH & SAFETY CODE § 1704.5 (West Supp. 1985) (failure to disclose
alternative treatments of breast cancer constitutes unprofessional conduct). For a further discussion of
the inadequacy of the present alternatives rules, see infra Part II().
86. For example, in one Washington case, Thornton v. Annest, 19 Wash. App. 174, 574 P.2d
1199 (1978), the defendant doctor performed a hysterectomy during exploratory surgery on the plaintiff, who had severe pelvic inflammatory disease. The plaintiff appealed a jury verdict for the defendant on the ground that the doctor's total failure to disclose alternatives should have entitled her to a
partial directed verdict on the issue of informed consent. Concluding that the plaintiff's alleged alternatives were alternative methods of diagnosing rather than alternative treatments, the court upheld
denial of the directed verdict. It stated that "[o]nly feasible and available treatment must be disclosed."
Id. at 179, 574 P.2d at 1203. Under the court's analysis, no disclosure of alternative approaches to
diagnosis is required, at least as a matter of law. As in Karlsons, when disclosure disappeared from
the analysis, only the interest in professionally competent care remained. The court held that the jury
might legitimately have concluded that the defendant had complied with the proper standard of care
"in performing the exploratory surgery as a method of diagnosing plaintiff's symptoms." Id. at 180,
574 P.2d at 1203 (emphasis added).
Many cases discussed in this article (Karlsons, Roark, Sinkey, Keogan) involve disclosure about
alternative diagnostic theories; others could be so characterized if artificial pre/post-care distinctions
were abandoned (Kelton). In these cases the notion of disclosing alternatives did not convince the
courts to accept an informed consent theory of the case. In Karlsons, disclosure under an alternatives
theory was expressly rejected.
87. Judgments made by a doctor in the aftermath of medical intervention could be described as
being the diagnostic process for a potential next stage of decision. Thus diagnosis could, at its most
expansive, mean "alternative knowings." As this Article shows, however, courts have tended to freeze
concepts like "diagnosis" and "consent" into narrow and literal meanings. Some more flexible and
generic concept would be necessary to trigger adequate disclosure.
88. 92 Wash. 2d 246, 595 P.2d 919 (1979), rev'g 20 Wash. App. 81, 579 P.2d 374 (1979).
HeinOnline -- 95 Yale L.J. 242 1985-1986
Protecting Patient Choice
possible, deprived her of the choice whether to consent to his recommendation that no further investigation of glaucoma be undertaken. Under the
traditional approaches discussed above, the interest in that choice might
well have been ignored.
No battery action would lie in Gates because there was no unconsented
touch. Nor, under the interpretation employed by cases like Karlsons,89
Kelton"0 or Roark,91 would there be an obligation to disclose under informed consent doctrine. In accord with such interpretations, the defendant in Gates claimed that because he was still engaged in diagnostic assessment and had not yet recommended treatment, no duty to disclose had
arisen. 92 But unlike courts in the cases discussed above, the Washington
court did not allow the absence of a proposal to touch to eliminate the
patient's interest in disclosure and choice.93
Although it purported only to apply existing informed consent law, the
Washington court articulated what might have become a significant new
standard: It ordered that a doctor should disclose whenever he has knowledge of a potentially dangerous abnormality in the patient's body." Under
this test, the occasion for disclosure would not be a proposed act of touching, but the doctor's possession of significant knowledge about the medical
condition of the patient. The Gates standard, if allowed to develop, could
have superceded touch-based boundaries that limit protection of patient
choice interests to particular stages of treatment or to types of proposed
intervention.
89. 57 A.D.2d 73, 394 N.Y.S.2d 933 (1977). See supra text accompanying notes 50-65 (discussing Karlsons).
90. 413 A.2d 919 (D.C. 1980). See supra text accompanying notes 66-71 (discussing Kelton).
91. 633 S.W.2d 804 (Tex. 1982). See supra text accompanying notes 72-79 (discussing Roark).
92. The intermediate court explicitly adopted this view in affirming the trial court's refusal to
give informed consent instructions. 20 Wash. App. at 87, 579 P.2d at 377. Timeline factors per se
continue to be suggested by some authors as appropriate considerations in creating duties of disclosure. See Comment, Informed Consent in Washington: Expanded Scope of Material Facts that the
Physician Must Disclose to His Patient, 55 WASH. L. REv. 655, 667-70 (1980).
93. Unlike the court in Thornton, 19 Wash. App. 174, 574 P.2d 1199 (1978) (for discussion of
case, see supra note 86), the court in Gates also refused to distinguish between diagnostic theories and
proposals to treat for purposes of triggering a patient's interest in disclosure. Gates, 92 Wash. 2d at
250, 595 P.2d at 922.
94. In its summary of the facts, the court mentions that Ms. Gates asked about the result of the
doctor's tests and was told that everything was all right, but the court does not characterize the problem as misrepresentation. The court phrased its broadened concept of an affirmative duty as follows:
"The existence of an abnormal condition in one's body, the presence of a high risk of disease, and the
existence of alternative diagnostic procedures to conclusively determine the presence or absence of that
disease are all facts which a patient must know in order to make an informed decision on the course
which future medical care will take." 92 Wash. 2d at 251, 595 P.2d at 923 (emphasis added). Although an improvement on a touch-based trigger, this formulation is still not optimal. It might not,
for instance, encompass disclosure of the availability of amniocentesis.
Although the Gates court purported only to apply the rule of Miller v. Kennedy, 11 Wash. App.
272, 522 P.2d 852 (1974), affd, 85 Wash. 2d 151, 530 P.2d 334 (1975), affid en banc, 91 Wash. 2d
155, 588 P.2d 734 (1978), it is Gates which created a new standard not dependent on touching.
HeinOnline -- 95 Yale L.J. 243 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
Gates was decided in a context of intense conflict over the adequacy of
95
standards of professional negligence. As a by-product of its dissatisfaction with the limitations of traditional competence regulation, the court
seems to have become particularly sensitive to the unfairness entailed in
9
relying on the judgment of professionals instead of patients. " However,
faced with new medical facts, the same court, in Keogan v. Holy Family
Hospital,9" 7 sharply limited the potential of Gates. In circumstances analytically similar to Gates,9" five justices refused in Keogan to impose a
99
duty of disclosure based on possession of knowledge. Instead, like the
court in Karlsons, they reverted to the assumption that, apart from instances involving touching, regulation of the doctor's competence was the
only important issue. 100
Gates has been criticized as interfering with proper professional judgment, indulging in the fantasy that a patient can "correct the reasonable
errors of his physician." 10 Yet Gates glimpsed what Keogan lost sight of:
Although the patient is not more competent in making judgments assessing the likelihood of a particular disease, she is more competent in deciding whether she wishes to undergo more tests and spend more money in
order to be more certain about the diagnosis in her case.
95. In Helling v. Carey, 83 Wash. 2d 514, 519 P.2d 981 (1974), the court had held that reasonable prudence may require a standard of care higher than the prevailing professional standard. The
state legislature responded by enacting WASH. REv. CODE ANN. § 4.24.290 (West Supp. 1986), requiring a plaintiff to demonstrate that the practitioner "failed to exercise that degree of skill, care, and
learning possessed at that time by other persons in the same profession . . . ."The Gates court,
however, interpreted the statute in a fashion that allowed it to continue to apply the doctrine of
Helling. See 92 Wash. 2d at 253-54, 595 P.2d at 924.
96. Because doctors are incompetent only when they fail to act as other doctors would, there is no
external referent by which to judge what are, essentially, questions of utility, e.g., how much certainty
about the presence of glaucoma is adequate. The patient could provide a determinate referent by
deciding the personal value of additional tests or procedures. Another source for such evaluations
would be social decisions allocating collective resources through collective cost-benefit analyses. In
Gates the court used both tools. It expanded a doctor's obligation to respect patient choice, and it
permitted 'a societal cost-benefit analysis of the adequacy of care by allowing a jury to decide that
reliance on the prevailing professional standard was insufficient.
97. 95 Wash. 2d 306, 622 P.2d 1246 (1980).
98. A doctor had obtained results suggesting but not confirming that heart disease was a possible
cause of plaintiff's chest pains. As in Gates, the doctor concluded that problems other than heart
disease lay at the root of the patient's symptoms, and made no disclosure of the possibility of heart
disease or of the test results. The patient later died from a heart attack. Id. at 307-08, 622 P.2d at
1249.
99. Id. at 330-31, 622 P.2d at 1261. The precedential effect of the opinion is uncertain. Justice
Horowitz's opinion, purporting to be the opinion of the court, was joined by two other justices on the
matter of disclosure under Gates. The dissent, refusing to apply Gates, however, was signed by five
justices. A rehearing was granted, but the case was settled before the rehearing was held. Thus Keogan did not directly reverse Gates, but it did undermine the holding.
100. Five justices signed an opinion stating that, "[i]f Dr. Snyder was negligent because he should
have discovered Keogan's diseased heart and failed to do so, that is what should be alleged and proved
.... This court with its benefit of hindsight should not now enter the fray. . . with rulings as a
matter of law as to what the doctor should have told the patient." Id.
101. Comment, supra note 92, at 673.
HeinOnline -- 95 Yale L.J. 244 1985-1986
Protecting Patient Choice
Several recent California cases have also taken steps away from physical contact as the definitive occasion for choice-oriented disclosure. In
Truman v. Thomas,102 a thirty-year-old woman several times declined to
have a Pap Smear that her gynecologist recommended, sometimes saying
she could not afford it, sometimes saying simply that she did not feel like
it. She eventually died of cervical cancer, and a wrongful death suit was
filed alleging the doctor's negligent failure to inform her of the risks of
refusing the test. The majority concluded that the case was controlled by
Cobbs v. Grant,10 3 the leading California case on informed consent. It
held that instructions should have been given allowing the jury to consider
whether the doctor had breached a duty by failing to inform the patient of
the risks of refusing the recommended test.1 0 '
Although the majority appropriately rested its holding on Cobbs, the
facts in Truman differed from the traditional informed consent facts to
which Cobbs is typically applied. The fact that the patient decided not to
have the test meant that the doctor would not touch her. Both the court of
appeal majority and the supreme court dissenters stressed that requiring
disclosure in such a situation would be a serious and undesirable expansion of the basic doctrine. 0 5 In their view, apart from traditional requirements for consent to physical contact, people should do as their competent
doctors tell them,10 6 and protection against incompetent care is therefore
102. 27 Cal. 3d 285, 611 P.2d 902, 165 Cal. Rptr. 308 (1980), rev'g 155 Cal. Rptr. 752 (Ct.
App. 1979).
103. 8 Cal. 3d 229, 502 P.2d 1, 104 Cal. Rptr. 505 (1972).
104. Truman, 27 Cal. 3d at 294-95, 611 P.2d at 307-08, 165 Cal. Rptr. at 313-14. A Pap
Smear involves physical touching and would therefore require both consent and information about
risks. However, no disclosure of remote or minor risks is required. A Pap Smear is virtually risk-free.
The risks attach primarily to refusal to undergo the recommended test. It is therefore, technically, the
alternatives branch of informed consent doctrine that actually requires disclosure. Refusal is a selfevident alternative, but the risks of not having a Pap Smear might not be common knowledge and,
therefore, ought to be disclosed.
105. 155 Cal. Rptr. at 757-59; 27 Cal. 3d at 297-301, 611 P.2d at 909-11, 165 Cal. Rptr. at
315-17.
106. According to the court of appeal majority, "[ilt does not follow that the doctor should be
required to protect his patient from the patient's lack of judgment." 155 Cal. Rptr. at 757. Furthermore, "[ilt is nonsensical to claim that [a patient] goes to the doctor for advice he will not thereafter
follow. . . ." Id. at 759. Similarly, according to Judge Clark, writing for the supreme court dissenters, "[when a patient chooses a physician, he or she obviously has confidence in the doctor and
intends to accept proffered medical advice [, and it is] reasonable to assume that a patient who refuses
advice is aware of potential risk." 27 Cal. 3d at 299, 611 P.2d at 910, 165 Cal. Rptr. at 316.
The supreme court dissenters raise one legitimate issue. There seems a danger here that doctors
will be held liable if they fail to get patients' consent, or if a "bad" outcome results from honoring the
patient's choice. The Truman majority forestalled that problem by holding that if the plaintiff would
have refused the test even if fully informed, no liability should result. 27 Cal. 3d at 294, 611 P.2d at
907, 165 Cal. Rptr. at 313. Thus, although it ordinarily uses a reasonable person standard of causation in informed consent cases, the court adopted a subjective person standard here. That decision
handled the immediate problem but failed to address the analogous unfairness created by the use of
the reasonable person causation standard in the more usual informed consent case. See also infra text
accompanying notes 127-31, 212-13, 294-309.
HeinOnline -- 95 Yale L.J. 245 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
7
all that should be required. The closeness of the outcome in Truman,""
decided in a jurisdiction where concern for patient autonomy has traditionally been vigorous, suggests that the perception of an essential nexus
between physical contact and choice is still very strong.
On its facts, Truman created a rather modest extension of duty. How108
ever, a subsequent California case, Jamison v. Lindsay, described a potentially broader obligation to disclose. The plaintiff sued a pathologist for
falling to inform either the surgeon or herself of the presence of immature
tissue in a tumor removed from her body or of a controversy among pathologists over whether or not such tissue increases the odds of later malignancy. Again, because there was no proposed physical contact, the court
instrucof appeal upheld the trial court's refusal to give informed consent
10 9 Unlike
doctors.
defendant
the
for
verdict
jury
tions, and affirmed a
other courts, ° however, the Jamison court did not allow its conclusion
that informed consent did not apply to eliminate disclosure as a duty imposed to protect patient autonomy. It observed that, pursuant to the
broadened duties of disclosure suggested by Truman, an instruction regarding a duty to disclose information necessary for informed decisionmaking regarding whether to seek additional treatment following surgery
1
would have been appropriate." Such a duty would necessarily make possession of information rather than proposed physical contact the occasion
for disclosure.11 2 Despite the breadth of its theories, Jamison's rather
18
stringent procedural rulings barred any actual implementation of those
107. The intermediate court's 2-1 decision that no duty was appropriate, 155 Cal. Rptr. 752, was
Cal. Rptr.
reversed by a bare 4-3 majority of the supreme court, 27 Cal. 3d 285, 611 P.2d 902, 165
308.
108. 108 Cal. App. 3d 223, 166 Cal. Rptr. 443 (1980).
and
109. "(N]o treatment or tests had been proposed by respondents. The informed consent theory
proposed
appellant's proposed instructions on the duty of a physician to disclose. . . 'regarding the
not
. . . postoperative treatment' were inapposite [because] . . . [alfter the surgery, respondents did
decision."
propose any therapy as to which appellant would have been entitled to make an informed
Id. at 230, 166 Cal. Rptr. at 446.
110. See supra Section B(1).
that
111. 108 Cal. App. 3d at 231, 166 Cal. Rptr. at 447. Despite this observation, the court held
because the trial court had no responsibility to edit proposed instructions, no reversible error resulted.
Id.
112. Using that new standard, Jamison would require post-intervention disclosure in much the
same way that Gates extended pre-intervention disclosure.
113. For example, the plaintiff argued on appeal that the nondisciosure should have been analyzed under the reasonable person rather than the expert standard of care. Although agreeing that the
broadened duty of disclosure as judged by the lay standard could legitimately have been alleged under
held
ordinary negligence law rather than informed consent doctrine, the appellate court nevertheless
stanexpert
the
incorporating
instruction
an
requesting
by
error
the
invited
herself
that the plaintiff
dard of care. 108 Cal. App. 3d at 232, 166 Cal. Rptr. at 448. Yet, presumably, the plaintiff's lawyer,
having characterized the issue as one of informed consent, would have assumed that the lay standard
regarding disclosure would have been invoked under that theory, and therefore would have seen no
need to request such instructions with reference to the ordinary negligence aspect of the case. The
assumption that informed consent doctrine would apply does not seem unwarranted. The appellate
recogcourt itself cited Truman and Cobbs (both labeled as informed consent cases), and indicated its
HeinOnline -- 95 Yale L.J. 246 1985-1986
Protecting Patient Choice
theories in the case.114 It thus remains uncertain whether future plaintiffs
can use Jamison to achieve broader protection for patient autonomy.
These cases strain to transcend the traditional, touch-based limits on
disclosure duties protecting patient autonomy. The standards they suggest
would require more disclosure of a different genre of information. A
Gates or Truman standard would require that doctors disclose the process
and reasons by which they arrive at their recommendations, rather than
simply provide boilerplate warnings about recommendations that are to be
accepted as foregone conclusions. It would also require that, even if a doctor deems care satisfactorily completed, and thus recommends no further
treatment, she would have to disclose what she knows about the patient's
condition and prospects as a result of earlier interventions.
Such information could produce a different kind of patient participation
than has resulted from disclosure requirements under traditional informed
consent. 1 5 Rather than being a yes/no gatekeeper regarding a single
preselected option, 1 the patient could act on broader information that
would provide the basis for meaningful participation in medical decisionmaking. These cases envision a patient actually making choices, including
choices that differ from what the doctor deems "reasonable." 1 1 7
nition that the disclosure issue was essentially one involving patient choice. Yet the court refused
either to flex its interpretation of the scope of informed consent doctrine or to allow the plaintiff
instructions appropriate to a choice theory of the case, however labeled. When the court transferred
the disclosure analysis from informed consent to ordinary negligence, the parallel request for an
appropriate standard of care instruction should have been deemed similarly transferred.
The plaintiff also challenged the trial judge's instruction, which stated that the pathologist had a
right to make a professionally competent judgment between alternative medical procedures.
The
pathologist, the plaintiff argued, was choosing between beliefs or conclusions that should have been
disclosed. The court agreed that the alternative methods instruction was inappropriate. But because
the plaintiff did not properly raise this argument at trial, the court refused to consider it on appeal.
Id. at 233, 166 Cal. Rptr. at 448.
114. In another substantive ruling that prevented broadened disclosure, the court refused to apply
the principles of express warranty to the doctor's assertions about the plaintiff's condition, even
though he did not disclose known controversy regarding those assertions. Id. at 234, 166 Cal. Rptr.
at
449. Although the court's conclusion probably reflected traditional resistance to the use of contract
principles in medical cases, it was explained on premises linked to a touch-oriented model. The court
said the statement came after the surgery, and no consent to treatment was based on it. However,
from the vantage point of choice, the inaction that was based on the statement was just as significant.
115. Some commentators seem to have given up on the law's ability to create more than a pro
forma role for patients. The President's Commission, describing informed consent as having created
only a "duty to warn," 1 MAKING DECISIONS, supra note 2, at 20, seems to have adopted a rather
pessimistic view of the law's role, and has turned its attention to medical education and public opinion
to create the "shared decisionmaking" it envisions.
116. See id.at 24, 29 (litigation process forces examination of disclosure of risks concerning particular procedure rather than evaluation of entire doctor-patient relationship).
117. Thus in Truman the patient made her "unreasonable" choice at least partly for economic
reasons. The Pap Smear seemed of less value to her than it seemed to the doctor and she rejected
his
advice. 27 Cal. 3d at 290, 611 P.2d at 904, 165 Cal. Rptr. at 310. In some sense the case is like
that
of a Jehovah's Witness whose refusal of a blood transfusion is honored. See, e.g., In re Melideo,
88
Misc. 2d 974, 390 N.Y.S.2d 523 (Sup. Ct. 1976). But our willingness to give public recognition to the
conflicts of value in the two cases may differ. Although religious conviction is a palatable as well as
a
constitutionally protected reason for being "unreasonable," we are less sure about lack of funds,
a
HeinOnline -- 95 Yale L.J. 247 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
Although the expansions of duty undertaken in Gates, Truman and
Jamison could be significant, the potential impact of these three cases is
questionable. Although not directly overruled, Gates was emasculated by
Keogan. Truman requires disclosure about a course of inaction, but its
informed refusal stance is unusual and renders it vulnerable to a narrow,
touch-oriented interpretation. The relevant parts of Jamison are dicta, expounded by an intermediate court, and the case has not been followed or
much discussed. There is little indication of these approaches being
adopted in other jurisdictions. If the duty suggested in these cases were
actually to take root and develop, it would increase protection for patient
autonomy. Whether even such an expanded protection would be adequate
to ensure patient choice, however, is the subject of the next section.
Other Elements of the Analysis Which Distort Patient Choice
Although the Gates-Truman approach could bring some instances of
patient choice not involving physical contact within the ambit of negligence duties, adoption of this approach would not affect the distortion of
choice that occurs at other stages of the informed consent analysis. Informed consent doctrines remain embedded within the different, and often
inconsistent, interest in physical well-being. Although various of the spe-8
cific doctrinal rules of informed consent have been extensively criticized,"
the role played by interest definition in diluting the vindication of autonomy has not been adequately challenged.
3.
a. Standard of Care
Advocates of autonomy have argued that the standard for disclosure in
informed consent cases should be what a reasonable patient would want to
know rather than what the average competent doctor would actually disclose.119 Because doctors are trained to take active responsibility and are
choice will
vastly more frequent factor influencing patient choice. Increased protection of individual
care
highlight potentially unpleasant realities about the correlation of wealth and access to health
conunder our system. But the difficulty of those issues is no excuse for abandoning analytic clarity
cerning the relevant interests. See infra Part IV(C).
informed
118. Autonomy-oriented critics have objected to two principal aspects of the doctrine of
(criticizing
consent: the standards of care and of causation. See, e.g., J. KATz, supra note 7, at 48-84
of individprofessional standard of care and noting that causation requirements conflict with dignity
of care
ual and right to self-determination); Katz, Informed Consent, supra note 15, at 169 (standard
proceproposed
of
risks
establish
to
knowledge
medical
for
need
confuses
disclosure
to
respect
with
dures with need for medical judgment to establish limits of disclosure); Note, Restructuring, supra
partly
note 30, at 1555-59 (criticizing professional standard of disclosure). These rules have survived
because they are logical outgrowths of the way interests are defined under current doctrine.
119. See, e.g., J. KATz, supra note 7 (advocating disclosure based on patient self-determination
frank
rather than medical expertise); J. KATZ & A. CAPRON, supra note 30, at 114 (goal of full and
55
partnership between physician and patient); Comment, Informed Consent in Medical Malpractice,
Note, RestructurCALiF. L. Rav. 1396, 1407 (1967) (proposing full disclosure of all known risks);
HeinOnline -- 95 Yale L.J. 248 1985-1986
Protecting Patient Choice
concerned first and foremost with outcomes, historically they have been
reluctant to disclose risks and share decisionmaking. 2 0 If doctors evaluate
the adequacy of disclosure, even when it is designed to protect choice, they
will naturally respond in terms of these traditions. 2 '
Professional expertise, therefore, is not the appropriate determinant of
how much disclosure is desirable or adequate for purposes of patient
choice. Where the ultimate issue is defined as protection from physical
injury, however, absent a painstaking parsing of the sub-issues, professional expertise will seem both central and sufficient to measure duties of
care. A substantial minority of states, responding to the identification of
the subsidiary choice interest, have adopted a reasonable patient standard
to measure the content and adequacy of disclosure.' But most states, responding to physical well-being as the protected interest, have chosen
professionalized standards of care as both natural and justified.' 2 3 The
difference of rules reflects the internal conflict within the hybrid doctrine;
the dominance of the professional standard reflects the dominance of physical well-being as the ultimately protected interest. If choice were an independently protected interest, the role of medical expertise could more appropriately be delimited.
b.
Causation
Medical cases potentially impose enormous liability. Fearing that patients' testimony would be self-serving and biased by hindsight, 2 " courts
have felt it necessary to subject hypothetical reconstructions of individual
choice to standardized criteria.'2 5 In informed consent cases, plaintiffs
ing, supra note 30, at 1559-66 (urging adoption of "reasonable patient standard"). Perhaps the best
known objection to the professional standard was expressed by Judge Robinson in Canterbury v.
Spence: "Nor can we ignore the fact that to bind the disclosure obligation to medical usage is to
arrogate the decision on revelation to the physician alone." 464 F.2d 772, 784 (D.C. Cir.), cert.
denied, 409 U.S. 1064 (1972).
To some degree, the conflict over the standard for disclosure reflects the difference between contractual and tort norms. Causes of action in contract look to the reasonable expectations of the promisee;
professional negligence actions in tort look to the specialized competencies and practices of the professional group.
120. This fact is thoroughly documented by J. KArz, supra note 7, at 1-29.
121. For example, in Dunham v. Wright, 423 F.2d 940 (3d Cir. 1970), a case involving death as
a result of thyroid surgery, the court affirmed the trial court's refusal to grant a judgment n.o.v. on the
issue of informed consent after a jury verdict for the doctor. The plaintiff claimed that the doctor's
failure to disclose alternative methods of treatment rendered the patient's consent defective. The court,
however, quoted with apparent approval the doctor's assertion that the patient had had no alternatives. Id. at 946. It seemed completely unaware that the doctor may have simply characterized his
expert recommendation in a fashion that justified the foreclosure of the patient's interest in choice.
122. J. AREEN, P. KING, S. GOLDBERG & A. CAPRON, supra note 32, at 384 n.4.
123. Id. (twenty-six states use professional standard of care).
124. See, e.g., Canterbury v. Spence, 464 F.2d at 790-91 (subjective standard "places the physician in jeopardy of the patient's hindsight and bitterness").
125. Id. at 791. But see Scott v. Bradford, 606 P.2d 554, 559 (Okla. 1980) (refusing to jeopardize
HeinOnline -- 95 Yale L.J. 249 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
made, a reasonable permust show that, had the contested disclosure been
1 28
treatment.
the
to
consented
son would not have
Such a standard invites both juries and doctors to make the too easy and
superficial assumption that reasonable people do what their competent
doctors tell them to do. 2 ' Moreover, as commentators have argued, the
choices made by reasonable others are not an appropriate screening crite128
rion where the value at issue is personal autonomy. Yet where physical
well-being is the protected interest, choice is placed in the role of factual
cause, linking breaches of duty to the occurrence of harm. Analyzed in
this way, what began as a concern for individual autonomy almost necessarily comes to be subjected to standardizing and oversimplifying criteria
that are alien to individuality.
objective
"right to know" by imposition of "reasonable man" standard). In some unusual instances an
that
jurisdictions
even
instances
such
In
doctors.
disadvantage
to
work
will
causation
standard of
selfof
problem
the
course,
Of
one.
subjective
a
to
switch
may
ordinarily employ an objective standard
stanserving testimony is removed in these instances, but the substantive illegitimacy of the objective
Truman v.
dard in the ordinary case is also indirectly acknowledged by these decisions. See, e.g.,
"pruThomas, 27 Cal. 3d 285, 294, 611 P.2d 902, 907, 165 Cal. Rptr. 308, 313, (1980) (satisfying
Jabaay,
v.
Guebard
also
see
plaintiff);
by
recovery
for
sufficient
not
but
dent person test" necessary
117 Ill. App. 3d 1, 10, 452 N.E.2d 751, 757-58 (1983) (upholding jury verdict for defendant/physiperson
cian because, given plaintiff's active sports life, she would have been less likely than reasonable
to choose undisclosed alternative to doctor's suggested treatment).
supra
126. Almost all jurisdictions have adopted the objective standard. 3 MAKING D cisiONS,
what
note 2, at 197. In tort law generally, decisions about how to handle causal questions regarding
treatment
would have happened had greater information been provided have not always paralleled the
cases involving
of the issue in medical informed consent cases. Thus, for example, in products liability
have been
failure to warn, courts have not asked whether a warning, if one had been given, would
be asheeded. These cases consequently never reach the question of whether that judgment should
Co., 278
sessed in individual or in reasonable person terms. See, e.g., McCormack v. Hankscraft
found evidence
Minn. 322, 154 N.W.2d 488 (1967) (in reversing judgment n.o.v. for defendant, court
where manusufficient to uphold jury verdict in favor of child injured by boiling water in vaporizer
But
facturer failed to warn despite lack of evidence regarding effectiveness of hypothetical warning).
of statistically
see Cunningham v. Charles Pfizer & Co., 532 P.2d 1377 (Okla. 1975) (failure to warn
entitled
small risk of paralysis rendered consent defective; case remanded for new trial with plaintiff
objective
being
test
ultimate
with
heeded
been
have
would
warning
that
to rebuttable presumption
Liabilone). For a discussion of failure to warn in products liability cases, see Keeton, Products
may differ
it)--Inadequacy of Information, 48 TEx. L. REv. 398 (1970). Manufacturers and doctors
of patients'
in their ability to absorb the costs of accidents, but doctors can probably prevent invasions
personcomparatively
The
products.
from
injuries
prevent
can
manufacturers
than
choice more easily
of warning
alized setting of the doctor-patient relationship should make discussion and evaluation
in prodinformation more useful than written product warnings. The different treatment of the issue
currently
ucts liability cases at least suggests that it is not essential to employ the causation analysis
294-308.
used by the courts in informed consent. See also infra text accompanying notes 212-13,
127. See, e.g., supra note 106.
when
128. See Meisel, Expansion, supra note 6, at 112. Judge Burger made the point forcefully
discussing an opinion by Justice Brandeis. He observed that:
Nothing in this utterance [by Brandeis] suggests that Justice Brandeis thought an individual
or wellpossessed these rights only as to sensible beliefs, valid thoughts, reasonable emotions,
and
unreasonable
foolish,
many
great
a
include
to
intended
he
founded sensations. I suggest
at great
even absurd ideas which do not conform, such as refusing medical treatment even
risks.
(D.C. Cir.
Application of the President & Directors of Georgetown College, 331 F.2d 1010, 1017
1964) (Burger, J., dissenting) (emphasis in original).
HeinOnline -- 95 Yale L.J. 250 1985-1986
Protecting Patient Choice
In many tort actions, factual cause is relatively clear. Even in those
actions where actual causation is less certain, an all-or-nothing resolution
may yet be justified on grounds that, under a balance of probabilities test,
the injury either was or was not causally connected to the negligent act.1 29
Invasions of autonomy, however, involve an especially complex and probabilistic analysis. There are multiple issues: Did the doctor's nondisclosure
materially invade the patient's interest in choice? What would the patient
have chosen had her choice been protected? What would have happened,
medically, had the alternate choice been made? Such complexities are unmanageable within the yes/no framework of factual cause; compressed
into a single question, they become oversimplified.
Under such a simplified analysis, if individual criteria were used, the
very existence of any injury would seem to turn solely on the rather shaky
reed of the plaintiff's hindsight testimony. It is not surprising that courts
faced with such a compacted ultimate issue moved to adopt objective standards of reasonableness to address the question. 13 1 If, on the other hand,
choice were an independently protected interest, the factual cause issue
would be narrower and simpler-whether the patient's right to choose
had been encroached upon as a result of a doctor's failure to disclose.1 "
To be sure, difficult problems of uncertainty, prediction and credibility
would remain regarding what would have happened had the patient been
given the choice. However, with choice as the protected interest, these
problems would be assessed as questions of the valuation of an injury that
was acknowledged to have taken place. The framework of valuation is
better adapted to the resolution of such probabilistic issues than is the
traditional analysis of factual cause. Moreover, questions regarding what
redress should be available to remedy the invasion of choice would then be
analyzed, appropriately, as issues of sanctioning policy rather than of the
factual existence of harm.
c.
Categories of Compensable Harm
Where battery was preoccupied with physical touch, negligence vindicates physical well-being. 3 2 Many invasions of patient autonomy do re129. For example, even if the defendants were negligent, if their negligence was, more probably
than not, not a substantial factor in plaintiff's injury, no liability is warranted. See J. FLEMING, AN
INTRODUCTrON To THE LAW OF TORTS 109-11 (1967). And where the question of factual cause is
complicated by the presence of two tortious factors, if either could, more probably than not, have
produced the injurious result, liability is warranted. Id.
130. See King, supra note 27 (criticizing courts' frequent conflation of issues of causation and
valuation).
131. Some standard of materiality would be needed. See infra text accompanying note 285.
132. While physical injury is the central theme of negligence doctrine, other types of injury may
be cognizable. For example, damages for emotional distress may be recovered in conjunction with
infringement of other interests, see RESTATEMENT (SECOND) OF TORTS § 905(b) (1977), or in their
HeinOnline -- 95 Yale L.J. 251 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
sult in physical injury as it is traditionally defined. Although such injuries
would seem to fit standard negligence definitions of harm, they often go
unredressed because, as demonstrated above, the analysis of duty, breach,
or cause differs where the protected interest is physical well-being rather
than choice. In addition, preemption of patients' authority by doctors may
also give rise to injuries that are real but intangible, or to physical outcomes that are arguably not "injurious" except from the individual's vantage point. These outcomes may be excluded from negligence doctrine's
definitions of harm. Thus, a patient not told about a method of sterilization that is more reversible than the one performed may have difficulty
133
physical injury.
convincing a court that nonreversibility is a cognizable
A patient who alleges that, properly informed, she would have chosen a
lumpectomy rather than a radical mastectomy might find it hard, under
existing negligence rules, to characterize the successful operation that re"injured" her.134
moved her breast and eradicated her cancer as having
Similarly, the patient with a desire to go home or to a hospice to die, who
is instead maintained alive by hospital machinery, might have difficulty
establishing "injury" under definitions of an interest in physical well1 35 And, at least when such cases were first litibeing rather than choice.
own right, under rules governing the interest in freedom from emotional distress, see id. at § 46.
133. See, e.g., Masquat v. Maguire, 638 P.2d 1105 (Okla. 1981). The doctor admitted he did not
court
inform Ms. Masquat of alternative procedures or of their differing degrees of reversibility. The
action
held that the doctor's nondisclosure would not support either a battery or an informed consent
because of the absence of "causal linkage between some unrevealed risk and the injuries complained
of." Id. at 1107.
Intangibility of injury might also play a role in barring redress in cases where doctors' nondisclosure arguably deprived patients of a chance to obtain a preferable outcome. If more sophisticated
concepts of measurement were used to value a chance, these "ordinary" physical injuries might arguais lost
bly be analyzed as injuries attributable to a lapse of professionally competent care. If the chance
depbecause the doctor's conduct fell below professional standards, that would be appropriate. Many
of
uncertainty
The
choice.
to
injuries
as
characterized
better
are
however,
cases,
rivation of a chance
of
medicine dictates that doctors should not be accountable for perfect outcomes. Precise measurement
sublost chances might suggest such accountability. Competence standards can and should require a
to professtantial threshold of knowledge, judgment, and action. Yet beyond demanding conformity
the judgto
taking
worth
are
chances
which
of
weighing
the
leave
should
law
the
sional standards,
ment of the individual affected. See infra text accompanying notes 217-31.
134. Despite the controversial nature of the surgery, no case has been discovered dealing with
deprivation of choice regarding lumpectomy as opposed to mastectomy, which suggests that such a
cause of action would not presently be cognizable. Cf. Hanks v. Doctors Ranson, Swan & Burch,
Ltd., 359 So. 2d 1089, 1093 (La. Ct. App.) (for discussion of case, see infra text accompanying notes
147-49), cert. denied, 360 So. 2d 1178 (La. 1978).
the
135. The "right to die" examples, although aneedotally familiar, do not seem to have reached
courts-at least as disputes between individuals. But see Court Hears Dead Man's Arguments in
Right-to-Die Case, San Francisco Chron., Nov. 9, 1984, at 10, col.1 (family filed $10 million civil
damage suit against hospital for not respecting family member's wish to be disconnected from respirator). The paucity of such litigation in the literature suggests lacunae in the cause of action as presently
conceived. In addition, the societal interest in such decisions is still being litigated. See generally J.
AREEN, P. KING, S. GOLDBERG & A. CAPRON, supra note 32, at 1077-1147 (presenting interdisciplinary viewpoints on legal, social, and ethical issues in regulation of death decisions); PRESIDEN'S
BEHAVIORAL RESEARCH, DECOMM'N FOR THE STUDY OF ETHICAL PROBLEMS IN MEDICINE AND
decisions are and
CIDING TO FOREGO LIFE-SUSTAINING TREATMENT (1983) (re-examining way
HeinOnline -- 95 Yale L.J. 252 1985-1986
Protecting Patient Choice
gated, the patient who gives birth to an unwanted child because of a doctor's failure to provide choice-protecting information may have difficulty
13 6
showing harm.
Conclusions about the reality of injury and about whether conduct is
deserving of sanction ultimately depend both on how the underlying interest is defined and on how accurately the consequences of its invasion have
been traced. In both respects, existing analysis that vindicates patient autonomy only in an indirect fashion has produced a pattern of protection
for that interest that is flawed in critical ways.
C.
The Insufficiency of Negligence Protection
The preceding sections document how choice is subordinated and distorted at every stage of the negligence analysis. Under the prototype identified at the outset, this weak form of protection offered by informed consent doctrine governs only comparatively peripheral questions. Thus,
where the doctor proposes surgical intervention, the occurrence of that intervention will necessarily be within the patient's awareness, and, hence,
to some degree, control. Furthermore, medical custom has accepted a need
for at least some basic consent in such circumstances.1 37 Finally, most
states retain some legal requirement under battery doctrine that the patient's consent be sought.1 8 Under modern views, that requirement does
not demand that the patient's consent be highly informed. However, because courts often treat blanket consents as being of questionable validity,13 9 the basic consent requirement for battery in fact necessitates some
disclosure regarding the specific proposal. Under the prototype, what is
ought to be made about whether to forego life-sustaining treatment). Once the parameters of patient
autonomy in regard to death decisions are more clearly established, litigation over infringement of
those rights by private parties such as doctors and hospitals will likely begin in earnest, as was the
case when resolution of constitutional disputes over abortion gave rise to large numbers of private law
actions concerning wrongful birth, life, conception, etc.
136. Wrongful birth cases will be discussed in Part III as examples of the emergence of an interest in choice similar to the one suggested here.
137. 2 MAKING DECISIONS, supra note 2, at 81. Virtually all doctors report obtaining consent for
inpatient surgery.
138. 1 MAKING DECISIONS, supra note 2, at 22; see also supra text accompanying notes 28-29.
139. See, e.g., Rogers v. Lumbermens Mutual Casualty Co., 119 So. 2d. 649 (La. Ct. App. 1960)
(operation for removal of reproductive organs tortious when plaintiff only consented to removal of
appendix); Gray v. Grunnagle, 423 Pa. 144, 167, 223 A.2d 663, 674 (1966) (general consent signed
upon admission to hospital may be found inadequate). But see Kennedy v. Parrott, 243 N.C. 355,
362, 90 S.E.2d 754, 759 (1956) (absent proof to contrary, consent to major operation will be construed as general in nature and surgeon may extend operation to remedy any condition in area of
original incision). See generally 23 OHIO REv. CODE. ANN. § 2317.54 (Page 1981) (example of
statute requiring consent forms to be particular and specific); 1 MAKING DECISIONS, supra note 2, at
106 (Joint Commission on Accreditation of Hospitals requiring separate consent forms be signed for
any procedure or treatment "for which it is appropriate"); A. RosonF, supra note 29, at 283 (warning doctors that blanket consent forms are often legally defective).
HeinOnline -- 95 Yale L.J. 253 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
left to be litigated under a negligence/informed consent theory is mainly
14
nondisclosure concerning collateral risks.
Disclosure of collateral risks tends to be additive and elaborative, comparatively unimportant,14 1 particularly if only one course of action is
under discussion.14 2 As between doing nothing and accepting the recommendation, a patient's decision is driven by the pains and problems of the
disease or illness that brought her to the doctor in the first place. Subjecting the patient's interest in receiving such information to professional
standards of care and reasonable patient standards of causation seems relatively unobjectionable.
In nonprototypical situations, the contrast is stark. Where the doctor in
effect recommends inaction (or continuing action or non-touching action),
that recommendation can be implemented without the patient's awareness.
Such decisions cannot be physically "sensed" by the patient. As a matter
of medical custom, they are viewed as issues of professional competence
1 43
Existing touch-oriented legal doctrines inaprather than patient choice.
propriately reinforce that conclusion.
Yet in these instances, information does not merely elaborate collateral
risks but is itself the sine qua non of choice, the sole means by which the
patient can become aware of highly consequential courses of medical management. Information in these cases is far more important than is most
collateral risk information; without disclosure, self-executing judgments
will be made by the doctor. The problem is illustrated by Hanks v. Ranson.144 In Hanks, the plaintiff's healthy breast was amputated after a
competently administered but mistaken diagnostic test. The plaintiff complained that the doctor had not told her of a more diagnostically accurate
two-step procedure that would have separated surgery from the diagnostic
biopsy. No protection regarding such an alternate choice will be forthcoming under battery doctrine. Consent was given to the very procedure per140. Under negligence analysis the predominant issue is disclosure of risks of the proposed proce195.
dure. See 3 MAKING DECISIONS, supra note 2, at 195; Andrews, supra note 85, at
This
141. Current doctrines are preoccupied most with that information which is least important.
calls
may explain why the doctrine itself is commonly perceived to be rather insignificant. Thus Katz
note 6,
supra
Expansion,
Meisel,
also
see
15;
note
supra
Consent,
Informed
Katz,
Tale."
it a "Fairy
legal
at 90 (calling informed consent in 1960's a "paper tiger"). Doctors frequently perceive the
J.
doctrine to be an albatross. See, e.g., Katz, Informed Consent: Is It Bad Medicine? 126 WESTERN
and that
MED. 426 (1977) (anesthesiologist argues that "informed consent" is useless legal doctrine
patients complain of being told "too much"). Seventy-nine percent of the public feels that the primary
purpose of consent forms is to protect doctors from lawsuits. See 2 MAKING DECISIONS, supra note 2,
at 160.
be
142. Given the weakness of present rules regarding disclosure of alternatives, this will typically
84-87.
notes
accompanying
text
supra
See
situation.
the
fall-off in
143. As the intervention becomes less like the prototypical surgery, there is a rapid
is true
perceived need to seek consent. See 2 MAKING DECISIONS, supra note 2, at 168, 335. This
id. at 335.
even though medication decisions, for example, may involve greater risks than surgery. See
144. 359 So. 2d 1089 (La. Ct. App.), cert. denied, 360 So. 2d 1178 (La. 1978).
HeinOnline -- 95 Yale L.J. 254 1985-1986
Protecting Patient Choice
formed; no consent is presently required for the doctor's decision not to
use the two-step alternative because that decision involved no physical
contact. Moreover, despite acknowledged non-disclosure of the alternative, 145 the court upheld a verdict for the defendant even though jury instructions on informed consent made no reference to disclosure of alternatives. Recharacterizing the issue as one of risk, the court held that the risk
of a wrong diagnostic outcome was "remote," and thus need not be
14
disclosed.
Information about alternative courses of action is different than information about collateral risks. Though remote side effect risks of a single
option may be comparatively unimportant, even a slightly greater risk, for
example, of unnecessary amputation may be sufficient to cause the patient
to choose an alternative. Under current informed consent doctrine, however, protection of the opportunity to choose an alternative may either
wholly disappear, as in Hanks, or be eroded by the distortions that flow
from the negligence interest definition.147 Thus, even had the court in
Hanks required disclosure of alternatives, vindication of a breast cancer
patient's interest in making the choice between one- and two-step procedures should not depend upon whether doctors typically disclose that information, upon whether reasonable patients would make the same choice
as the patient, or upon standardized definitions of physical injury. Yet
with regard to nonprototypical choices, that will be the result if only current negligence analysis is employed.
Current doctrine perceives nearly all informational issues to be
problems of professional knowledge and duty; negligence has become the
ordinary category for analysis of nondisclosure.' 4 Thus, when informa145. Id. at 1091. Even if disclosure is required, some fact patterns involving alternatives may not
be characterized as "injury" where the protected interest is physical well-being. See Masquat v.
Maguire, 638 P.2d 1105, 1106-07 (Okla. 1981).
146. 359 So. 2d at 1093.
147. Cf. Masquat, 638 P.2d 1105 (for discussion of case, see supra note 133); Thornton v. Annest, 19 Wash. App. 174, 574 P.2d 1199 (1978) (for discussion of case, see supra note 86). In part,
these cases reflect appellate deference to jury verdicts on the basis of norms regarding the judge-jury
function. But they also make clear the inadequacy of current legal protection of patient autonomy
under existing doctrines.
148. One authority did not succumb to this perception. Professor Plant suggested that nondisclosure of information about the "nature and character" of proposed treatment should trigger a claim for
battery, while information about collateral risks could be adequately treated under negligence. Plant,
Analysis, supra note 27, at 648-50 (correlating different types of nondisclosure with different legal
actions). The suggestion, however, has not been widely followed. Professor Plant himself later observed that "[uiltimately almost all informed consent cases came to be treated as falling in the negligence area." Plant, Decline, supra note 30, at 92. An occasional court picked up the distinction, see,
e.g., Hales v. Pittman, 118 Ariz. 305, 309, 576 P.2d 493, 497 (1978); Gaston v. Hunter, 121 Ariz. 33,
57, 588 P.2d 326, 350 (Ct. App. 1978), but their efforts have had little effect. Indeed, it is not likely
coincidental that the jurisdiction whose courts incorporated this approach became the first to adopt a
statute abolishing actions for medical battery. See ARIz. REv. STAT. ANN. § 12-562 (Supp. 1979).
Moreover, Plant's approach still assumes that important choices will involve physical touching. The
HeinOnline -- 95 Yale L.J. 255 1985-1986
Vol. 95: 219, 1985
The Yale Law Journal
tional gaps in the protection accorded to choice have been perceived at all,
they have been "corrected" by expanding the scope of the duty to secure
informed consent. 149 The narrow disclosure-of-alternative-treatments ap50
proach or the broader Gates-type extension of informed consent would
bring more information concerning such decisions to the attention of the
patient than do other formulations of the duty to secure informed consent.
But the solution is less than adequate. The protection offered to patient
autonomy under the informed consent doctrine is weak, distorted by an
analysis rooted in the standard negligence interest definition. Although
such analysis arguably provides adequate vindication for choice interests
regarding collateral risk information, it is difficult to justify when it constitutes the sole protection accorded to patient choices regarding entire
courses of action (or inaction). Yet that will be the result where battery/
consent requirements are not triggered because no touching is proposed.
Only a few states have even extended the doctrine of informed consent
to impose any duty of disclosure in circumstances that involve no physical
contact.151 Others have actually moved in the opposite direction: They
have abolished all actions for medical-care battery, leaving only a negli15 Such a policy does
gence action to protect all aspects of patient choice.
nothing to correct existing weaknesses of negligence doctrine; at the same
time, it deprives even those instances of choice that do involve physical
contact of any vindication other than that available under negligence doctrines. The move to abolish battery seems to be an outgrowth of legitimate
frustration with certain aspects of the intentional tort analysis. Yet it may
also derive from an uncritical assumption that if negligence/informed consent analysis seems to be doing an adequate job with some aspects of patient choice protection, it may legitimately be applied to all aspects. Such a
conclusion is seriously mistaken. If that mistake is not corrected, and some
better approach to the problem put forth, the move to abolish medical
battery actions might expand, further undermining the limited protection
the law currently accords to patient autonomy.
types of information under discussion here are not easily characterized as being about the "nature and
character" of a proposed intervention.
149. This is the result, for example, of an expansion to cover prescription of drugs, to require
disclosure of alternative treatments, or more broadly, to require disclosure of important information,
as the Washington court did in Gates.
150. See Gates v. Jensen, 92 Wash. 2d 246, 595 P.2d 919 (1979) (en banc). For further discussion of this case, see supra text accompanying notes 88-101.
151.
See supra Part II(B)(1)(b).
152. See, e.g., ARIz. REv. STAT. ANN. § 12-562 (Supp. 1979).
HeinOnline -- 95 Yale L.J. 256 1985-1986
Protecting Patient Choice
III.
A.
TOWARD GREATER PROTECTION OF PATIENT CHOICE
Factors Strengthening Patient Choice
In the case law, two factors seem to strengthen patient claims to information and choice. When there is a relatively crystallized conflict of interest, or when there is a recognizably heightened electiveness, the protection
accorded to patient autonomy is likely to be stronger than it would otherwise be. Although neither factor has been clearly articulated or consistently applied, each can be discerned as an emerging influence.
1. Conflict of Interest: Disqualifying the Doctor
The principal conflict of interest within the doctor-patient relationship
derives from the fact that doctors' incomes rise when patients consume
health care services that those same doctors recommend and provide."8 3
While that issue is frequently implicit in conflicts about care, it is rarely
raised in litigation, presumably because most patients have only an indirect economic stake in the health care they receive. 1 " In a few areas, however, litigation raising clear problems of conflict of interest does arise.
a. Expansion of Battery Doctrine
Only one court has accepted a battery theory as applicable to the prescription of drugs.1 5 5 Although the holding in Mink v. University of Chi153. Although the incentive to over-service is somewhat reduced because many doctors are paid
according to time expended rather than procedures performed, this is not true of some specialists. For
example, surgeons are paid mainly when patients accept their recommendations. This relatively direct
conflict of interest may partly explain why patient choice is most aggressively protected under battery's consent requirements in surgical cases. Moreover, the potential pharmaceutical, hospital, or
laboratory profits that are available to doctors greatly undermine the effectiveness of the payment-fortime concept in offsetting the conflict of interest. See Relman, Dealing With Conflicts of Interest, 313
NEw ENG. J. MED. 749 (1985) (expressing concern that doctors' entrepreneurial profit-making
desires weaken both professional ethics and public trust). The concept also does not remove doctors'
economic incentives to recommend time-use itself.
154. Despite their indirect stake in, for example, insurance premiums, patients are not the most
obvious financial losers when excess care is provided. Few patients file suit alleging only that they
were given more care than they needed, i.e. that they paid for unnecessary care. See Salis v. United
States, 522 F. Supp. 989, 994-96 (M.D. Pa. 1981). The main objectors to such care would be thirdparty payors. These organizations tend to pursue political and market methods of cost control. See
Blumstein & Sloan, Redefining Government's Role in Health Care: Is a Dose of Competition What
the Doctor Should Order?, 34 VAND. L. Rxv. 849, 856-59, 863-64 (1981); Havighurst, Competition
in Health Services: Overview, Issues and Answers, 34 VAND. L. Rxv. 1117, 1123 (1981). Moreover,
since the main conceptual structure of physician accountability is professional competence rather than
patient (or indirectly, payor) choice, over-care will rarely be negligent, though it may constitute a
violation of patient choice.
155. Several courts have applied negligence theories of informed consent to drug prescription. See,
e.g., Hamilton v. Hardy, 37 Colo. App. 375, 549 P.2d 1099, 1104-05 (1976); Trogun v. Fruchtman,
58 Wis. 2d 596, 592-604, 207 N.W.2d 297, 307-15 (1973). At least one case holds drug prescription
to be strictly a therapeutic decision, requiring no disclosure to or consent by the patient. See Malloy v.
Shanahan, 280 Pa. Super. 440, 421 A.2d 803 (1980).
HeinOnline -- 95 Yale L.J. 257 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
cago156 was unusual, the federal district court apparently decided that
only battery, from among existing doctrinal tools, provided sufficient protection to these patients' autonomy in the face of their doctors' conflict of
interest.
During their prenatal care, the Mink plaintiffs were administered DES
pursuant to a research experiment evaluating the effectiveness of the drug.
Suing under both battery and negligence theories, plaintiffs claimed injuries from the increased risk of cancer to their daughters and from personal
emotional distress arising from this threat to their children.
As noted in Part II, the application of battery analysis to these facts
would ordinarily have been questionable on several grounds.1 57 Although
pills may have serious physical consequences, prescribing them does not
involve the kind of touching traditionally associated with battery. Moreover, the plaintiffs had consented to prenatal care. Under prevailing standards, 58 miscarriage prevention would seem to fall within the ordinary
scope of prenatal care; the fact that patients voluntarily ingested the pills
would also typically have constituted consent.1 59 Yet when the plaintiffs'
claim was characterized as one alleging professional negligence, it failed.
Even if nondisclosure were found to violate medical community standards,
a questionable matter at best, the plaintiffs alleged no injury cognizable
under negligence doctrine.160
Thus under existing medical consent law, the plaintiffs in Mink were
trapped between the absence of unconsented touch on the one hand and
the lack of injury to physical well-being on the other. Yet the court sensed
an injury to patient choice that it was unwilling to ignore; it accepted a
156. 460 F. Supp. 713 (N.D. Ill. 1978).
157. See supra note 45.
158. The court itself stated the Illinois standard for battery to be a "total lack of consent by the
patient." 460 F. Supp. at 717 (footnote omitted).
159. Exactly what the patients were told is a matter of some uncertainty. According to Mark
Debofsky, of the office of the plaintiffs' attorney, various plaintiffs testified that doctors told them, "I
want you to take these pills to help you through your pregnancy," "I want you to take these vitamins." Mr. Debofsky did not feel that the court's acceptance of the battery claim was premised on
affirmative misrepresentation. Telephone interview with Mark Debofsky (Oct. 1984).
160. 460 F. Supp. at 720. Cf. Rogers v. Okin, 478 F. Supp. 1342, 1388 (D. Mass.) (no injury
under negligence/informed consent because patients have not yet developed tardive dyskinesia, the
primary risk/side-effect of medications they were forced to take), affid in part, rev'd in part on other
grounds, 634 F.2d 650 (1st Cir. 1979). The inability to recover under negligence concepts of harm is
a common thread in Mink and Rogers. The Rogers court's award of injunctive relief against forced
medication based on constitutional grounds is analogous to the acceptance of a battery characterization
in Mink: Each provided vindication for the patients' interests in choice independent of the definitions
and standards of professional negligence. Given the Rogers court's constitutional holding, in which it
explicitly stated that the competence of the professional recommendation was not sufficient to overcome or satisfy the patients' privacy right to refuse forced medication, it is theoretically though not
pragmatically surprising that the court rejected a battery action and held that, although an informed
consent action might be appropriate, it would be judged by standards of professional practice. Id. at
1387.
HeinOnline -- 95 Yale L.J. 258 1985-1986
Protecting Patient Choice
battery characterization of the nondisclosure in Mink rather than allow
that injury to go unredressed. The court set aside the absence of physical
contact, stating that "[t]he gravamen of a battery action is the plaintiff's
lack of consent, not the form of touching."1 1 It also determined that the
issue regarding consent was sufficient to go to the jury under a battery
theory.
In explaining its decision, the court referred to performance of "substantially different acts."16' 2 The description seems inapposite, however,
for it is not the acts that were different. Rather, the essential complaint in
Mink was that the patients were not given crucial information before consenting. Although nondisclosure about medication is typically analyzed, if
at all, as an issue of professional negligence,"' the particular nature of
this undisclosed information caused the Mink court to deviate from that
pattern. What the doctors did not disclose was their research purposes.'
Although the court does not use the term "conflict of interest," that seems
the essential framework from which its decision derives.
A doctor's specialized knowledge and powerful role make her a fiduciary to those who depend on her. Consequently, the doctor owes undivided
loyalty to her patients.1 65 This duty of loyalty constrains and legitimates
whatever power the doctor has to advise or act on behalf of the patient. In
Mink, however, the doctors' judgments were potentially influenced by research motivations. This violation of the doctor's fiduciary obligation of
loyalty increased the court's concern that the dependent patient have the
opportunity to control decisions.166 Typically, where conflict of interest is
161. 460 F. Supp. at 717 n.4. The court observed that if the medicine had been administered by
injection, touching would have been present. Id. at 718.
162. Id. (emphasis added).
163. See supra note 155. Certainly omitting to tell patients the content of a drug is common, see 2
MAKING DECISIONS, supra note 2, at 334, and such omission is not perceived as subjecting a doctor
to liability. Failure to disclose side effect risks is treated as an issue of professional negligence.
164. Thus it is not the nature of the doctors' medical intervention that is at issue but their motivation for intervening.
165. RESTATEMENT (SECOND) OF AGENCY § 387 (1957).
166. See RSTrATEMENT (SECOND) OF AGENCY § 390 comment a (1957). But see Burton v.
Brooklyn Doctors Hosp., 88 A.D.2d 217, 452 N.Y.S.2d 875 (1982), which held that researching
doctors were negligent in subjecting a premature baby to an experiment involving usage of high quantities of oxygen that eventually blinded the baby. Incompetent care, together with occurrence of injury
cognizable under negligence analysis, rescued Burton from the problem faced by the court in Mink;
there was no discussion of battery here. The court's analysis of disclosure, however, was seriously
flawed. The court reversed a jury verdict of negligence against the treating doctor for not disclosing
the experimentation by his colleagues, about which he knew, to the baby's parents. In so doing, it
stated that the doctor did nothing wrong. It concluded that because "no evidence was offered of any
continuing obligation on his part to obtain informed consent once his order was countermanded by a
superior, the verdict against him based on failure to obtain informed consent cannot stand." Id. at
227, 452 N.Y.S.2d at 881-82. Although it is correct to say there was no incompetent care by this
doctor, dismissal of the disclosure issue seems wrong. The court's "no evidence" characterization is
puzzling, because it stated that the hospital imposed on the treating doctor the duty of informing
parents, and that he testified that he could not remember informing them. Id. Although here the loss
HeinOnline -- 95 Yale L.J. 259 1985-1986
Vol. 95: 219, 1985
The Yale Law Journal
involved, even a showing of competence is insufficient to immunize a fiduciary from liability.16 7 A fiduciary must not only justify the substantive
adequacy (the competence) of the transaction, but must also disclose it and
seek the agreement of the client. 6" The Mink court's acceptance of a battery characterization is what prevented either the presence of professional
competence or the absence of an injury to physical well-being from immunizing the conduct of the doctors.
The Mink decision demonstrates circumstances that call for stronger
vindication of patient choice. It suggests the needed transition from physical parameters of consent toward a more intangible notion of medical
choice. Furthermore, it transcends the dichotomy that places basic consent
under battery doctrine while assigning issues of disclosure to negligence
theory. Finally, it reveals that guaranteeing professionally competent care
of physical well-being does not sufficiently safeguard patient autonomy.
b.
Application of General FiduciaryPrinciples
When medical intervention leads to harm, the doctor who may be guilty
of malpractice and does not wish to disclose pertinent facts has interests
that directly conflict with those of the patient." 9 Such situations do not
lend themselves to a Mink-type battery analysis. Some courts, sensing the
underlying conflict, have achieved a similar result by analyzing these issues of disclosure under general principles of fiduciary duty. The duty to
disclose for purposes of informed consent is a specific instance of such a
fiduciary duty,'"0 yet, for several reasons, a generic fiduciary duty to discould be laid on other more culpable defendants, the court's analysis of nondisclosure leaves much to
be desired. Again, the inadequacy stems in part from a refusal to examine disclosure in nonprototypical situations.
167. See RESTATEMENT (SECOND) OF AGENCY § 381 comment d (1957). A conflict of interest
might, of course, cause the doctors to provide a level of care that falls below the professional standard
of competence. If so, they would be liable for the results of that incompetence. There was no contention in Mink that the doctors had provided incompetent care at the time the drugs were administered.
At the time, DES (a man-made estrogen) was widely prescribed as a miscarriage preventive. Note,
DES and a Proposed Theory of EnterpriseLiability, 46 FORDHAM L. REv. 963, 964 & n.4 (1978).
The FDA did not withdraw DES from the market until 1971. Sindell v. Abbott Laboratories, 26 Cal.
3d 588, 594, 607 P.2d 924, 925, 163 Cal. Rptr. 132, 133, cert. denied, 449 U.S. 912 (1980).
168. Where an agent, even with the knowledge of the principal, acts on the agent's own account
in a transaction, that agent "has a duty to deal fairly . . . and to disclose . . . all facts which the
agent knows or should know would reasonably affect the principal's judgment, unless the principal
has manifested that . . . he does not care to know them." RESTATEMENT (SECOND) OF AGENCY §
390 (1957). The limitation implied here by the qualifier "reasonably" is not nearly so restrictive as
the reasonable person causation standard employed in informed consent cases: The actual individual
client remains the decisionmaker, and the "unless" clause clearly tilts the requirement in favor of
disclosure. Another analogy may be drawn from lawyer-client relations. See MODEL RuLrs OF PaOFESSIONAL CONDUCr Rule 1.7(b) (1982) (requires both adequate representation and disclosure/
consent).
169. See Delgado & Vogel, supra note 69, at 52 (proposing duty on all members of medical team
to disclose any malpractice they have witnessed).
170. Canterbury v. Spence, 464 F.2d 772, 782 (D.C. Cir.), cert. denied, 409 U.S. 1064 (1972),
HeinOnline -- 95 Yale L.J. 260 1985-1986
Protecting Patient Choice
close sometimes more effectively vindicates patient interests in autonomy
than do the narrower duties that have crystallized under ordinary rules of
medical consent.
Fiduciary responsibilities are imposed in order to regulate relationships
marked by dependency and disparity of power.17 1 Although technical expertise is the main reason a power inequity exists, such expertise is not
the sole measure of the fiduciary's responsibility. Where the possibility of
conflict of interest exists, the fiduciary's accountability for disclosure and
accountability for competence are separate and cumulative, not alternative.
Thus, analysis under general fiduciary principles is less likely than analysis under informed consent doctrines to narrow the obligation of disclosure
or to confuse the issues of patient choice with those of professional
competence.
17 2
For example, where there is failure to disclose in the aftermath of a
medical intervention, no proposal to touch or treat is made, and thus battery doctrine will not protect patient autonomy. The absence of proposed
physical contact could also mean that no disclosure would be required
under rules of informed consent. 173 Yet conflict of interest may trigger
analysis under general fiduciary principles, requiring disclosure of all information that might be material to the patient.17 4 Unlike the duties specified under battery or informed consent doctrines, such a duty has no limitation based on time or stage of treatment. It is triggered not by physical
contact, but by the more flexible and general criterion of possession of
relevant information.17 5 Thus, under fiduciary principles, where malpractice has likely occurred, nondisclosure of the relevant facts may be remediable in an independent cause of action,1 76 or may provide plaintiff a reason to toll the statute of limitations in a malpractice action that would
1 77
otherwise be barred.
draws on this earlier and broader concept of fiduciary duty to impose the specific informed consent
duty.
171. See Frankel, Fiduciaty Law, 71 CALIF. L. REv. 795 (1983) (emphasizing control of discretionary and unequal power as common theme in fiduciary law).
172. But see Delgado & Vogel, supra note 69, at 67 (noting that, although doctors are frequently
characterized as fiduciaries, duties of disclosure imposed on them have been less extensive than those
imposed on other fiduciaries). Outside of conflict of interest, fiduciary law, too, exhibits some ambivalence about the role of professional standards. See infra note 187.
173. See, e.g., cases cited supra Part II(B)(1); see also Delgado & Vogel, supra note 69, at 69-71
(noting limited scope of informed consent duties).
174. See Nixdorf v. Hicken, 612 P.2d 348, 354 (Utah 1980) (duty to disclose applies to material
information).
175. Such a standard is like the one developed in Gates v. Jensen, 92 Wash. 2d 246, 595 P.2d 919
(1979) (for discussion of case, see supra Part II(B)(2).
176. See, e.g., Nixdorf, 612 P.2d at 352 n.7.
177. Some courts impose an affirmative duty under which mere silence is tantamount to concealment. See, e.g., Stafford v. Shultz, 42 Cal. 2d 767, 270 P.2d 1 (1954). Other courts require active
misrepresentation or silence regarding a known certainty before tolling the statute. See, e.g., Nardone
v. Reynolds, 333 So. 2d 25, 34-40 (Fla. 1976).
HeinOnline -- 95 Yale L.J. 261 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
Moreover, once a duty of disclosure is imposed under general fiduciary
principles, there may be less tendency at other stages of the analysis to
confuse the patient's interest in the information with issues of competent
physical care. 178 Thus, expert testimony may be less likely to be required
to establish the standard of care, 179 and defenses might be more narrowly
drawn."" As in battery analysis, factual cause will tend to be presumed;
no inquiry will likely be made regarding what a reasonable person would
have done had the disclosure been made." Remedy for a breach of fiduciary duty in circumstances involving conflict of interest may also be
broader than if the action were one in ordinary negligence.' 82
Despite their tendency to broaden liability and remedy, however, several factors limit the capacity of general fiduciary duties to resolve
problems identified here. First, a relatively crystallized conflict of interest
may be necessary before courts decide that such principles should apply. If
even the conflict of interest inherent in situations of alleged malpractice
178. But see Nardone, 333 So. 2d 25. The plaintiffs urged that doctors' nondisclosure of a possible cause of their son's deteriorated condition after brain surgery should toll the statute of limitations
in their action for malpractice. The plaintiffs knew of the boy's condition but not why he had suddenly become worse. They were informed only that "these things happen." Id. at 29. The court held
that the doctors had no affirmative duty to disclose their speculations as distinct from their certainties
regarding causes. Id. at 39. Given the prevalence of uncertainty in medical knowledge, this distinction
places severe constraints on the affirmative duty to disclose. See also infra text accompanying notes
215-31. Nardone illustrates that even the taint of fiduciary conflict of interest will not always be
sufficient to cause a court to strengthen the protection it gives to patient autonomy.
179. Where disclosure of possible malpractice is at issue, the physical injury arises because of
allegedly incompetent care, but the legal problem regarding disclosure of the harm has nothing to do
with professional expertise. This may be more easily recognized in these cases because, intuitively,
outcome disclosure is less likely than pretreatment disclosure to be characterized as a question of
professional competence. But see Lopez v. Swyer, 115 N.J. Super. 237, 251, 279 A.2d 116, 124
(1971) (employing medical community standard of disclosure despite issue of possible malpractice).
The issues of standard of care here may simply parallel the debate over the standard of care in
informed consent cases. States that use a reasonable person standard to decide issues of informed
consent are likely to use the same criterion under general fiduciary analysis. Yet the influence probably flows the other way. Awareness of the breadth of fiduciary responsibilities may make a court more
likely to choose the lay standard of disclosure even where the context is proposed treatment.
180. For instance, the defense of therapeutic privilege may not apply. Utah's informed consent
statute, for example, has such a privilege, UTAH COnE ANN. § 78-14-5(2) (1977), which would
presumably not apply where a court invoked the broad duty derived from fiduciary principles in
Nixdorf, 612 P.2d 348.
181. See, e.g., Nixdorf 612 P.2d 348; Stafford, 42 Cal. 2d 767, 270 P.2d 1; Lopez, 115 N.J.
Super. 237, 279 A.2d 116.
182. E.g., had the court imposed a duty to disclose facts to the plaintiff in Kelton, 413 A.2d 919,
she might have recovered damages flowing from deprivation of her interest in making decisions not
only about possible medical intervention, but also about litigation, possible adoption, etc. See supra
text accompanying notes 66-71.
Delgado and Vogel discuss injuries that may result from nondisclosure of malpractice that are
different from those arising from the malpractice itself. Economic harm in particular may be more
easily recognized where the conceptual framework is one of general fiduciary responsibility rather
than disclosure in the context of treatment per se. Delgado & Vogel, supra note 69, at 89; see also
Hart v. Browne, 103 Cal. App. 3d 947, 163 Cal. Rptr. 356 (1980) (lost opportunity to file malpractice action because of negligence of doctor advising about whether another physician was guilty of
malpractice).
262
HeinOnline -- 95 Yale L.J. 262 1985-1986
Protecting Patient Choice
does not always trigger a duty to disclose, 18 general fiduciary principles
would probably be ineffective to generate more aftermath disclosure where
malpractice is unlikely." 4 Yet even where no malpractice judgment would
be likely, doctors possess information that could be vital to their patients.
Second, decisions involving broad fiduciary principles of disclosure have
most frequently arisen in the context of requests to toll the statute of limitations. In such cases, the duty to disclose is imposed for the limited purpose of removing a technical defense to malpractice. The underlying protected interest is still in some sense the same as that in professionally
competent physical care." 5 It is less clear whether such a fiduciary duty
of disclosure would be recognized as an independent action-i.e., as protecting an interest in patient knowledge and choice rather than as a reason
to remove barriers to redress of injuries resulting from a violation of professional care." 6
Thus, general fiduciary principles can ameliorate limitations on protection of patient choice under existing medical consent doctrines, but such
principles are, at present, likely to be invoked only in limited circumstances for limited purposes. A more general fiduciary principle might yet
emerge with sufficient coherence and force to solve the problems identified
in this Article.18 7 In the meantime, principles governing fiduciary conflict
of interest have important implications for the protection of patient choice.
183. See Kelton v. District of Columbia, 413 A.2d 919 (D.C. 1980) (no duty to disclose scarring
of fallopian tubes that occurred during caesarian birth); Nardone, 333 So. 2d at 39 (no duty to
disclose conjecture about cause of relapse).
184. See, e.g., Roark v. Allen, 633 S.W.2d 804 (Tex. 1982) (no duty to disclose factual condition
about which doctors made competent but mistaken judgment). But see Tresemer v. Barke, 86 Cal.
App. 3d 656, 150 Cal. Rptr. 384 (1978) (duty to warn former patient when later learned of dangers
of IUD).
185. Damages in the delayed malpractice action may not be identical to what they would have
been had the problem been disclosed at the outset. See Delgado and Vogel, supra note 69, at 89.
186. One court has recognized an independent action. Nixdorf, 612 P.2d 348. Another, without
clearly addressing the relevant issues, declined to recognize any independent duty of aftermath disclosure. Kelton, 413 A.2d at 922.
187. LeBlang & King, Tort Liabilityfor Nondisclosure: The Physician's Legal Obligations to
Disclose Patient Illness and Injury, 89 DIcK. L. REv. 1, 24-26 (1984), suggest that a broad fiduciary duty of disclosure is already emerging. However, many of the cases they annotate could be
grounded on other analyses, such as the duty to provide competent care. Thus, the extent to which
such cases really rest on and protect patients' autonomy interests, as the authors claim, is
questionable.
Moreover, although a broad fiduciary duty surmounts some problems identified here, it is not immune to others. An agent's duty to "act on behalf of the principal," see RESTATEMENT (SECOND) OF
AGENCY § 13 comment a (1957), may conflict with his accountability to the control of the principal.
Id. at §§ 13, 14. In most agency relationships, this tension is resolved by contractual allocation of
authority. See id. at § 376. But in professional relationships, contract is a weak and rarely used tool.
In its absence, achieving adequate control by the principal (in professional settings, the client) is
further complicated by incorporation of professional standards of conduct into agency rules. Id. at §
379 comment c (discussing obligations of professional agent in traditional agency terms). To the extent
that professional standards provide a basis for measuring competent performance, they constitute no
problem. To the extent that they become a substitute for control by the client, they present the same
difficulties that are inherent in the rules now governing medical consent.
HeinOnline -- 95 Yale L.J. 263 1985-1986
Vol. 95: 219, 1985
The Yale Law Journal
2.
Heightened Electiveness: Empowering the Patient
Just as conflict of interest sometimes prompts courts to require special
choice-protecting disclosure, so a factor that might be called "heightened
electiveness" seems to produce more aggressive and independent protection
of patient choice. I use the term "heightened electiveness" to denote cases
where the special role of personal values or preferences causes a court to
have greater than ordinary concern about patient choice. Two types of fact
patterns seem presently to be perceived as possessing this characteristic:
(1) those involving elective, especially cosmetic surgery; and (2) those involving procreation, including sterilization, pregnancy, and birth. Both
cosmetic surgery and birth/procreational procedures are, to a significant
degree, optional. No progressive and threatening disease drives the patient
to undergo medical treatment. The patient seeks some affirmative outcome
instead of warding off an encroaching evil. Furthermore, both types of
decisions are recognized as highly personal, involving either the uniqueness of personal appearance or the deep personal values and consequences
inherent in procreation and parenting.18 8 Where heightened electiveness is
present, courts sometimes analyze a case in terms of doctrines other than
professional negligence, or alternatively, they may use the framework of
professional negligence but adapt it in some fashion that increases the protection afforded to choice.
a. Analysis Under ContractPrinciples
A court may strengthen protection of patient choice by classifying the
problem as one of contractual obligation, but few medical cases are analyzed under contractual theories. Those that are so analyzed emphasize
the definiteness of the doctor's promise. 8 9 Although that factor may play
a role, it is probably not generally dispositive. The certainty of the doctor's promises in such cases is not markedly greater than that in disputes
where contractual analysis is denied. 9 '
188. See, e.g., Roe v. Wade, 410 U.S. 113 (1973) (right to abortion based on constitutional right
to privacy); Griswold v. Connecticut, 381 U.S. 479 (1965) (access to birth control information protected by constitutional right of privacy). Procreation cases have to a considerable degree been analyzed as situation-specific, and they are indeed unique. However, these cases also illustrate unresolved
issues that are generally present in current doctrines protecting choice.
189. See Depenbrok v. Kaiser Found. Health Plan, 79 Cal. App. 3d 167, 144 Cal. Rptr. 724
(1978) (patient may recover for breach of contract if doctor dearly promised a particular result);
Guilmet v. Campbell, 385 Mich. 57, 188 N.W.2d 601 (1971) (allegation of specific, clear, and express promise sufficient to go to jury on contract theory). The critical problem is to distinguish between therapeutic reassurance and binding commitment in a setting of inherent uncertainty. As one
annotator has aptly commented, "Despite the statement of some courts that [therapeutic reassurance
and contractual promise] are quite distinct, there appears to be a 'gray area' . . . ." Annot., 43
A.L.R.3d 1221, 1226 (1972).
190. Thus promises such as "to cure him of bladder trouble," Marty v. Somers, 35 Cal. App.
HeinOnline -- 95 Yale L.J. 264 1985-1986
Protecting Patient Choice
The more important factor leading courts to apply contractual doctrines
is the electiveness of the medical intervention. A striking number of the
cases employing contractual analysis involve cosmetic surgery or procreative choice.191 The choice in these cases is often whether to have medical
treatment at all, rather than simply what kind of treatment to choose for a
condition that requires some treatment. Although courts ordinarily interpret doctors' positive descriptions of potential outcomes as harmless therapeutic reassurance, they are more willing in cases of elective treatment to
believe that the doctor induced the treatment,19 2 and accordingly more
willing to hold the doctor accountable for her promised results rather than
solely for the competence of her efforts.
Under tort law, the patient's interest in choice is analyzed under the
rules of battery and informed consent; vindication of choice is often diluted. By contrast, contract vindicates patient choice directly. Injury to expectation is judged from the perspective of the actual patient and encompasses, but is not confined to, physical injury; physical contact is
irrelevant; professional competence is not dispositive of duty.
b. Analysis Under Battery Principles
Battery doctrine, too, provides greater protection for choice than does
negligence analysis. Battery applies where there has been no consent at
all. 193 Decisions about whether a battery theory may be invoked are most
difficult where the finding of no consent rests on whether a treatment is
deemed to exceed a consent given to some different or lesser procedure.
Although fact patterns vary considerably, cosmetic surgery or procreation
182, 182, 169 P. 411, 412 (1917), and that a doctor "could and would cure" an osteopathic problem,
Kershaw v. Tilbury, 214 Cal. 679, 689, 8 P.2d 109, 113 (1932), have been held not to state a cause of
action in contract. A promise to make the patient's hand "100% good," Wilson v. Blair, 65 Mont.
155, 177-78, 211 P. 289, 297 (1922) (Farr, J., dissenting), overruled on other grounds, Klemens &
Son v. Reber Plumbing & Heating Co., 139 Mont. 115, 360 P.2d 1005 (1961), was unenforceable
because it was not given for separate consideration, an approach that is merely a different way to
defeat a contract characterization. See also Herrera v. Roessing, 533 P.2d 60 (Colo. Ct. App. 1975)
(doctor's statement that patient would not get pregnant after tubal ligation mere opinion, not contractual guarantee). According to Lane and Hirsh, "Courts will look for 'buzz' words such as 'I guarantee' or 'I promise you,"' before upholding a cause of action in contract. Lane & Hirsh, The Broken
Promise: Physician's Breach of Warranty, 89 CASE & COM., Sept.-Oct. 1984, at 3, 8.
191. See, e.g., Depenbrok v. Kaiser Found. Health Plan, 79 Cal. App. 3d 167, 144 Cal. Rptr.
724 (1978) (promise that tubal ligation would prevent pregnancy); Sullivan v. O'Connor, 363 Mass.
579, 296 N.E.2d 183 (1973) (plastic surgery on nose); Stewart v. Rudner, 349 Mich. 459, 84 N.W.2d
816 (1957) (promise to perform Caesarian section); Hawkins v. McGee, 84 N.H. 114, 146 A. 641
(1929) (plastic surgery on hand). But see Herrera v. Roessing, 533 P.2d 60 (Colo. Ct. App. 1975)
(statements about tubal ligation did not create contract).
192. See, e.g., Hawkins, 84 N.H. 114, 146 A. 641 (relying on repeated solicitations of doctor to
uphold contract claim).
193. See Cobbs v. Grant, 8 Cal. 3d 229, 240, 502 P.2d 1, 8, 104 Cal. Rptr. 505, 512 (1972).
HeinOnline -- 95 Yale L.J. 265 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
cases also stand out among those that allow the battery characterization in
94
such circumstances.2
c. Negligence Analysis Under Altered Interest Definition
Heightened electiveness sometimes causes courts to adapt standard negligence doctrines in ways that strengthen protection of patient choice. Such
adaptations are particularly likely to occur in cases involving procreation.
Actions for wrongful birth have encountered multiple analytic difficulties under standard negligence doctrines. When these cases first emerged,
courts were extremely reluctant to define the birth of a child,1 95 particularly a healthy child, 9 ' as an injury. When constitutional protection of
procreation deflected public policy objections to birth-as-injury, 7 analytic
problems came to the fore. The birth of a child, especially a normal child,
was difficult to encompass within a definition of protected interest as
avoidance of physical injury. 9 8 If the unwanted child was physically defective, the absence of physical harm might seemingly be finessed, but new
194. In his exhaustive analysis of informed consent cases, Professor Rosoff discusses the circumstances under which a limited consent may be used to validate a more extensive operation. See A.
ROSOFF, supra note 29, at 8-13. As cited by Rosoff, the cases refusing to extend consent include a
sterilization, Wells v. Van Nort, 100 Ohio St. 101, 125 N.E. 910 (1919), a cosmetic addition to
surgery, Lloyd v. Kull, 329 F.2d 168 (7th Cir. 1964), and a hip prosthesis, Cathemer v. Hunter, 27
Ariz. App. 780, 558 P.2d 975 (1976). By contrast, the cases accepting an extension of consent as valid
involved the reduction of a fracture, McGuire v. Rix, 118 Neb. 434, 225 N.W. 120 (1929), and an
appendectomy and rupture of cysts, Kennedy v. Parrott, 243 N.C. 355, 90 S.E.2d 54 (1956). Rosoff
specifically mentions that where medical intervention affects reproductive capacity, ordinary rules
about extension of consent may not apply. A. RosoFF, supra note 29, at 8-12. For other decisions
upholding battery claims for surgery, see Meretsky v. Ellenby, 370 So. 2d 1222, 1224 (Fla. Dist. Ct.
App. 1979) (tip of nose alteration presents battery claim despite statute deeming written consent sufficient to relieve doctor of liability); Kinikin v. Heupel, 305 N.W.2d 589 (Minn. 1981) (breast reduction beyond patient consent constitutes battery); Bang v. Charles T. Miller Hosp., 251 Minn. 427, 88
N.W.2d 186 (1958) (consent to prostate resection does not bar claim of battery where possible severance of sperm cords not disclosed).
195. Gleitman v. Cosgrove, 49 N.J. 22, 31, 227 A.2d 689, 693 (1967).
196. See, e.g., Shaheen v. Knight, 6 Lyc. Rptr. 19, 11 Pa. D. & C.2d 41 (Lycoming County Ct.
1957); Comment, Wrongful Pregnancy: Recovery for Raising a Healthy Child, 10 N. Ky. L. Rv.
341 (1983).
197. See Boone v. Mullendore, 416 So. 2d 718 (Ala. 1982); Cockrum v. Baumgartner, 95 Ill. 2d
193, 447 N.E.2d 385, cert. denied, 464 U.S. 846 (1983); Christensen v. Thornby, 192 Minn. 123,
255 N.W. 620 (1934); Gleitman v. Cosgrove, 49 N.J. 22, 227 A.2d 689 (1967); Collins, supra note
52, at 691-93.
198. See, e.g., Howard v. Lecher, 42 N.Y.2d 109, 112, 366 N.E.2d 64, 66, 397 N.Y.S.2d 363,
365 (1977) (parents "were made to bear no physical or mental injury, other than the anguish of
observing their child suffer"); Hickman v. Myers, 632 S.W.2d 869, 870 (Tex. Ct. App. 1982) ("A
parent cannot be said to have been damaged by the birth and rearing of a normal, healthy child."). In
significant part this concern is discussed as a problem of damages-their intangibility, their uncertainty, and the difficulty of balancing benefits and burdens. See, e.g., Public Health Trust v. Brown,
388 So. 2d 1084, 1085-86 (Fla. Dist. Ct. App. 1980) ("[Tlhe intangible but all-important, incalculable but invaluable 'benefits' of parenthood far outweigh any of the mere monetary burdens involved."); Rieck v. Medical Protective Co., 64 Wis. 2d 514, 518, 219 N.W.2d 242, 244 (1974)
("Every child's smile, every bond of love and affection, every reason for parental pride. . . . [T]hese
are intangible benefits, but they are nonetheless real.").
HeinOnline -- 95 Yale L.J. 266 1985-1986
Protecting Patient Choice
analytic problems arose. It was hard to view the child's physical defect as
a harm to the parent,19 9 yet parents often brought suit on their own behalf, alleging breach of duties owed to them.20 0
Delineations of duty also derive from definition of interest. It was easier
to tolerate the unusual nature of the injury involved in these cases if the
breach was of a duty that was familiar to negligence doctrine. Such a duty
might be an ordinary one of competent care, breached, for instance, by a
botched procedure or misdiagnosis. 20 1 Or it might be a duty of disclosure
stemming from the subsidiary choice interest under informed consent,
breached for instance, by a failure to disclose risks. 22 Where, however,
not only the injury but the duty as well was unusual, it was doubly difficult for a court to recognize this cause of action.20 3
One court solved these analytic problems in a procreation case by redefining the protected interest from physical well-being to an intangible concept of choice. Berman v. Allen 20 4 spoke of the plaintiff's injury as one of
being "deprived of the option of making a meaningful decision as to
whether to abort the fetus."' 20 5 There is little difficulty in recognizing the
birth of either a healthy or a defective child as an injurious consequence if
the protected interest is choice. Further, such a formulation easily recognizes parents as plaintiffs in their own right. The redefinition of protected
interest also allowed Berman to impose a rather unusual duty. In circumstances similar to those in which the court in Karlsons found no such
duty,208 the Berman court recognized an obligation to preserve the parents' choice by informing them about the availability of amniocentesis
199. See Howard v. Lecher, 42 N.Y.2d 109, 112, 366 N.E.2d 64, 66, 397 N.Y.S.2d 363, 365
(1977). Capron, supra note 52, at 642, effectively critiqued the treatment of parents as bystanders.
200. For a list of cases, see Collins, supra note 52, at 693-95 nn.81-84.
201. One of the first cases to allow recovery for the birth of a healthy child was Custodio v.
Bauer, 251 Cal. App. 2d 303, 59 Cal. Rptr. 463 (1967), in which a negligent surgical sterilization
failed to prevent pregnancy. See Troppi v. Scarf, 31 Mich. App. 240, 187 N.W.2d 511 (1971) (negligent filling of birth control pill prescription resulted in unwanted pregnancy); Dumer v. St. Michael's
Hosp., 69 Wis. 2d 766, 233 N.W.2d 372 (1975) (wrongful birth derived from negligent failure to
diagnose rubella).
202. See Harbeson v. Parke-Davis, Inc, 98 Wash. 2d 460, 472, 656 P.2d 483, 491 (1983) (health
care providers have duty "to impart to their patients material information as to the likelihood of
future children's [sic] being born defective").
203. Although in Karlsons v. Guerinot, 57 A.D.2d 73, 394 N.Y.S.2d 933 (1977), the court did
recognize the parents as appropriate plaintiffs, the only duty the court recognized was that of providing competent physical care. See supra text accompanying notes 50-65.
204. 80 N.J. 421, 404 A.2d 8 (1979).
205. Id. at 430-31, 404 A.2d at 13. Harbeson, 98 Wash. 2d at 478, 656 P.2d at 494, involving a
more typical breach of duty-the failure to inform a woman plaintiff of risks involved in becoming
pregnant while taking epilepsy medication-uses similar language: "Mrs. Harbeson ought to have
been informed in order to intelligently exercise her judgment whether to have further children."
206. Like the plaintiff in Karlsons, 57 A.D.2d 73, 394 N.Y.S.2d 933, the plaintiff in Berman
was in her late thirties at the time of the pregnancy. Both women claimed their doctors had a duty to
inform them about amniocentesis in order to determine possible fetal abnormality. In both cases the
babies were born with Down's Syndrome.
HeinOnline -- 95 Yale L.J. 267 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
even though no analysis of informed consent would ordinarily be triggered.20 7 With choice as the directly protected interest, such a duty of
disclosure was both more obvious and more legitimate. The touch-based
boundaries of disclosure under informed consent doctrine could be
discarded.
Despite its imposition of an unusual duty, Berman did not follow its
redefinition of interest with an altered standard of care. 08 Other courts,
however, have altered the standard of care in cases of heightened electiveness. The Iowa Supreme Court, for instance, although acknowledging that
it usually employed a medical community standard in cases of informed
consent, adopted a reasonable person standard in a case involving a vasectomy.20 9 It explicitly rested that decision on the elective and personal nature of the medical treatment involved. 0
Berman's redefinition of protected interest did alter its analysis in another crucial way. Although choice is routinely assessed by an objective
standard in informed consent actions where it provides the causal link to
Br
court used a subjective, individualized
physical injury, 211 the Berman
standard, asking only what these particular parents would have done. 12
This individualized approach emerges naturally where choice itself becomes the protected interest.2 13
Finally, with reference to analysis of damages, Berman asserted that
because the doctors neither caused nor could have ameliorated the baby's
defects, a comprehensive award of tort damages would be out of proportion to defendants' culpability. The court held that the parents could recover damages for emotional distress, but could not recover for the costs of
207. See supra Part II(B)(1).
208. It adopted the medical community standard. The court accepted without discussion the plaintiffs' characterization of the issue as one to be measured by "accepted medical standards." 80 N.J. at
424, 404 A.2d at 10.
209. Cowman v. Hornaday, 329 N.W.2d 422 (Iowa 1983); Cf. Cross v. Trapp, 294 S.E.2d 446
(W. Va. 1982) (adopting reasonable person standard in case involving male impotence as result of
prostate resection). West Virginia previously had no case law or statute governing informed consent.
210. Cowman, 329 N.W.2d at 427. In jurisdictions that routinely employ the lay standard in
informed consent cases any special impact of the procreation fact pattern is obscured. See, e.g., Harbeson, 98 Wash. 2d 460, 656 P.2d 483.
211. See 3 MAKING DEcISIoNs, supra note 2, at 197 (virtually all states employ objective standard of causation).
212. Berman, 80 N.J. at 431-32, 404 A.2d at 14. New Jersey apparently has no case or statutory
law on causation apart from Berman. Cf Harbeson, 98 Wash. 2d 460, 656 P.2d 483, which also
adopts without comment a subjective analysis of causation in a procreation case. Id. at 482-83, 656
P.2d at 494. Washington clearly employs an objective standard of causation in other informed consent
cases, making this decision unique to the procreation context. Cf Steele v. St. Paul Fire & Marine
Ins. Co., 371 So. 2d 843, 851 (La. Ct. App.), cert. denied, 374 So. 2d 658 (La. 1979) (in hysterectomy case involving young woman, court applied objective person standard but used "in the circumstances" provision of the standard to particularize it to an unusual degree to individual plaintiff).
213. The court did not comment on its divergence from the norm of objective causation. It may
simply have responded intuitively to the shift it had made in the structure of the action.
HeinOnline -- 95 Yale L.J. 268 1985-1986
Protecting Patient Choice
rearing the child. As a matter of sanctioning policy, the court's decision
might be wise, but the assertion that the doctors did not cause the baby's
defects is argumentative at best. It is true that the doctors did not cause
the baby's defects in the way that professional malpractice might have
caused such injuries. By the court's own logic, however, the protected interest here was choice. The doctors did cause a deprivation of choice. The
culpable conduct was a failure to inform that prevented the parents from
being able to choose to terminate the pregnancy. Thus, a more apt phrasing of the court's concern would be that invasion of the parents' interest in
competent care was more culpable than invasion of their interest in choice,
thereby justifying a difference in damages. 14 Regardless of whether one
agrees with Berman's resolution of the damages problem, the issue is appropriately one of policy about sanctioning, rather than one of causal connection. Defining choice as the protected interest helped to clarify the difference; Berman's explicit focus on comparative culpability rather than on
the existence of injury was a step in the right direction.
B.
The Limits of Professional Competence: Generalization of the
Patterns
Conflict of interest undermines the doctor's claim to authority, and special values or preferences intensify the patient's right to control. If either
factor is present, courts may alter the overall classification of the action or
shift sub-elements within standard analyses in order more clearly to identify and protect patient choice. These exceptions ought not to be narrowly
delimited. The issues are pervasive and the rationale is more inclusive
than courts have yet recognized. Individuals and their preferences are inherently unique. Professional competence can provide only limited resolu214. Berman might be characterized as specifying an additional set of circumstances in which
there may be recovery for infliction of emotional distress. It is probably more accurate, however, to
describe the court as limiting, for reasons of sanctioning policy, the recovery it is willing to allow for
invasion of its newly defined interest in choice. Other courts might prefer different solutions to the
problem of determining what extent and type of recovery is appropriate in these cases. A decision to
protect choice does not settle difficult issues of sanctioning policy; it merely clarifies the questions. See
Collins, supra note 52, at 695-700; infra Part IV.
The court's decision in Berman that emotional distress damages are the only appropriate remedy
for this deprivation of choice finds a precursor in Vaia v. Drago, 24 A.D.2d 888, 264 N.Y.S.2d 660
(1965). In Vara, the defendant failed to tell the plaintiff of his discovery, two months before her due
date, that her fetus had died. On appeal, the court affirmed an award of damages for mental anguish
only. Unless these damages are conceived as the result of outrageous intentional conduct, they were
unusual, especially in 1965. RESTATEMENT (SECOND) OF TORTS §§ 46-48, 905 comment c (1977).
Freestanding emotional distress damages have only recently been accepted in negligence actions. See,
e.g., Molien v. Kaiser Found. Hosp., 27 Cal. 3d 916, 616 P.2d 813, 167 Cal. Rptr. 831 (1980). Vara
can possibly be explained as a variant on the cases having to do with handling of dead bodies, PRosSER & KEETON, supra note 18, at 328-30, but the court makes no reference to those cases. More
likely, there was some inchoate sense of an injury akin to what I have labeled here as an interest in
patient knowledge and choice.
HeinOnline -- 95 Yale L.J. 269 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
tion of limited questions. The proposition that regulation of professional
competence cannot substitute for vigorous protection of patient choice is,
then, a matter of general rather than exceptional applicability.
1. Medical Uncertainty: The Basis of Electiveness
Implicit in any assumption that competence can adequately substitute
for choice is a medical world in which expertise either points in a single
direction or, at least, can best choose such a direction from among competing alternatives. Though the idea is comforting, no such world exists.215 A
moment's reflection will suggest long lists of things medicine does not fully
know or understand, from the cause of the common cold to the cure for
cancer. Although doctors are trained to present a face of decisiveness to
patients, they are often only sure about their uncertainty. 1 6
Medical uncertainty destroys the possibility of a single right answer,
leaving many answers to compete. Specializations and schools of thought
in medicine have strenuous and unresolved differences. In terms of quality
of care, traditional malpractice law can assure that a doctor's behavior is
proper as judged by the norms of at least one recognized school of
thought.2"' Strictly as a matter of competence regulation, it is fair that a
doctor should not be penalized for a careful, good faith recommendation of
a given reputable approach to care. Yet conformity to a particular school
of thought does not assure that that viewpoint best fits the circumstances
and preferences of the patient. Because there is no certainty about who is
right, the patient should receive information about divergent views and be
allowed to arrive at her own decision.
Moreover, competence regulation does not guarantee the best possible
or the most care; doctors are penalized only if they fall below a minimum
threshold of reasonable professional behavior, evidenced legally by what
the average doctor in good standing in her profession actually does.2"' Although doctors' knowledge or advice can be appropriately evaluated by
reference to what other competent doctors do, questions of how sure one
should be about the fact that no skull fracture is present,2" 9 that an abor215. J. KATz, supra note 7, at 165-206. Professor Katz, himself a doctor, comprehensively documents both the prevalence of uncertainty and doctors' discomfort in admitting it to their patients.
216. "There is no certainty about the available knowledge, but its uncertainty can be specified."
J. KATz, supra note 7, at 183-84 (emphasis in original). Medical training forces doctors to present at
least an impression of certainty, even in the face of uncertainty. Id. at 184.
217. PRossER & KErEON, supra note 18, § 32, at 186-87.
218. Id. Prosser notes that while the "averageness" concept refers to professionals in good standing, even "of these it is not the middle but the minimum common skill which is to be looked to." Id. at
187. Furthermore, some states still adhere to the locality rule, a further limitation on the standard of
care. Id. at 164.
219. Roark v. Allen, 633 S.W.2d 804 (Tex. 1982).
HeinOnline -- 95 Yale L.J. 270 1985-1986
Protecting Patient Choice
tion is complete, 220 that immature cells in excised tissue do not signal cancer, 221 that glaucoma, 22 2 heart disease,22 or cervical cancer2 24 is not developing, that a fetus does not have detectable genetic defect, 225 or that a
breast is not unnecessarily amputated 226 are not ultimately questions that
can be settled by medical expertise. At bottom, they are questions of allocating scarce resources, personal or societal.227 Expertise per se provides
data and experiential wisdom to inform decisions, but it cannot strike the
ultimate balance. Like other questions of utility and value, such issues
should be referred to the individual who will enjoy the benefits and suffer
the consequences of the choice.228
Both courts and doctors have been quick to accept uncertainty as a justification for flexing and softening the standards of professional competence, and appropriately so. They have been less quick to recognize uncertainty's implications for patient self-determination. 2 29 Although medical
220. Stills v. Gratton, 55 Cal. App. 3d 698, 127 Cal. Rptr. 652 (1976).
221. Jamison v. Lindsay, 108 Cal. App. 3d 223, 166 Cal. Rptr. 443 (1980).
222. Gates v. Jensen, 92 Wash. 2d 246, 595 P.2d 919 (1979) (en banc).
223. Keogan v. Holy Family Hosp., 95 Wash. 2d 306, 622 P.2d 1246 (1980) (en banc).
224. Truman v. Thomas, 27 Cal. 3d 285, 611 P.2d 902, 165 Cal. Rptr. 308 (1980) (en banc).
225. Karlsons v. Guerinot, 57 A.D.2d 73, 394 N.Y.S.2d 933 (App. Div. 1977).
226. Hanks v. Doctors Ranson, Swan & Burch, Ltd., 359 So. 2d 1089 (La. Ct. App.), cert.
denied, 360 So. 2d 1178 (La. 1978).
227. Societal judgment might, of course, override individual choice, as in the case of scarce resources (e.g., organ transplants), or because of public financing of health care. Helling v. Carey, 83
Wash. 2d 514, 519 P.2d 981 (1974) (en banc) (failure to test for glaucoma in patients under 40 is
negligent despite conformity to professional standard of care), offers one example of a judicially
administered societal intervention into the question of how much care/certainty is enough. The practical effect of Helling is to demand a standard response by all doctors to all similarly situated patients,
and to brand as culpable any professional who fails to make that response. This issue might be better
addressed, however, at the level of individual choice, as suggested here. For further discussion of
societal considerations, see infra text accompanying notes 326-33.
228. There is a reluctance to think of professions in market terms. For a classic description, see,
e.g., Hughes, Professions, 92 DAEDALUS 655, 657 (1963) (identifying distinctive traits of a profession,
including principle of credat emptor rather than caveat emptor); Newmark v. Gimbel's, Inc., 54 N.J.
585, 258 A.2d 697 (1969) (distinguishing professionals from commercial actors in terms of which
legal rules are appropriate for regulation of conduct). This view has hindered awareness of medical
care as a service traded for money, or as necessitating active consumer evaluation in order to maximize
utility. Recently a great deal more attention has been given to the need for active consumer decisionmaking and tighter market controls in health care, mainly as a result of concern about rising medical
costs. See, e.g., 42 C.F.R. § 476.2-3 (1984) (Department of Health and Human Services regulations
requiring disclosure by professional standards review organizations of costs and performance records
of medical care providers); A Special Symposium: Market-OrientedApproaches to Achieving Health
Policy Goals, 34 VAND. L. REv. 849 (1981).
Several major developments in law reflect and support this trend. See, e.g., Arizona v. Maricopa
County Medical Soc'y, 457 U.S. 332 (1982) (applying antitrust laws to professions); Bates v. State
Bar of Arizona, 433 U.S. 350 (1977) (analogous case, Arizona Supreme Court rule prohibiting professional advertising held unconstitutional); Virginia State Bd. of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U.S. 748 (1976) (state statute prohibiting professional advertising by pharmacists held unconstitutional).
229. The Washington Supreme Court's unusually intense awareness of how uncertainty limits the
effectiveness of competence regulation (in Helling and Gates) probably motivated its expansion of
protection for choice in Gates. See supra note 96.
HeinOnline -- 95 Yale L.J. 271 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
uncertainty heightens the need for professional advice, it also strengthens
the case for lay choice.
Choosing among alternative courses of action implicates individual
characteristics of the patient, such as career, age, and gender, as well as
personal attitudes and values, from religious belief to risk aversion. A
preference for surgical treatment of a back problem or, alternatively, for
long-term rest and traction, may depend on the patient's job or lifestyle. A
woman's preference for a radical mastectomy as opposed to conservative
surgery or chemotherapy may depend more on her body image, age and
career than on medically expert knowledge.2"'
The elective quality of plastic surgery and the personal character of
procreation are easily perceived. Less recognizable is the fact that the uncertainty and diversity of medical opinion necessarily turn much of medical decision-making into an exercise in electiveness. If cost-benefit comparisons (financial and otherwise) are not made by the patient, there is a
significant danger that decisions will reflect the doctor's attitudes and values rather than the patient's.23 1
2.
Conflict of Values: Close Analogue of Conflict of Interest
The conflict of loyalties in Mink v. University of Chicago was crystallized in a way that made recognition of the affront to patient autonomy
relatively easy. Conflicts of interest differ from what might be labeled conflicts of judgment or value mainly in terms of the degree to which such
230. For an illustration, see J. KATZ, supra note 7, at 90-93, 182-84. Similarly, a woman's
preference for separating the diagnostic biopsy from the final surgery may be controlled by her attitudes toward bodily disfigurement and risk aversion. Cf. Hanks, 359 So. 2d 1089 (no liability for
nondisclosure of possibility of separating biopsy from amputation; breast removed after erroneous
diagnosis).
There are countless other examples suggesting the need for a different allocation of authority between doctors and patients. Is it better to exercise or not to? Experts disagree. See A Warning About
the Exercise Experts, San Francisco Chron., Sept. 14, 1984, at 24, col. 1. The choice may depend
more on how vital exercise is to the patient's mental and emotional well-being or to a family's pattern
of activities than on purely medical recommendation. What form of birth control is best? The answer
may depend on sexual habits more than on medical conclusions about safety or effectiveness. Are
coronary artery bypass operations desirable? That could depend on what changes a person is willing
and able to make in exercise and stress patterns. Yaeger, Bettering the Odds: Cardiologists Focus
Efforts on Prevention of HeartAttacks, Wall St. J., July 2, 1984, at 1, col. 1; see also M. MILLMAN,
THE UNKINDEST CuT: LIFE IN THE BACKROOMS OF MEDICINE 90-151, 217-49 (1978) (discussing
medical management of mistakes and the overselling of coronary bypass in the face of its uncertain
value). Is it better to take time to do a fluid culture to determine whether antibiotics will be effective
against an infant's constant ear infections or to turn immediately to standard drugs? The choice may
depend on the weight one places on unknown risks associated with powerful chemicals. Etcetera.
231. J. KArz, supra note 7, at 173-74, describes a doctor whose recommendations in favor of
radical surgery rather than more conservative approaches were influenced more by his attitudes toward his father and his mentor than by scientific evidence. For other examples, see id. at 131-41,
175-84; Hilfiker, Making Medical Mistakes, HARPERS, May 1984, at 59. Hilfiker argues that doctors need to accept and have others accept their errors. This is dearly true. It also, however, underscores the need for sharing difficult decisions with the people whose lives depend on them.
HeinOnline -- 95 Yale L.J. 272 1985-1986
Protecting Patient Choice
crystallization is present. Even the best doctor is influenced by various
motives, concerns and goals-many of which may not be as benign from
the vantage point of patient interests as the medical research in Mink.
Concerns about income, prestige, and professional independence affect a
doctor's recommendations in indirect ways. Even relatively nonselfish factors such as commitments to particular scientific assumptions or medical
techniques, as well as personal or professional values, affect the judgments
made by doctors regarding the care of patients.
Moreover, doctors as a group are, like other groups, subject to error, to
intellectual or professional fashion, to blindspots and myopia. Grouptypical behavior is not necessarily either wrongful or incompetent, but
neither will it be carefully evaluated by standards of competence that are
set by the actual practice of the group itself. Blindspots associated with
group characteristics can bend doctors' preferences in ways that differentiate those preferences from those of their patients. For example, doctors as
a group are professionally committed to lifesaving. In the past they have
tended to interpret "lifesaving" as meaning the maximum extension of
biological life. Recently, with demands for "death with dignity," it has
become apparent that large numbers of lay persons have different preferences than those that doctors have chosen to implement.2 3 2 Similarly, for
years most doctors have been male. As a result, doctors as a group may
well have had different preferences about gender-related medical issues
such as birth control, procreation, and breast cancer than the population
they treated.23 3
232. Public policy concerning the right to die is being reconsidered in light of changing medical
facts and social values. As that process advances, disputes about the relative authority of doctors and
patients can be expected to increase. In Bartling v. Superior Ct., 163 Cal. App. 3d 186, 209 Cal.
Rptr. 220 (1984), a patient suffering from five serious diseases asked the hospital to remove his lifesustaining apparatus. The hospital refused and the superior court upheld the hospital. Although the
decision was reversed on appeal, reversal came after the patient died. More courts are accepting right
to die pleas. See, e.g., Satz v. Perlmutter, 362 So. 2d 160 (Fla. Dist. Ct. App. 1978), offd, 379 So. 2d
359 (Fla. 1980); In re Conroy, 98 N.J. 321, 486 A.2d 1209 (1985). But see Bouvia v. County of
Riverside, No. 159780 (Cal. Super. Ct. Dec. 16, 1983) (order denying preliminary injunction against
forced feeding). Despite this trend, many people still find themselves in conflict with individual doctors, hospitals and courts over their rights to control decisions about their treatment where death is
likely. An estimated five million Americans have made living wills stating their wish not to be artificially kept alive where there is no hope of recovery. Nelson, Doctors Debate Right To Stop 'Heroic'
Effort To Keep Elderly Alive, Wall St. J., Sept. 7, 1982, at 1, col. 1, 20, col. 2.
233. See generally G. COREA, THE HIDDEN MALPRAGrICE: How AmERICA MEDICINE
TREATs WOMEN AS PATIENTS AND PROFESSIONALS (1977) (tracing effects of male domination of
medical profession on health care received by women); B. EHRENREIC AND D. ENGLISH, COMPLAINTS AND DISORDERS: THE SEXUAL POLITICS OF SICKNESS (1973) (tracing history of institutional sexism in health care); Morrow, Women's Health Care and Informed Consent: Who Should
Decide What Is Best For Women-Patients or Doctors, 9 GoLDEN GATE U.L. REv. 553 (197879) (discussing factors that have contributed to current problems in women's health care and capacity
for change through use of informed consent); Note, Coerced Sterilization Under Federally Funded
Family PlanningPrograms, 11 NEw ENG. L. REv. 589 (1976) (discussing effectiveness of regulations guarding against coercive sterilization); Watchorn, Midwifery: A History of Statutory Suppres-
HeinOnline -- 95 Yale L.J. 273 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
Decisions made in a climate of conflicting values or judgments are every
bit as consequential to patients as those made when there are conflicts of
interest. And it is, after all, the fact that she will have to live with the
consequences of the medical choice upon which the patient's claim to authority preeminently rests. As the above examples suggest, there is good
reason to be alert to professional advisors' inchoate divisions of loyalty.
Indeed, such divisions should be presumed as a given of human nature;
the only meaningful safeguard is assurance of the opportunity for patient
rather than professional choice.
Not only have some types of conflict been more readily identifiable,
some have also been thought more culpable. Conflict of interest has been
traditionally viewed as a more serious problem than conflict of value and
judgment. The lesson of Mink, however, is that there need be no implication of doctor wrongdoing in the narrow sense of that word to justify aggressive protection of patient choice.
In comparison to the crass self-interest of acting to suppress a lawsuit,
the conflict of interest in Mink23 4 was relatively benign, involving research
presumably undertaken for the good of society. The comparatively good
intentions of these doctors might once have shielded them from criticism. 23 5 However, sensitivity concerning patient self-determination has increased. The concept of wrong itself has evolved: Wrongfulness here is the
illegitimate usurpation of authority. No matter how benign their goals, or
how competent their recommendations, the University doctors' research
28
had the potential to distort their decisionmaking. ' Mink allowed the jury
to consider, without reference to the constraints of negligence analysis,
whether the doctors' substitution of their judgment for that of their patients was permissible.2 3
The problem of unjustified paternalism in medicine transcends the
sion, 9 GoDEN GATE U.L. REV. 631 (1978-79) (tracing legislation affecting the midwife's
opportuni'ties).
234. Mink v. University of Chicago, 460 F. Supp. 713 (N.D. Ill. 1978).
235. For extraordinary documentation of relatively casual attitudes toward patient consent where
the advancement of science is involved, see J. KATZ, EXPERIMENTATION wrrH HUMAN BEINGS
(1972); Boffey, Medical Experimentation:Everybody's a Critic, San Francisco Chron., Dec. 2, 1984,
at 19, col. 1.
236. Biased decisionmaking is the reason universities and other research organizations funded by
the government are now required to conduct extensive institutional review of all research conducted on
human beings. E.g., Dept. of Health and Human Services, Regulations on Protection of Human
Subjects, 45 C.F.R. §§ 46.101-46.117 (1984). See Robertson, The Law of Institutional Review
Boards, 26 UCLA L. Rv. 484 (1979). In addition, many private hospitals and treatment centers are
establishing their own ethics and decisionmaking review procedures. See Kleiman, Hospital Philosopher Confronts Life's Basic Issues, N.Y. Times, Jan. 23, 1984, at B1, col. 5.
237. 460 F. Supp. at 717. In the context of bodily contact, differences of judgment and value are
presumed and autonomy protected; it is not necessary to demonstrate a conflict of interest in order to
justify the patient's right of control. As I have demonstrated, however, current legal doctrines do not
adequately protect against unjustified paternalism where only information is involved. See supra Part
II.
HeinOnline -- 95 Yale L.J. 274 1985-1986
Protecting Patient Choice
boundaries of traditional conflict of interest.23 8 The medical profession
holds relatively strong ethical and disciplinary ideals proscribing conflicts
of interest, 23 9 but its norms about deference to patient autonomy in instances of conflict of judgment or value are seriously underdeveloped.
Moreover, legal analysis of fiduciary and professional responsibility is
only gradually becoming attentive to these subtler problems of allocating
authority. 2 0 Because contractual arrangements that govern divisions of
authority in most commercial agency relationships are all but absent in
professional-client settings, 4 1 the need for legal rules that establish the
limits of professional authority is especially great.
238. See generally J. CHILDRESS, WHO SHOULD DECIDE? PATERNALISM IN HEALTH CARE
(1982) (classifying types of paternalism and analyzing when it is unjustifiable in medical
decisionmaking).
239. One of the seven Principles of Medical Ethics adopted by the A.M.A. in 1980 states that "a
physician shall . . . strive to expose those physicians deficient in character or competence, or who
engage in fraud or deception." JUDICIAL COUNCIL, AM. MEDICAL AWS'N, CURRENT OPINIONS at ix
(1984) (Principle II); see also id. § 4.04 at 14 (requiring physician to disclose ownership in health
care facilities); id. § 6.03 at 24 (prohibiting fee-splitting). "The symbol of the profession ... portrays
a group whose members have altruistic motivations and whose professional activities are governed by
a code of ethics which ... condemns misuse of professional skills for selfish purposes." Becker, The
Nature of a Profession, in 2 THE SIXTY-FIRST YEARBOOK OF THE NATIONAL SOCIETY FOR THE
STUDY OF EDUCATION 27, 36 (1962).
Although the A.M.A. Code includes general requirements that the physician shall "respect the
rights of patients", id. at ix (Principle IV), and shall "deal honestly with patients," id. (Principle II),
Katz traces the long history of nondisclosure and failure to share decisions with patients. J. KATz,
supra note 7, at 1-29. Professor Katz notes the direct conflict between the medical norm of "custody"
and the legal norm of personal liberty. Id. at 2.
240. The doctrine of informed consent in medicine is part of that effort. J. KATz, supra note 7, at
28, points out that the "recent interest in disclosure and consent," following a "history of silence,"
could not be expected, as yet, to have altered centuries of contrary medical tradition.
Struggles over professional authority and client control have also emerged within the legal profes-
sion. See Bell, Serving Two Masters: IntegrationIdeals and Client Interests in School Desegregation
Litigation, 85 YALE L.J. 470 (1976); Burt, Conflict and Trust Between Attorney and Client, 69
GEO. L.J. 1015 (1981); Spiegel, Lawyering and Client Decisionmaking: Informed Consent and the
Legal Profession, 128 U. PA. L. REV. 41 (1979); c. Jones v. Barnes, 463 U.S. 745 (1983) (failure of
attorney to honor client directive regarding nature of arguments made on appeal did not deprive client
of effective counsel). The new ABA Model Rules of Professional Conduct make a greater effort to
deal with the issue of client choice and professional disclosure than either the comparable medical
rules or the earlier legal rules. MODEL RULES OF PROFESSIONAL CONDUCT Rule 1.4 (1982). Explanations of the legal background for the model rules refer to the RESTATEMENr (SECOND) OF AGENCY
§ 381 (1958) as a basis for broad disclosure obligations. See T. MORGAN & R. ROTUNDA, PROBLEMS
AND MATERIALS ON PROFESSIONAL RESPONSIBILrrY 175, 177 (2d ed. Supp. 1983).
241. RESTATEMENT (SECOND) OF AGENCY § 376 (1958) anticipates that the extent of an agent's
authority will be determined by the terms of the agreement between agent and principal. Although the
doctor-patient relationship typically originates from a contract, the terms of the relationship are not
usually explicitly bargained for. Instead, the professional model has presumed unquestioning client
trust, see Hughes, supra note 228, at 657, and what courts are fond of calling "abject dependence."
See, e.g., Cobbs v. Grant, 8 Cal. 3d 229, 242, 502 P.2d 1, 9, 104 Cal. Rptr. 505, 513 (1972). These
presumptions are obviously in conflict with respect for patient consent and control.
HeinOnline -- 95 Yale L.J. 275 1985-1986
The Yale Law Journal
c.
Vol. 95: 219, 1985
Conclusion of Part III
Where conflict of interest disqualifies the doctor or heightened electiveness intensifies the patient's claim to autonomy, courts have acted to overcome distortions and omissions that have plagued existing legal protection
of patient autonomy. The rationale for such exceptions needs to be generalized. Professional competence does not satisfy the goals of patient autonomy. Medical uncertainty forces a high degree of election in decisionmaking, and extra-medical values necessarily shape resulting choices. The
conflicts of value and judgment that are inherent in all human decision
are both consequential and problematic. Thus, preemption of authority in
circumstances where competent adults may be consulted regarding their
own fate is unjustifiable. Patient choice ought to be a fully and independently protected legal interest.
IV.
A.
A
NEW INTEREST IN MEDICAL CHOICE
The Interest is Desirable and Congruent with Other Doctrines
Serious deficiencies exist in the protection presently accorded to patient
autonomy. Both as a matter of omission and as a matter of exception, the
present doctrinal schema is inadequate and inconsistent. A new model for
the allocation of authority between doctors and patients is needed. A patient should be able to avail herself of a doctor's services without depriving
42 Patients
herself of the opportunity to control significant care choices.
should, of course, be free to delegate authority, but such delegation should
not be required or presumed. Giving patients control over medical choices
would delimit doctors' authority and their responsibility. At the same
time, such control implies that new obligations would be placed on doctors
to facilitate and defer to patients' choices. To effectuate such a relationship, the direct creation of an independent interest in medical choice
would be preferable to the indirect vindication now derived through protection of other, related interests.
Creating direct legal protection for patient autonomy would be consis242. The principle of patient control involves a degree of conflict with professional ideology. Doctors are not supposed to be accountable, insofar as professional decisions are involved, to laypersons.
Freidson observes that "the most strategic distinction [between professions and non-professions] lies in
legitimate, organized autonomy. . . the right to control its own work. . . . [P]rofessions are deliberately granted autonomy, including the exclusive right to determine . . . how the work should be
done." E. FREIDSON, PROFESSION OF MEDICINE: A STUDY OF THE SOCIOLOGY OF APPLIED
71-72 (1972). Although the conflict might in theory be resolved by differentiating between the "how" and the "whether" of work, doctors have not traditionally been attentive to such
distinctions. See J. KATz, supra note 7, at 30-47. The concern about professional independence is
partly directed against organizational control by laypersons rather than against control by individual
patients. Even medicine's organizational independence, however, is threatened by increasingly bureaucratic forms of health care delivery and lay-initiated financial controls on medical decisionmaking.
KNOWLEDGE
HeinOnline -- 95 Yale L.J. 276 1985-1986
Protecting Patient Choice
tent with other recent legal developments. The legal duty to facilitate patient choice by disclosing information is likely to be construed as an affirmative duty, and perhaps, therefore, to be resisted under common law
traditions. Yet the line dividing an affirmative duty from a prohibitive one
is largely semantic. The duty to protect patient choice may also be described as prohibitive: It instructs the doctor-agent not to substitute her
judgment for that of the patient. Viewed in this light, an interest in choice
is similar to intangible interests in personal sanctuary that are already
protected by the law.
Constitutional privacy cases often involve medical or medically-related
decisions. 24" These cases, involving birth control, 2" abortion, 4 5 form of
treatment, 246 and refusal of care at death,24 stress the importance of individual autonomy. The government interventions that are constitutionally
prohibited in medical privacy cases involve interference with decisionmaking, rather than physical intrusion.2 48 Constitutional development is often
broadly rooted in common law principles; in this instance the public law
is somewhat in advance of the private.2
49
The intangible decisionmaking
focus of the constitutional privacy interest presages the change in private
law proposed here.
As a matter of private law, no explicit extension of a privacy interest to
medical contexts has yet been attempted. 250 But existing tort privacy ac243. Factually, these cases may involve either a doctor allied with a patient against the intrusions
of the government, e.g., Roe v. Wade, 410 U.S. 113 (1973) (antiabortion laws), or a doctor allied with
the state in seeking to bar a patient's medical choice, e.g., Andrews v. Ballard, 498 F. Supp. 1038
(S.D. Tex. 1980) (patient's right to die). They do not directly address the allocation of authority as
between the doctor and the patient. Indeed, there is some evidence that the constitutional privacy right
has been vested more in doctors than in patients. Asaro, JudicialPortrayalof the Physician in Abor-
tion and Sterilization Decisions: The Use and Abuse of Medical Discretion, 6 HARV.
WOMEN'S
L.J.
51, 51-52 (1983).
244. Griswold v. Connecticut, 381 U.S. 479 (1965).
245. Roe, 410 U.S. 113.
246. Andrews, 498 F. Supp. 1038.
247. See, e.g., Bartling v. Superior Ct., 163 Cal. App. 3d 186, 209 Cal. Rptr. 220 (1984); Satz v.
Perlmutter, 362 So. 2d 160 (Fla. Dist. Ct. App. 1978), affid, 379 So. 2d 359 (Fla. 1980).
248. Some of the constitutional decisions are, like their private law analogues, preoccupied with
degree of physical intrusion as the way to calibrate patient autonomy. E.g., In re Quinlan, 70 N.J.
10, 355 A.2d 647, cert. denied, 429 U.S. 922 (1976) (right of permanently noncognitive patient to be
terminated from life support systems). But many involve no physical intrusion, e.g., Andrews, 498 F.
Supp. 1038 (patient's right to choose form of treatment), and most emphasize patient autonomy as the
underlying goal, e.g., Griswold, 381 U.S. 479 (right to use birth control). A similar transition from
physical to intangible definitions of privacy occurred in Fourth Amendment protection of personal
sanctuary. Katz v. United States, 389 U.S. 347 (1967) (electronic surveillance through phone taps
may violate Fourth Amendment even though there was no physical entrance into area occupied by
petitioner).
249. Cf. Rogers v. Okin, 478 F. Supp. 1342 (D. Mass. 1979) (for discussion of case, see supra
note 160), affd in part, rev'd in part on other grounds, 634 F.2d 650 (1st Cir. 1980).
250. In discussing possible expansion of the tort interest in privacy, Prosser observed the potential
connection between constitutional and tort privacy cases. PROSSER & KEETON, supra note 18, § 117,
at 866-67. He did not comment specifically on medical applications, however. Cf. RESTATEMENT
(SECOND) OF TORTS § 652A comment c (1976) (discussing possible connections between constitu-
HeinOnline -- 95 Yale L.J. 277 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
tions do protect other interests in intangible personal sanctuary, such as
2 51
reputation or'security, from unwanted publicity. Among presently recognized privacy actions, invasion of a patient's interest in choice somewhat
resembles appropriation.25 The tort of appropriation asserts a kind of
property interest in one's likeness and name. An interest in medical choice
would assert what amounts to a proprietary interest in information possessed by a doctor concerning one's condition, options and fate. The appropriation at issue in medical cases would be usurpation of the decisional
authority that depends upon possession of such information.
The opportunity for maximum feasible control of medical fate would
certainly seem to be as important an interest as control of name or likeness, reputation or seclusion. Moreover, although government intrusions
on privacy may be justifiable as necessary means to fulfill social goals, a
doctor's authority derives only from the patient and should extend no furPrivate usurpation of authority, even with
ther than the patient decrees.
25
.
culpable
is
good intentions,
The interest in freedom from emotional distress also has much in common with the interest in medical choice. Both interests protect aspects of
personal sanctuary. The physical context of health care choice provides an
analogue to the physical injury or proximity to physical danger that many
2
jurisdictions require for redress of freestanding emotional distress. "4 In
recent years, courts have become less skeptical about the genuineness of
freestanding emotional injury.2 55 Because objections to recovery for emotional distress paralleled the resistance to stronger protection of patient
choice, the increased acceptance of emotional injury claims may ease the
way for an interest in medical choice.
Although emotional disturbance is one common consequence of deprivational privacy and tort privacy doctrines).
251. Tort privacy doctrines proscribe invasion of what Professor Fleming calls the "more sophisticated" int~hgible interests. J. FLEMING, supra note 129, at 193.
252. See id. at 210-11 (discussing appropriation); PROSSER & KEETON, supra note 18, § 117, at
851-54.
253. Some privacy torts require extreme offense or outrage to create a cause of action. RESTATEMENT (SECOND) OF TORTS §§ 652B, 652D, 652E (1976) (intrusion upon seclusion; publicity given to
private life; and publicly placing person in a false light). Cf Hirsh, CurrentPerspectives on the Tort
of Outrage, 87 CASE & COM., Nov.-Dec. 1982, at 9 (discussing application of action to medical
patient's rights). But not all privacy actions require extremely culpable behavior. Thus, appropriation
need not involve offensive or outrageous conduct. RES'ATEMENT (SECOND) OF TORTS § 652C (1976).
Moreover, not all privacy disputes involve unmitigatedly antisocial conduct. News professionals sometimes run afoul of privacy strictures although they, like doctors, are presumptively engaged in a sodally legitimate enterprise. Likewise, agents who usurp authority of principals are not excused by
their good intentions. RESTATEMENT (SECOND) OF AGENCY § 14 (1958). See supra Part III(B)(2).
254. These requirements occur mainly where the action is for negligent infliction, or where the
emotional injury is to a bystander. See, e.g., RESTATEMENT (SECOND) OF TORTS §§ 46-48, 312-13
(1976).
255. See Molien v. Kaiser Found. Hosp., 27 Cal. 3d 916, 616 P.2d 813, 167 Cal. Rptr. 831
(1980) (recovery of emotional damages for husband of woman negligently informed she had syphilis).
HeinOnline -- 95 Yale L.J. 278 1985-1986
Protecting Patient Choice
tion of choice, 256 leaving injuries to choice to be redressed only as a subset
of the emotional distress action would be inappropriate. Such an approach
would ignore the frequency and centrality of physical consequences arising from invasions of patient choice. More important, just as analysis of
patient choice under traditional informed consent doctrine defines that interest as a particular species of damage rather than as an interest in autonomy per se,257 subsuming choice under emotional distress would have
the same effect. Injury-specific characterizations distort the interest in
choice, subjecting it to inappropriate analytic and remedial restraints. 2 8
Analogies to present doctrine may also be found if the protection of
choice is understood as imposing an affirmative duty. Although the common law does not readily create such duties, it will do so if the defendant's
relationship with the plaintiff makes the imposition of such duties appropriate.259 The medical relationship is founded upon a contract whose substance is caretaking and whose character is fiduciary. 2 0 There seems little
theoretical reason to balk at broadened duties of disclosure to protect patients' interest in an informed choice when other affirmative obligations
are already imposed on doctors.2"'
Doctors are universally conceded to be fiduciaries; as such they have
special duties to serve their clients' interests. Patients have been redefining
their interests in the direction of more active participation in decisionmaking.2" 2 In the wake of such redefinition, the nature of fiduciary obligation
must also change to stress more advising and less deciding. Even if the
doctor is conceded to have some authority as an agent, agency doctrine
256. E.g., Berman v. Allan, 80 N.J. 421, 404 A.2d 8 (1979) (recovery of damages for emotional
anguish suffered by parents as result of deprivation of opportunity to choose abortion of Down's
Syndrome fetus).
257. See supra Part II(B).
258. Some limits on recoverable damages for an invasion of choice might be desirable. See infra
text accompanying notes 294-314.
259. See PRossvt & KEErON, supra note 18, § 56, at 373-82; RESTATEMENT (SECOND) OF
TORTS §§ 314-314B (1976).
260. There are other such relationships to which a duty to enhance choice might theoretically be
extended. Lawyers also have contractual and caretaking responsibilities. See supra note 240. Yet the
underlying concerns of the medical relationship-bodily condition and fate-are ones in which claims
to personal autonomy are uniquely compelling. That context seems an appropriate one in which to
explore affirmative duties to protect autonomy.
261. Doctors currently may be held liable in tort for several kinds of inaction. Failure to treat can
constitute violation of duties of competent care. Failure to disclose risks of treatment may generate
liability under informed consent rules. Cf. Tresemer v. Barke, 86 Cal. App. 3d 656, 150 Cal. Rptr.
384 (1978) (by virtue of fiduciary relationship doctor obligated to warn former patient about dangers
of IUD of which doctor was made aware after treatment of patient).
262. See 2 MAKING DECISioNS, supra note 2, at 221-25 (96% of patients want to be told about
diagnosis of terminal illness; 94% want to be told everything about their medical condition; and in
deciding between aggressive and supportive therapy in terminal cases, 79% of public thought decision
should be made by patient; only 24% of doctors agreed and only 19% said such decisions actually are
made that way).
HeinOnline -- 95 Yale L.J. 279 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
2 63 Both
emphasizes the agent's accountability to the principal's control.
agency and fiduciary principles require disclosure by the agent-fiduciary
as an obligation independent of the substantive fairness (here competence)
of the transaction, especially where there is any possible division of loycall for increased
alty.2' " Particularly with regard to the professions, the
26 5
intensified.
has
expert-fiduciaries
client control over
Even commercial actors who are neither fiduciaries nor professionals
are increasingly subject to affirmative duties to disclose. In ordinary contract settings, each party is presumed to be acting in her own interest; yet
disclosure may be demanded to enable a weaker party to make more informed decisions. Nondisclosure may create a claim for damages or a de26 6
fense to enforcement of a contract. The trend toward requiring disclosure is strongest when a party has special knowledge or expertise. As
imposed in tort law generally, affirmative duties to disclose potential dangers may help to prevent physical injury. But because such duties to warn
are not necessarily coupled with a requirement to remove the risk itself,
the emphasis on disclosure to enhance plaintiff's choice-making is analogous to the situation here. If a landlord may be required to disclose known
2 67
crime risks of a neighborhood, or manufacturers the possible hazards of
268 imposing affirmative duties upon doctors to inform their paa product,
tients' choices seems readily justifiable.
Although the duties proposed here are in some ways broader than these
analogous examples, 6 9 the differences are appropriate. Unlike most product manufacturers and landlords, doctors have a one-to-one relationship
with their patients, facilitating personal consultation and discussion. Furthermore, unlike either landlords or manufacturers, doctors have fiduciary
were
and professional responsibility for their patients. Those constraints
270 they
deference;
and
traditionally seen as obviating a need for disclosure
provided a basis for trusting the doctor more than the ordinary contractual
263. RESTATEMENT (SECOND) OF AGENCY § 14 (1958).
264. See supra note 168.
265. See supra note 240.
claim of
266. See, e.g., Obde v. Schlemeyer, 56 Wash. 2d 449, 353 P.2d 672 (1960) (basis for
Wash.
damages where seller fails to disclose presence of termites in apartment); Sorrell v. Young, 6
App. 220, 491 P.2d 1312 (1971) (rescission available where seller knew of defect, buyer was unaware,
and defect was not apparent); Slater v. KFC Corp., 621 F.2d 932 (8th Cir. 1980)(duty to disclose
also
where superior knowledge is not within the reasonable reach of less experienced party); see
REsTATEmENT (SECOND) OF CONTRACTS § 161(b) (1979).
487
267. See, e.g., O'Hara v. Western Seven Trees Corp., 75 Cal. App. 3d 798, 142 Cal. Rptr.
(1977).
268. E.g., Davis v. Wyeth Laboratories, Inc., 399 F.2d 121 (9th Cir. 1968). For a discussion of a
causation standard in relation to product liability warnings, see supra note 126.
269. Manufacturers have no duty to tell about alternative products, or landlords about alternative
buildings, but only to warn regarding the particular product or building they place on the market.
270. See Becker, supra note 239, at 27 (discussing code of ethics as guarantee of professionals'
trustworthiness and altruism).
HeinOnline -- 95 Yale L.J. 280 1985-1986
Protecting Patient Choice
actor. Increasingly, however, the professional's responsibility is viewed not
as a mandate to preempt authority, but as a responsibility to educate, and
then defer to, the patient's own decisions. Accordingly, that doctors' affirmative responsibility for disclosure should be greater than that of commercial actors is unsurprising.
B.
How The Interest Would Work
An interest in medical choice could be administered under either contract or tort doctrine. Either analysis would derive the duty to protect the
patient's interest in choice from the doctor-patient relationship-one in
order to vindicate reasonable expectations, the other in order to protect the
substantive value of autonomy itself.
1. Contract Analysis
Because patients have been deemed incapable of individual bargaining
about expert services, duties undertaken through a contract for professional care have been given content and specificity through negligence policy rather than through contract analysis. I have argued, however, that
although patients may be incapable of supervising the quality and administration of care, they are capable, indeed uniquely so, of balancing ultimate costs and benefits of care decisions. Moreover, they are capable of
determining the extent to which they wish to allocate decisionmaking authority to their doctors. Thus, the rationale for adopting a standardized
tort analysis does not extend to issues of decisionmaking and allocation of
authority; these matters could appropriately be analyzed under contract
doctrine. Were such an approach adopted, the entire analytic paradigm
would be reversed. Rather than invasion of patient choice being one subtype of injury causation within a professional negligence framework, professionally negligent care would constitute one species of breach of
contract.
Explicit agreements concerning the general allocation of authority or
about specific care choices will likely be honored under either analytic
regime. Traditionally, however, patients have not bargained with their
doctors about decisionmaking authority. In the absence of explicit agreement, the problem, contractually, would be to imply a reasonable term.
How much delegation of authority should be implied?
Current doctrines show courts to be unwilling to deem decisionmaking
authority automatically transferred by the act of hiring a doctor.27 ' Be271. See supra note 139. General contract law also frowns on unilateral decisionmaking that
results from disparate power or adhesion contracting. E.g., Tunkl v. Regents of Univ. of Cal., 60 Cal.
2d 92, 383 P.2d 441, 32 Cal. Rptr. 33 (1963) (invalidating agreement between hospital and entering
HeinOnline -- 95 Yale L.J. 281 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
yond the existing rules, both as a matter of expectation and reliance, patients today assume that they will be given the opportunity to make decisions if they wish. Anecdotal and empirical evidence suggest that most
patients wish to be informed and to participate more actively in the management of their care."' 2 Indeed, patient dissatisfaction with health care
frequently centers on doctors' failure to provide sufficient information and
dialogue."' 3 However, even patients who wish to exercise choice remain
dependent on doctors for information. Although public knowledge about
medicine has greatly increased, patients still rely on doctors for education
and options.27 4
and counsel, particularly about their individual condition
By contrast, assuming that payment schemes are adjusted to compensate
for time spent, doctors themselves have little, if any, legitimate reliance
interest in resisting disclosure or in controlling patients' decisions; they
certainly have a less intense interest than do patients. Moreover, doctors
are already adapting to patients' growing expectations of involvement,
making responsibility to disclose to and consult with patients less foreign
7 5 Even if doctors' expectations about authan it might once have been.
thority differ, they have reason to know, and under contract principles
27
therefore to be bound by, the intent of patients. 6 Particularly in their
role as fiduciaries, doctors should be held to defer to the expectations and
reliance of their patients.
Thus, although express bargaining should be encouraged, where no explicit term is agreed to, patient control of decisionmaking should be the
term implied into the contract. Given the tradition of medical paternalism,
so.
patients who wish to opt out of such responsibility could easily do
277
A
protection.
legal
greater
need
Those who seek involvement will likely
of
pattern
other
any
for
doctor could be expected to bargain 2explicitly
78
require.
to
decisionmaking she wished
patient exculpating hospital).
2, at
272. See 1 MAKING DECISIONS, supra note 2, at 17; 2 MAKING DECISIONS, supra note
survey).
empirical
of
(results
221-24
273. See Enright, supra note 10, at 76, 90 (discussing evidence that high percentage of malpractice suits are in part caused by failure of doctor to establish good communication with patient).
and 78
274. See J. KATz, supra note 7, at 227 (speaking on doctor's crucial role as teacher)
information).
for
needs
individual
patients'
to
respond
to
doctors
(noting need for
275. See supra note 14.
276. See RESTATEMENT (SECOND) OF CONTRACTS § 201 (1979).
277. See J. KATz, supra note 7, at 1-29 (tracing the history of paternalism in western medicine).
and
For an especially thoughtful series of articles contrasting medical paternalism, legal imperialism,
& MFD.
other alternatives, see Relman, The Saikewicz Decision: A Medical Viewpoint, 4 AM. J.L.
AM. J.L.
233 (1978); Baron, Medical Paternalismand the Rule of Law: A Reply to Dr. Relman, 4
& MED. 337 (1978); and Buchanan, Medical Paternalismor Legal Imperialism:Not the Only Alternativesfor Handling Saikewicz-type Cases, 5 AM. J.L. & MED. 97 (1979).
278. The danger that doctors would use their power to insist that patients give them broader
Condecisionmaking authority than the patient might wish would be forestalled by three factors: (1)
adhetracts would be subject to review under principles governing unconscionability, good faith and
sion; (2) Doctors may actually be relieved to reduce their responsibility for others' fate; (3) Doctors
HeinOnline -- 95 Yale L.J. 282 1985-1986
Protecting Patient Choice
Once the term governing authority is established, contract analysis
would be straightforward, though not simple. A patient explicitly or impliedly contracting for maximum disclosure and choice would be deprived
of her expectation if the doctor failed to give material information. To put
her where she would have been had the contract been performed, it would
be necessary to determine the difference in value between where she
would have been had she been given information and choice, and where
she is now. Because tort analysis of an interest in choice would also require projection and valuation of what a patient would have chosen, this
process would, to a substantial degree, parallel the analysis proposed in
279
the next section.
2.
Tort Analysis
Although existing tort approaches are seriously deficient, the weight of
tradition may cause tort analysis of these issues to be continued. Tort
analysis could provide adequate protection if patient choice became an independent and fully protected interest in its own right. Analysis of an
independent interest in choice would differ from existing analyses in a
number of ways.
Duties involving disclosure of knowledge lend themselves to analyses of
intentional conduct. Where a defendant has fiduciary and contractual responsibility to a plaintiff, and has primary access to information essential
to the plaintiff's recognition and exercise of choice, the defendant must
know that failure to disclose that information is substantially certain to
cause invasion of the plaintiff's interest. On the other hand, because the
doctor typically claims other justifications for nondisclosure and has no
specific intent to harm, the simpler and more punitive assumptions frequently applied to intentional torts seem inappropriate to most situations
of medical nondisclosure. Privacy torts encompass both negligent and intentional conduct without being confined to one or the other,2"' and this
approach would also be appropriate for patient choice.
A duty to disclose would be triggered by the possession of information
are increasingly having to compete for health care dollars. Patients wishing more authority in decisionmaking are likely to have the bargaining power to achieve that goal.
279. In Gates, for example, where the patient became blind, she probably would not have lost her
sight had she opted for further testing. To assess her loss under contract theory, her expectation
should be determined by figuring the value of sight, discounted both by the possibility that she might
not have undertaken the additional tests and by the projected outcome of having the tests, offset by any
value there is in being blind. Calculation of damages based on prediction of probabilistic outcomes has
been undertaken in contracts cases, e.g., Locke v. United States, 283 F.2d 521 (1960) (projected share
of business); Rombola v. Cosindas, 351 Mass. 382, 220 N.E.2d 919 (1966) (projected winnings of
racehorse). For a discussion of damages under the tort theory, see infra text accompanying notes
294-314.
280. See RESTATEMENT (SECOND) OF TORTS §§ 652A-652E (1976).
HeinOnline -- 95 Yale L.J. 283 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
important and relevant to the patient, rather than by a proposal to touch.
This approach would reverse the relationship between information and
choice that is created under current doctrine. At present, the requirement
of consent determines the necessity of disclosure. Because, however, choice
and reflection, the essential
arises out of and depends upon knowledge
28
itself.
knowledge
point of access must be
Where disclosure is triggered by touching, occasions for performance of
the duty are rather clearly defined. Although a duty to disclose based
upon possession of knowledge per se might seem comparatively unbounded,28 2 it would not be as unlimited as it might seem. As a function
of their existing responsibilities for care, doctors already have the knowl2s8
edge needed to carry out this proposed new duty. The doctor's duty to
know would continue to be regulated in present fashion as a matter of
professional competence. 2 " At issue in relation to an interest in choice
would be what the doctor must disclose of what she does know.
Certainly the duty would not be to disclose all that the doctor knows,
but only what is materially relevant to the patient at the time of the disclosure. The doctor should affirmatively offer the following information:
(1) material clinical observations or test results that describe the condition
of the patient at any stage of care; (2) interpretation of this information by
the doctor and her advisers, including material judgments and conclusions
based on the data; and (3) material possible responses that the patient
might elect in light of the information and the possibilities known to the
28 5 must be understood to
doctor. In each aspect of the duty, "material"
281. The importance of proper disclosure of information is heightened by the fact that the same
information that would allow the patient to control medical choices would also permit her to make
other personal choices that depend on her understanding of her medical condition.
282. Even under the earlier rule, courts worried about imposing a duty that would make it necessary to give patients a medical education. See, e.g., Canterbury v. Spence, 464 F.2d 772, 782 n.27
(D.C. Cir.), cert. denied, 409 U.S. 1064 (1972).
283. Because doctors are uniquely (both in the sense of "well" and in the sense of "exclusively")
situated to provide information vital to patient choice, the duty to avoid the costs of nondisclosure
should be placed on them. See generally G. CALABRESI, THE COSTS OF ACCIDENTS (1969) (presenting theory of cheapest cost avoidance as most appropriate method of allocating cost of accidents).
284. Lack of knowledge, however, would be more exposed under the proposed rule than under
the present one. See infra note 289.
285. Useful work has already been done to define the concept of materiality for purposes of determining the scope of medical disclosure. The concept of materiality under informed consent doctrine
inversely relates probability and seriousness: Even relatively remote possibilities may have to be disclosed if their consequences are very serious. See, e.g., Canterbury v. Spence, 464 F.2d at 788.
Waltz and Scheuneman originally proposed the test of materiality and attempted to define it. Waltz
& Scheuneman, supra note 30, at 638-41. See also Halligan, The Standard of Disclosure by Physicians to Patients: Competing Models of Informed Consent, 41 LA. L. REv. 9 (1980) (discusses tort of
deceit as antecedent upon which courts have relied in developing medical disclosure rules); 2 MAKING
see
DECISIONS, supra note 2 (gives empirical indications of what public thinks is material). But
Epstein, supra note 17, at 124 (arguing that there are "no principled limits" to doctrine's requirements for disclosure).
Traditional informed consent doctrine might offer other acceptable methods of narrowing the dis-
HeinOnline -- 95 Yale L.J. 284 1985-1986
Protecting Patient Choice
extend beyond the doctor's certainties.2" 6 Although some data, judgments
and options will be too remote to require mention, disclosure of material
possibilities in each of the categories is essential. Inappropriate management of uncertainty has been responsible for many of the failures of choice
28 7
protection analyzed here.
A concept such as "focus of attention" would help to delimit the duty of
disclosure. Of the nondisclosures analyzed here as being vital to choice,
many occurred where a doctor made an underlying decision not to act. 288
That inaction was not a matter of inadvertance. It was an intentional
judgment, the focus of attention and decision, to which liability could have
attached had it been incompetent as a matter of professional care. If the
relevant knowledge and the decision based upon it were not actually
within the doctor's awareness, the issue again becomes whether it should
have been-a question of professional competence.2 89
Under a focus of attention approach, information of the type described
should be forthcoming with regard to any presenting symptom, problem
or condition that has brought the patient to the doctor for care. Material
information should also be forthcoming any time the doctor actually has
or acquires information as a result of observations made or initiatives undertaken during care, even if the information is not related to the condition that brought the patient to the doctor.2 90
The extent of the duty to disclose general information about possibilities that are not the focus of either the doctor's or the patient's concern at
the time of the encounter raises the most difficult issues. Anyone might
benefit from a cardiogram to screen for heart problems or a tonometry test
to identify glaucoma. Explanations of such possible courses of action and
reasons why they should or should not be explored would be endless.
closure obligation. Thus, an exception for matters of common knowledge might be employed. In addition, traditional exceptions for mental incapacity, emergency and waiver should apply.
286. J. KATz, supra note 7, at 166-69, 186-89.
287. E.g., Nardone v. Reynolds, 333 So. 2d 25 (Fla. 1976); Hanks v. Doctors Ranson, Swan &
Burch, Ltd., 359 So. 2d 1089 (La. Ct. App. 1978); Keogan v. Holy Family Hosp., 95 Wash. 2d 306,
622 P.2d 1246 (1980).
288. E.g., Roark v. Allen, 633 S.W.2d 804 (Tex. 1982) (decision that skull indentations did not
require x-ray to determine fracture); Gates v. Jensen, 92 Wash. 2d 246, 595 P.2d 919 (1979) (decision not to test further for glaucoma).
289. Such a rule will create a more precise accountability about knowledge under standards of
professional competence. Doctors might find themselves choosing whether to claim they knew but did
not disclose (thereby invading the interest in choice), or to admit that they did not know (thereby
falling below the threshold for competence).
290. Thus, a patient might come in for an asthma problem, and the doctor might note a skin
condition that could be cancerous. Even if the doctor competently concludes it is probably not cancerous, and that no care is necessary, the doctor should disclose the skin condition because it had been a
focus of the doctor's attention. If the doctor did not note the condition, the issue would again be one of
competence regarding whether she should have. See also Ratzan, Unsolicited Medical Opinion, 10 J.
MED. & PHIL. 147 (1985) (suggesting criteria for and techniques of offering unsolicited medical
opinions).
HeinOnline -- 95 Yale L.J. 285 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
They would threaten what courts have sought to avoid: the obligation to
give each patient a medical school education. However, as soon as specific
data linking general information to a particular individual is noted, atten291
tion is focused and a need to disclose would arise.
Once it is determined that a duty of disclosure applies, breach of that
duty ought to be judged not by the standards of expert behavior but by the
standards appropriate to protection of patient autonomy. Like the extent
of the doctor's knowledge, the accuracy of the doctor's disclosure would
still be a question of professional competence, evaluated as it is at present.
The occasion for and scope of disclosure, however, would be analyzed in
terms of what constitutes reasonable disclosure where one on whom another depends has special access to information relevant to that other's
interests, a standard akin to the agent-fiduciary standard. An agentfiduciary must disclose all information that she knows or should know
292 Given recent documentation
that the principal would desire to have.
that patients want much more medical disclosure and choice than they
currently enjoy,29 3 such a duty should extend disclosure well beyond the
norms of traditional medical practice.
Once duty and breach are established, sanctioning analysis would also
differ if choice were the protected interest. The crucial difficulties stem
from the uncertainty regarding what would have happened if the patient's
interest in choice had not been invaded. Cooper-Stephenson and Saunders 294 have argued persuasively that an analysis that balances probabili-
ties to achieve an either/or result is appropriate to determinations regarding events that have occurred. Questions of substantive liability are
generally of this genre. Those authors urge, however, that a proportional
analysis is fairer where a court must assess intrinsically uncertain events.
Such events cannot be analyzed as more probable than not because they
are inherently unknowable. Most issues of damage valuation are of this
type.
29 5
291. The disclosure required ought to be less extensive with regard to general background information than with regard to its individual application. Such a duty might require the doctor to note
briefly the general issue ("Smoking is dangerous to health"), and to provide more extensive information regarding the individual ("Your cough sounds as if you may be having lung problems. We can
check it out in the following way . . ... "). Also, the obligation to disclose ought not to recur more
than once for a single problem of a single patient unless the doctor acquires relevant new data about
the patient or the problem.
292. See RESTATEMENT (SECOND) OF AGENCY § 331 (1957).
293. See supra note 14.
294.
Throughout this section I rely upon K. COOPER-STEPHENSON & I. SAUNDERS, PERSONAL
83-114 (1981). They present both a theory and a review of the cases
with reference to appropriate measurement of damages.
295. Although both Canadian and English law have adopted this approach, courts in the United
States have been less enthusiastic. Id. at 113. Nevertheless, commentators in the United States have
endorsed the proportional approach to such determinations. E.g, King, supra note 27, at 1396-97;
Tom on Torts, Kentucky Allows Accident Victim's Recovery for Increased Risk of Future Harm, 27
INJURY DAMAGES IN CANADA
286
HeinOnline -- 95 Yale L.J. 286 1985-1986
Protecting Patient Choice
If one adopts Cooper-Stephenson and Saunders' analysis, a balance of
probabilities approach would appropriately serve to determine whether
the defendant's nondisclosure more probably than not materially deprived
the patient of an interest in choice. Two additional uncertainties would
then remain to be assessed in valuing any resulting injury. What would
the plaintiff, properly informed, have chosen? What, medically speaking,
would have been the result if a different course had been selected? Analysis of these issues should employ what Cooper-Stephenson and Saunders
call a simple probability test (proportional), because neither involves
events that have actually occurred. Each requires the reconstruction of
what would have happened had the defendant not interrupted the chain of
events by failing to disclose.
Such interruption, the invasion of choice, results in the loss of an uncertain chance of a preferable outcome. That chance can be valued as a matter of assessing damage.2 9 The familiar maxim that courts should not
readily allow uncertainty to prevent vindication of an invaded interest
comes into play. 297 Damages are adjustable along a monetary continuum.
Statistical tools are more readily employed to apportion and value uncertainty for purposes of remedy than to determine substantive liability. 298
Moreover, characterizing the issue as one of valuing injury allows the
multiple sub-issues involved in these cases to be resolved in an orderly and
sequential manner. The court does not have to compress the sub-issues
into a single ultimate question as it must where they are treated as a
single question of factual cause. 99 In order to determine the value of the
plaintiff's loss of choice, three questions must be answered.
First, how likely is it that an alternative path would have been chosen?
The subjectivity of this first branch of uncertainty makes fact-finding with
AM. TRIAL LAW. A. L. REP. 302, 302 (Sept. 1984). Some courts have also embraced such a perspective, e.g., Davis v. Graviss, 672 S.W.2d 928, 931-32 (Ky. 1984). But see Rosenberg, The Causal
Connection in Mass Exposure Cases: A "PublicLaw" Vision of the Tort System, 97 HARV. L. REV.
851, 874 n.99 (1984) (arguing that in sporadic accident cases proportional rule of recovery is less fair
than preponderance of evidence rule).
296. See King, supra note 130, at 1396-97.
The response of the courts to questions of causation and valuation involving preexisting
conditions and claims for future consequences has been largely unsatisfactory. Their failure to
distinguish between the functions of causation and valuation, or to identify and value rationally
the true interests lost, has created a serious gap in the remedial structure. Courts have had
difficulty perceiving that a chance of avoiding some adverse result or of achieving some
favorable result is a compensable interest in its own right.
Id. at 1354.
297. REsTATEMENT (SECOND) OF TORTS § 912, at 479 (1979).
298. K. COOPER-STEPHENSON & I. SAUNDERS, supra note 294, at 83-114.
299. See J. FLEMING, supra note 132, at 108 (noting problems created by placing "a wide spectrum of inquiries" within "monistic" doctrine of proximate cause). For a discussion of problems arising from courts' tendencies to treat what should be issues of valuation as questions of causation, see
King, supra note 130, at 1354.
HeinOnline -- 95 Yale L.J. 287 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
regard to it the most difficult. 30 0 Although no perfect way of determining
3 01 the
an answer exists, as determinations in other circumstances show,
court could derive some probabilistic estimate of alternative courses. The
projected decision of this individual, rather than "a reasonable person,"
should provide the standard; any other standard fails to protect the very
0
autonomy that lies at the heart of the interest. '
Subjective states are susceptible of objective proof; they are demonstra30 3
ble through evidence of conduct or words observable by others. Because
the issue here involves a hypothetical rather than an actual event, however, objective evidence will not directly or definitively demonstrate what
the individual's choice would have been. Objective evidence, however,
why this individual might or probably would have chosen
could elucidate
30 4
path.
a given
300. The word "subjective" sometimes refers to inner, i.e., mental or emotional, states and sometimes to particularization regarding a given individual. See Eisenberg, The Responsive Model of Contract Law, 36 STAN. L. REV. 1107 (1984) (discussing these as two main spectra within contract
theory). The reconstruction of patient choice actually involves both kinds of subjectivity. In the informed consent caselaw, both types of "subjective" standards have been almost totally rejected as being
too unreliable. See, e.g., Canterbury v. Spence, 464 F.2d 772, 790 (D.C. Cir.), cert. denied, 409 U.S.
1064 (1972).
301. For instance, where a plaintiff claims damage through reliance on a promise, the court will
have to reconstruct what that individual would have done in the absence of the promise, under circumstances in which hindsight and self-interest will taint the evidence. See, e.g., Feinberg v. Pfeiffer
Co., 322 S.W.2d 163 (Mo. Ct. App. 1959) (in determining whether plaintiff relied on promised
retirement stipend, one issue was whether plaintiff would have quit work if promise had not been
made). Like medical choice, reliance involves both reconstruction of what an individual would have
chosen to do, and what the outcome of that choice would have been. See Eisenberg, Donative
Promises, 47 U. CHI. L. REv. 1, 18 (1979); see also infra note 309; cf supra note 279 (discussing
contract cases calculating damages based on prediction of probabilistic outcomes).
302. The approach proposed here parallels Professor Eisenberg's recommendation that most contract law principles should be individualized, and that they may be either objective or subjective,
depending upon the circumstances. Eisenberg, supra note 300, at 1111. Patient choice cases present
more problems both of fairness and of administrability than does the average contract case, but in
light of what is at stake, those problems should be manageable. Eisenberg's justification for individualization-that "a major goal of contract law is to facilitate the realization of individual objectives,"
id. at 1111 -applies with much greater force here. Bodily fate is a more intense and personal interest
than are "highly differentiated goods," and no "cover" at all is possible.
303. See, e.g., Lucy v. Zehmer, 196 Va. 493, 84 S.E.2d 516 (1954) (court used objective words
and acts of sellers to determine whether they were joking when they agreed to sale of property);
RESTATEMENT (SECOND) OF CONTRACTS § 18 comments a and c, § 19 (1981) (explaining objective
standards for evaluating assent to contract); cf, e.g., Lange v. Hoyt, 114 Conn. 590, 159 A. 575
(1932) (Christian Scientists not required to have child undergo surgery to mitigate damages).
304. If, for example, she made her living as a graphic artist or a photographer, she could persuasively claim that she would have gone to more than average lengths to protect her sight. A few courts
have allowed consideration of such individual factors under the "in the circumstances" aspect of the
reasonable causation standard in informed consent. See, e.g., Steele v. St. Paul Fire & Marine Ins.
Co., 371 So. 2d 843 (La. App.), cert. denied, 374 So. 2d 658 (La. 1979) (implicitly reflecting special
attention courts give to individual choice in procreation cases). Objective information about individuals' hypothetical preferences regarding future medical care has become increasingly available. Several
states have adopted provisions allowing individuals to make binding advance statements about care in
circumstances of terminal illness. See, e.g., Natural Death Act, CAL. HEALTH & SAFETY CODE §§
7186-7195 (West Supp. 1985). Legal provisions have been adopted that effectuate clearly stated individual preferences regarding organ donation. See, e.g., Uniform Anatomical Gift Act, CAL. HEALTH
HeinOnline -- 95 Yale L.J. 288 1985-1986
Protecting Patient Choice
Moreover, a requirement to reconstruct the choice of a given individual
does not preclude the introduction of non-individualized objective information. Thus, the court would find useful testimony about choices that other
people actually make when confronted with similar risks and odds. 30 5
However, the role of such information should be delimited as bearing only
on the credibility of the claim. The individual nature of the ultimate question should be clearly preserved.
The potentially self-serving nature of the plaintiff's testimony should be
less worrisome if the court requires and uses objective evidence to determine only a proportional chance that a different path would have been
taken. If jury evaluation still seems likely to reflect an excess of sympathy
or lack of adequate evaluation of credibility, the court could apply some
intermediate protective device. For example, a court could employ a presumption that this individual would have decided as a reasonable person
would have, but could then allow the individual to rebut the presumption
by showing persuasive objective evidence of individual reasons for
deviation. 30 6
Second, the jury would have to determine what would have happened if
a different choice had been put into effect. 30 7 Expert testimony would
probably be essential here, but again, such testimony should be restricted
to assessment of factual medical data. Statistical predictions about outcome
odds in a given case would not prove what would actually have happened
in the case at bar. But courts routinely use such general data to determine
308
an individual's damages.
Finally, the value of where the plaintiff stands under the actual chain
of events would need to be compared with where she would have stood as
a result of the alternative course of events, discounted by the appropriate
probability assessments. Comparison of values could begin with reasonable averages, but should also factor in objectively demonstrable individual
& SAFETY CODE §§ 7150-7157 (West 1970).
305. Such data would be of two kinds: (1) statistics showing what people actually have chosen
when faced with comparable decisions; and (2) statistics regarding prospective preferences as established by public opinion polls. See, e.g., 2 MAKING DECISIONS, supra note 2, at 136, 244 (percent of
public desiring to be informed of condition, including diagnosis of cancer).
306. Cf. Cunningham v. Charles Pfizer & Co., 532 P.2d 1377 (Okla. 1975) (presumption plaintiff would have heeded warning on product rebuttable by evidence of what reasonable person would
have done).
307. In Gates v. Jensen, the answer would have been relatively easy. Further testing would have
had a high probability of discovering the glaucoma. 92 Wash. 2d 246, 250, 595 P.2d 919, 922 (1979).
In some cases, however, answers would present more difficulties. For example, if a woman alleges she
would have opted for a lumpectomy rather than a radical mastectomy, the jury would have to judge
the increased probability of having a recurrence of cancer had she followed the alternative course.
308. Thus, courts awarding damages for wrongful death do not know how long an actual individual would have lived, or how much she would have earned. But they use general actuarial probabilities (which take account of as many individual factors as possible) to render those uncertain projections adequately certain. See RESTATEMENT (SECOND) OF TORTS § 92 comments c, d, and e (1976).
HeinOnline -- 95 Yale L.J. 289 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
variables that affect medical preferences, such as age or occupation, or
even personal attitudes or circumstances where those could be established
by evidence more probative than the plaintiff's simple assertion.
Applying these proposed techniques would be difficult, but the incentive
to resolve the complicated problems associated with the vindication of
choice derives from the depth of the concern about patient autonomy. As
elsewhere, running some risks with inaccurate damages is preferable to
not following the logic of doctrine at all.30 9 Moreover, the methods of calculation suggested here should reduce concern about uncertainty and
about recoveries being unduly influenced by hindsight.
Issues regarding types of compensable harm would remain. The intangibility of the interest should not obscure the substantial and often physical nature of the consequences flowing from its invasion. As I have
demonstrated, the redefinition of interest suggested here would significantly alter the analysis of duty, breach, causation and value in regard to
traditional physical injury. Thus, even if courts were to restrict redress for
invasions of choice to physical consequences, realignment of the protected
interest would significantly increase legal protection of patient choice.
Identification of an intangible interest in choice could also allow recovery for less traditional categories of harm. Courts could evaluate consequences of a substantial but not necessarily "physical" or "injurious" (as
socially judged) harm, for example, the birth of an unwanted child or the
undesired prolongation of death. Yet, as a matter of sanctioning policy,
courts need not deem compensable all consequences of invasions of choice.
For example, although emotional distress damages would constitute a par31 0
ticularly likely result of invasions of this interest, courts could restrict
such recoveries." 1 ' However, categorical limits on emotional distress dam1
ages in medical contexts seem, if anything, to be eroding. '
Under certain conditions, an invasion of choice might not produce any
substantial consequences, physical or otherwise. Thus, for instance, if the
doctor had not told Ms. Gates of the test indicating possible glaucoma, but
she did not, in fact, develop the disease, the doctor would technically have
309. For other illustrations of legal redress in conditions of substantial uncertainty, see Davis v.
Graviss, 672 S.W.2d 928 (Ky. 1984) (projection of uncertain future injury); Feinberg v. Pfeiffer Co.,
322 S.W.2d 163 (Mo. Ct. App. 1959) (reliance on promise); Locke v. United States, 283 F.2d 521
(1960) (value of bargained-for chance); Rombola v. Cosindas, 351 Mass. 382, 220 N.E.2d 919 (1966)
(uncertain expectation); RESTATEmENT (SECOND) OF TORTS §§ 910, 924 (1979) (uncertain prospective consequences of present injury).
310. See, e.g., Capron, supra note 52, at 639-45.
311. See, e.g., Berman v. Allan, 80 N.J. 421, 404 A.2d 8 (1979). Most courts dealing with
wrongful birth actions have allowed emotional distress damages. See, e.g., Jacobs v. Theimer, 519
S.W.2d 846 (Tex. 1975); Naccash v. Burger, 223 Va. 406, 290 S.E.2d 825, 830-31 (1982).
312. See, e.g., Molien v. Kaiser Found. Hosp., 27 Cal. 3d 916, 616 P.2d 813, 167 Cal. Rptr. 831
(1980); Sullivan v. O'Connor, 363 Mass. 579, 587-88, 296 N.E.2d 183, 188-89 (1973); Hirsh, supra
note 253, at 9-10.
290
HeinOnline -- 95 Yale L.J. 290 1985-1986
Protecting Patient Choice
violated the interest in information and choice but the consequences here
would be mainly intangible and symbolic. Two possible methods of treating such cases emerge. The court could award nominal damages, or it
could award general damages, as is done in privacy actions. 3 13 The latter
approach would better redress the purely dignitary element of the new
interest. However, because so many demonstrably consequential invasions
of patient choice have been ignored under the existing doctrinal scheme,
even adoption of the former approach would allow dramatically strengthened protection of patient autonomy. An interest is not delegitimated because in a particular instance its invasion produced little demonstrable
harm. Finally, unless the doctor harbored specific or malicious intent to
14
deprive a patient of choice, punitive damages would be inappropriate.
3.
Contract or Tort?
Several possible grounds for choosing between the two analyses might
be suggested. First, the indexing of cases under one or the other doctrinal
category may influence the selection of norms and concepts to be used in
the analysis. Contract analysis is traditionally rooted in respect for individual choice and might more effectively protect that concern. By contrast,
the standard professional negligence action has so dominated tort analysis
of medical relationships that it might be difficult for an interest in choice
to achieve sufficient independence within the tort domain.
Second, tort has traditionally provided the locus of redress for physical
injury. However, both fields have expanded their compass, with contract
now recognizing physical and emotional injuries"' and tort redressing
even negligent injuries to economic, 318 emotional,31 7 and privacy31 8 interests, as well as to more traditional interests in physical well-being and
property. There seems little reason to distinguish on this somewhat outmoded basis.
Finally, a tort approach makes a stronger statement about public policy
regarding patient autonomy, while contract, deferring to the parties' election, remains more neutral. Under the proposals I make here, however,
the outcomes under the two analyses would, in fact, converge. Under my
tort analysis, the interest would emanate from public policy but would be
subject to individual choice (waiver) by the patient, or with greater diffiSee NV.
313.
PROSSR, HANDBOOK OF THE LAW OF TORTS
117, at 815 (4th ed. 1971);
STATEMENT (SECOND) OF TORTS § 904 (1976). These could be based on some standard amount
§
was more than nominal, or on a jury appraisal of the facts and circumstances.
314. See RESTATEMENT (SECOND) OF ToRT § 908 (1976).
315. See, e.g., Sullivan v. O'Connor, 363 Mass. 579, 296 N.E.2d 183 (1973).
316. See PROSSER & KEETON, supra note 18, § 95A, at 679-81.
317. RESTATEMENT (SECOND) OF TORTS §§ 46-48, 313 (1976).
318. Id. at § 652.
HeinOnline -- 95 Yale L.J. 291 1985-1986
REthat
Vol. 95: 219, 1985
The Yale Law Journal
culty, by the doctor if she required waiver as a condition of contracting to
provide services. Under the contract analysis, I have proposed that, based
both on empirical data about patient expectations and on public policy
values, courts should read into the contract, as an implied term, patient
control of decisions unless the parties explicitly agree to the contrary.
Again, then, the contract would include a term requiring patient control
as the background rule, alterable only by the affirmative and mutual decision of the parties.
Under contract analysis there remains some danger that, in implying a
term, courts would turn to custom of the trade and supply a term based
on traditionally paternalistic medical custom. But the needs and expectations of doctors should not exclusively define what is customary. Moreover, the expectations of both doctors and patients are changing. Public
policy points, as it does in other consumer contexts,"1 9 toward special judicial concern for protection of consumers' (patients') reasonable
expectations.
C. Policy Effects
1. Are Patients Capable of Choice?
The interest proposed here would place a great deal of responsibility on
patients. Some may fear that many individuals will be unable to discharge
the responsibilities that autonomy would impose.32 0 Rationality is inevitably limited; motives and wishes are frequently not conscious. Although
these are powerful arguments for seeking personal support and expert advice, the same limits apply to the rationality and motives of doctors, relatives or friends. The burden of consequences provides the compelling reason to place final authority with the affected individual rather than with
others who advise or care for that individual.
2.
Effects on Health Care Outcomes
If patients had more knowledge and control, experts would have less.
Would poorer health be the result? 2" Medical decisions involve both uncertainty and conflicts of judgment and value. Neither experts nor society
can judge what is best for an individual better than the individual herself.
The quality of patient choice will, of course, depend on the quality of
319. See, e.g., RESTATEMENT (SECOND) OF CONTRACrS § 211 (1981) (protection against contravention of consumer expectations in standardized agreements).
320. See J. Kxrz, supra note 7, at 26-27. See generally R. ButT, supra note 5 (discussing
complex motives affecting medical choice in a number of specific instances).
321. This is perhaps the core objection to proposals to strengthen patient choice. See J. KArz,
supra note 7, at 27.
HeinOnline -- 95 Yale L.J. 292 1985-1986
Protecting Patient Choice
information provided by the doctor. But assuming adequate performance
of that obligation, patient-made decisions should generally yield outcomes
that are preferable as evaluated by the ultimate consumer, the patient.
Moreover, there is evidence that patients who are well-informed progress better than do those who are treated more paternalistically.3 22 Such
patients implement treatment plans more fully3 3 and can aid the doctor's
diagnostic process. 24 More debatable, but not to be overlooked, is evidence that the mobilization of patients' psychic resources is critically important in the struggle for health. 2 5
3.
Effects on the Distributionof Health Care
Wealth already buys greater access to health care services. 32 6 Any sys-
327 If
tem that increased patient autonomy might exacerbate that effect.
choices were more thoroughly subjected to cost-benefit analysis by individ322. See Fiore, Fighting Cancer-One Patient's Perspective, 300 NEW ENG. J. MED. 284
(1979); Mumford, Schlesinger, Glass, Patrick & Curedon, A New Look at Evidence About Reduced
Cost of Medical Utilization Following Mental Health Treatment, 141 AM. J. PSYCHIATRY 1145
(1984). There is also evidence that information can reduce emotional problems which accompany
physical illness. See Comment, When the Truth Can Hurt: Patient-MediatedInformed Consent in
Cancer Therapy, 9 U.C.L.A.-ALASKA L. REV. 143, 151 & n.33, 152 & n.34 (1980).
323. Noncompliance is a serious problem in health care. See J. KATz, supra note 7, at xiv and
sources cited in Introduction n.2; Press, The Predispositionto File Claims: The Patient'sPerspective,
12 LAW, MED. & HEALTH CARE 53, 54 (April 1984) (patients' noncompliance reported by some
studies as high as 50-60%). Noncompliance likely indicates patients who have "voted with their feet"
when they do not feel in control of their treatment. The very term "noncompliance" suggests orders
given by the doctor to a passive patient.
324. See J. KATZ & A. CAPRON, supra note 30, at 89-90. Doctors not providing necessary instruction regarding self-care may be liable for negligent care. See, e.g., Crosby v. Grandview Nursing
Home, 290 A.2d 375 (Me. 1972); see also LeBlang & King, supra note 189, at 24-26.
325. See N. CoUsINS, ANATOMY OF AN ILLNESS AS PERCEIVED BY THE PATIENT: REFLECTIONS
ON HEALING AND REGENERATION 11, 14-23 (1979); K. PELLETIER, MIND AS HEALER, MIND AS
SLAYER 10-12 (1977); 0. SIMONTON, S. MATHEWS-SIMONTON, J. CREIGHTON, GETTING WELL
AGAIN 4-12 (1978).
326. Many people have no health insurance coverage. See BayerJustice and Health Care in an
Era of Cost Containment, 9 Soc. RFSP.: JOURNALISM, L. MED. 37, 43-47 (1983). For a discussion
of the severe problem of access to health care, see 1 PRESIDENT'S COMM'N FOR THE STUDY OF
ETHICAL PROBLEMS IN MEDICINE AND BIOMEDICAL AND BEHAVIORAL RESEARCH, SECURING ACCESS TO HEALTH CARE 90-113 (1983) [hereinafter all volumes cited as SECURING AccESs]; Starr,
Medical Care and the Pursuitof Equality in America, in 2 SECURING ACCESS, supra at 1-22. For a
discussion of the difficulty of access to high cost procedures like organ transplant or renal dialysis, see
J. AREEN, P. KING, S. GOLDBERG & A. CAPRON, supra note 32, at 852-861; Knox, Heart Transplants: To Pay or Not to Pay, 209 SCIENCE 570 (1980) (reviewing Department of Health and
Human Services criteria for Medicare/Medicaid reimbursement for new technologies), and for relatively low cost procedures such as abortion, see, e.g., Harris v. McRae, 448 U.S. 297, 329-37 (1980)
(Brennan, J., dissenting); id. at 337-48 (Marshall, J., dissenting).
327. See the discussion of equity concerns under a market controlled system in Havighurst, Competition in Health Services: Overview, Issues and Answers, 34 VAND. L. REv. 1117, 1142-43 (1981).
Individual decisionmaking may be less influential because third-party payors play such a central role
in access to health care. CONGRESSIONAL BUDGET OFFICE, CONTAINING MEDICAL CARE COSTS
THROUGH MARKET FORCES 1-2 (1982). But see Rushefsky, A Critiqueof Market Reform in Health
Care: The "Consumer-Choice Health Plan," 5 J. HEALTH POL., POL'Y & L. 720, 726-37 (1981)
(suggesting significance of third party reimbursement may have been overestimated).
HeinOnline -- 95 Yale L.J. 293 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
ual consumers, rather than being made mostly by supposedly wealthneutral professionals, would comparatively poorer patients end up with
even less equitable access to health care than they now have?
Doctor-directed decisionmaking deprives the poor as well as the rich of
autonomy and self-determination. Whatever professional ideology would
have us believe, professionally competent choice is not in fact immune to
economic incentives. 328 The tendency to preempt the choices of poor patients may be significantly greater than it is where wealthy patients are
involved.3 2 9 Thus, poor patients may have an even greater stake in
strengthened protection for autonomy than do the affluent. The difficulty,
however, is that self-determination is a hollow concept to anyone who
lacks the resources to implement choice.
The problem of social justice in access to health care is essentially
moral and political. Professional expertise cannot and should not provide
society with answers about how scarce resources should be distributed any
more than it should provide answers about individual utilities regarding
medical choice. If there are not enough kidneys to go around, it is disingenuous to pretend that medical expertise can do more than to advise
regarding some factors that are likely to be relevant in deciding who ought
to get one. Expertise can inform those who must make decisions regarding
comparative medical risks and benefits. But it should not decide who
should receive a transplant any more than it can decide who wants higher
levels of certainty. Admittedly, it is politically easier to respond to demands for a "right" to health care by citing the impersonal commands of
"medical indication" 330 than it is to articulate and endorse the social, political and economic criteria by which society will determine distribution
of health care. 332 Yet offering professional expertise as the source of an328. Some hospitals and doctors refuse to provide services without advance deposits or proof of
insurance. 1 SECURING ACCESS, supra note 326, at 100-101.
329. See 2 MAKING DECISIONS, supra note 2, at 20 (doctors tell less to patients they perceive as
being less able to understand, a factor highly correlated with socioeconomic status); Note, Coerced
Sterilization Under FederallyFunded Family PlanningPrograms, 11 NEw ENG. L. Rev. 589, 595
(1976) (describing coercive practices of medical providers in urging sterilization of welfare mothers).
330. See, e.g., Cowan v. Meyers, 3 Civ. 22987 (Cal. Ct. App. filed June 10, 1983) (challenging
such restrictions on state-paid health care services). One major effort to introduce the criterion of
medical necessity was the Professional Standards Review Organization (PSRO) legislation designed to
use peer review to cut costs for Medicare and Medicaid on the basis of whether services were necessary. 42 U.S.C. § 1320(C)(3) (1982). See Blumstein, The Role of PSROs in Hospital Cost Containment, in HOSPITAL COST CONTAINMENT.: SELECTED NOTES FOR FUTURE POLICY 461 (M. Zubkoff,
I. Raskin & R. Hanft eds. 1978) (assessing contribution PSROs can make to hospital cost containment). Private insurers also limit their coverage of discretionary services by contract in order to reduce
costs. Havighurst, supra note 327, at 1127.
331. See G. CALABRESI & P. BOBBITr, TRAGIC CHOICES 186-91 (1978); Baily, "Rationing"
and American Health Policy, 9 J. HEALTH POL., POL'Y & L. 489 (1984) (discussing reluctance of
President's Commission to use or explore concept of "rationing" as opposed to "allocation"); Havighurst, Health Care Cost-Containment Regulation: Prospects and an Alternative, 3 AM. J.L. &
MED.309, 313-15 (1977) (regarding difficulty of political confrontation with problems of cost con-
HeinOnline -- 95 Yale L.J. 294 1985-1986
Protecting Patient Choice
swers to questions of social justice is at best incomplete and at worst
deceitful.
Furthermore, if professional expertise is allowed to regulate the distribution of health care resources, its adequacy as a surrogate for individual
choice is further undermined. If doctors allocate care, their loyalty to the
best interests of their patients is necessarily diminished, just as it is by
other conflicts of interest. A doctor who accepted as optimal a given level
of certainty about the presence of glaucoma or about the completeness of
an abortion might claim to be acting not only as a rational expert, but also
as an implementer of social policy. 3 2 That doctor could not, however,
simultaneously claim to be acting on behalf of a patient who places an
unusually high value on her eyesight or on her lifestyle as a nonparent.
The strands of professional competence need to be disentangled from
those of social justice, just as they need to be separated more thoroughly
from those of individual choice. The question of what limits should be
placed upon individual choice by considerations of social justice is beyond
the focus of this Article.33 3 I have argued only that, in regulating medical
decisionmaking, doctrine should maximize individual autonomy rather
than decisionmaking by professional experts.
4. Effects on Consumption and Cost of Health Care
Increasing demand for medical services is a major cause of rising costs
of health care. 34 Increased cost is a problem for individuals, for organiza335
tions that finance health care benefits, and for the economy as a whole.
Given greater knowledge and control, individuals may increase some kinds
of health care consumption. At a minimum, more doctor time would be
required to inform and consult with patients regarding their options. Patients who are made aware that additional tests are available to determine,
for example, the health of a fetus, the risk of glaucoma, or the possibility
of skull fracture might opt for such additional tests.
However, a central premise of much of the current literature on medical cost control is that greater patient choice would yield a significant reduction in health care consumption. 3 ' Indeed, unless consumers have
meaningful control over the implementation of medical recommendations,
tainment and cost benefit analysis).
332. For example, by keeping costs down.
333. For a discussion of issues of social justice in health care, see, e.g., 1 SECURING ACCESS,
supra note 326, at 18-47; Veatch, What Is a "Just" Health Care Delivery?, in ETHICS AND HAL'r
POLICY 127, 131-42 (R. Veatch and R. Branson eds. 1976).
334. CONGRESSIONAL BuDGET OFFICE, supra note 327, at 1.
335. 1 SECURING ACCESS, supra note 326, at 184-85.
336. Blumstein & Sloan, supra note 154, at 860-61, 894-95; Marmor, Boyer & Greenberg, Medical Care and Procompetitive Reform, 34 VAND. L. REV. 1003, 1011-16 (1981).
HeinOnline -- 95 Yale L.J. 295 1985-1986
Vol. 95: 219, 1985
The Yale Law Journal
many cost-control incentives will be ineffective."' If fully informed, patients might decline many types of recommended surgery or drugs as being unnecessary or undesirable. For example, there has been much debate
about termination of care by patients demanding a right to die, or greater
control over the surroundings and conditions under which death or birth
will occur.3 38 In particular, a deepened knowledge about the value of selfcare and prevention, along with a healthier respect for the limits and uncertainties of medicine, would minimize the unrealistic search for absolute
protection against danger and mortality. 3 9
5.
Effects on Liability Burdens of Doctors
Recognition of an independent interest in patient information and
choice would impose a new or at least a more extensive responsibility, and
hence a broadened liability, on doctors. Commentators have expressed
concern that liability burdens on doctors are already too great. Yet given
what has been said both about patient expectations and about cost control,
this new responsibility does not seem inappropriate. Moreover, choice
protection would actually reduce doctors' liability in ways that would offset this expansion.
Empirical evidence suggests that even when undesirable medical outcomes occur, the greater the degree to which the patient participates and
3 40
is informed, the less likely she is to file a malpractice claim. Furthermore, greater protection of patient choice would relieve doctors of some of
the responsibility for decision risks because more decisions would be made
by patients. 4 Sharing authority with patients could, therefore, be both a
psychic and a legal relief.
Greater clarity in identifying and protecting an interest in choice might
337. Increasing the patient's direct financial responsibility for chosen medical services would also
decrease demand. Thus under a regime of increased individual choice, demand could be curtailed by
requiring patients who seek levels of care that exceed standards of competent practice to pay a higher
share of the cost of such care. Such a scheme might also, however, require a reciprocal program of
rebate to those who choose less care than professional norms would recommend.
338. See, e.g., Bartling v. Superior Ct., 163 Cal. App. 3d 186, 209 Cal. Rptr. 220 (1984) (competent adult wished to refuse medical care and be allowed to die); Watchorn, supra note 233 (discussing
midwifery as desirable and less expensive alternative to hospital birth).
339. Nor would increased protection of choice necessarily promote the practice of defensive
medicine. Choice protection does not change standards for what the doctor should know or recommend; it affects only what the doctor should disclose about what she knows. In any case, malpractice
suits may be a more limited factor in stimulating defensive medical practice than has been thought.
See Epstein, supra note 17, at 107 n.43.
340. Press, supra note 323, at 54 (relationship factors more important than harm per se in filing
malpractice suits); Enright, supra note 10, at 90.
341. Press observes that the myth of medical perfection creates a situation in which "the question
of who is in charge is easily converted into a question about who is responsible." Press, supra note
323, at 59. But see Adams & Zuckerman, supra note 35, at 484-85 (informed consent is significantly
associated with higher annual rate of malpractice claims after 1976).
HeinOnline -- 95 Yale L.J. 296 1985-1986
Protecting Patient Choice
also reduce unpredictability and irrationality in current malpractice recoveries. Because they receive ambiguous signals from the legal system, doctors have found it difficult to discern meaningful guidelines for conduct. 2
Moreover, growing public anger regarding medical paternalism may
sometimes cause juries and even judges to render decisions that are anomalous under established principles, because they feel that patient choice is
being inadequately protected by present doctrines. A clear and decisive'
mandate to disclose information for purposes of patient choice might actually improve predictability and reduce litigation.
Some have suggested that, in dealing with medical accidents generally,
no-fault systems are preferable to fault-based analysis under traditional
tort approaches.3 43 No-fault schemes assume the central issue to be compensation for an irreducible quantum of medical accidents. Such a rationale has force in the context of malpractice but is unpersuasive where the
interest is patient autonomy. There is nothing inevitable about the allocation of decisionmaking authority between patients and doctors. 3 4 Incentives aimed at furthering traditional tort goals of conduct guidance and
deterrence are especially appropriate here. Indeed, the whole point of
changed standards would be to avoid the very harms occasioned by invasion of choice by altering perceptions both of what is culpable and of what
is appropriate with regard to decisionmaking roles." 5
6.
Effects on the Doctor-PatientRelationship
Realignment of authority in the doctor-patient relationship raises fears
about the quality of that relationship.' 6 The spectre of doctor turned
puppet is not attractive. Nor are these simply self-serving fears of doctors
in renwishing to retain their uncontrolled authority. There are dangers
7
dering either party to a caretaking relationship choiceless.4
342. 2 MAKING DECISIONS, supra note 2, at 25 (only 32% of physicians surveyed agree that the
legal requirements of obtaining informed consent are clear and explicit).
343. See sources collected in Meisel, Expansion, supra note 6, at 143-51; Note, Comparative
Approaches to Liabilityfor Medical Maloccurrences,84 YALE L.J. 1141 (1975). Although Epstein
argues that consensual principles are generally preferable to strict liability in resolving medical disputes, he undermines his own theory in the context of informed consent by adopting paternalistic
medical custom rather than patient expectation as the appropriate measure of disclosure wherever
there is no explicit agreement to the contrary. Epstein, supra note 17, at 102-07, 120-28.
344. Professor Meisel comments on the conflicting functions of medical accident compensation and
informed consent, noting that "[i]t would be ironic if the informed-consent doctrine, which spawned
strict liability for medical accidents, were to contain the seeds of its own destruction .... " Meisel,
Expansion, supra note 6, at 151.
345. If the resistance to fault analysis stems from concern about excessive judgments against individual defendants, other forms of control, such as the proportional recovery system proposed here,
would be preferable solutions.
346. Fears about the doctor-patient relationship being disrupted by greater patient involvement
are mentioned in J. ITz, supra note 7, at 27; Miller, supra note 35, at 2100.
347. See generally R. BuRT, supra note 5 (discussing dynamics of doctor-patient relationship and
HeinOnline -- 95 Yale L.J. 297 1985-1986
The Yale Law Journal
Vol. 95: 219, 1985
A broadened obligation to disclose would not make the doctor an impersonal purveyor of technical information. She would remain professionally
and personally responsible for recommending and implementing decisions.
Nor could she hide behind disclosure of unassimilated information.
Rather, she would retain the role of responsible advisor. Accountability
for the advising function would have to be carefully monitored, lest the
doctor abandon it under the guise of deferring to patient choice.
Patients could not compel a doctor to do something to which the doctor
objects as a matter of personal or professional ethics or competence. 348
Given the varied views of doctors, however, patients should usually be
able to find someone who has no personal objection to implementing the
choice the patient wishes to make.4 9
CONCLUSION
Medical decisionmaking involves the interwoven, overlapping and often
competing claims of personal autonomy and professional competence. The
challenge of regulating medical decisionmaking is to allocate the proper
weight to each of these values. The task is both difficult and important,
for the conflict may be between professionally defined "correct" choices in
matters ultimately involving life and death, and no less a value than selfdetermination.
A new model for the allocation of authority between doctors and patients is needed. Existing legal protection for medical patients' autonomy
is more limited than has been recognized and more deficient than should
be tolerated. Present doctrines falsely equate the protection of autonomy
need for both parties to retain power and responsibility).
348. See, e.g., Doe v. Mundy, 378 F. Supp. 731, 736 (E.D. Wis. 1974) (medical personnel not
required to perform abortions against personal convictions), affd, 514 F.2d 1179 (7th Cir. 1975). A
right to refuse to perform such services is protected by federal legislation in Title II of The National
Research Service Award Act of 1974, Pub. L. No. 93-348, § 214, 88 Stat. 342, 353 (codified at 42
U.S.C. § 300a(b)(2), (c) (1982)). See Ruddick, Doctors' Rights and Work, 4 J. Ma. & PHIL. 192
(1979) (advocating doctors' "bill of rights").
349. See 2 MAKING DECISIONS, supra note 2, at 23 (36% of patients reported that they have
changed doctors because they disagreed with the doctor; 20% report that a doctor has told them to find
another doctor if they did not agree with that doctor's advice). A more complete identification of
doctor and patient preferences and the development of market-oriented shopping for a doctor whose
views are compatible with one's own would be desirable. If no qualified doctor would implement the
patient's choice, a difficult conflict occurs. See Bouvia v. Riverside County, No. 159780 (Cal. Super.
Ct. Dec. 16, 1983). A doctor's autonomy, like a patient's, deserves respect. Common law traditions
about omission and commission suggest that it may be worse to force a doctor to act against her values
than to deprive the patient of the opportunity to effectuate a desired course of action. The logic of the
market also suggests that if no one wishes to provide a service, the patient must go without. Yet the
tradition of medical paternalism is so strong that deference to such principles would invite the same
weakened protection of patient choice that now exists. Furthermore, access to medical care is widely
considered to be one of the most fundamental of necessities. Therefore, in this scenario the presumption of control should be with the patient, and the doctor should carry the burden of showing that
there is an unacceptable personal compromise.
298
HeinOnline -- 95 Yale L.J. 298 1985-1986
Protecting Patient Choice
with control over bodily contact. These doctrines also submerge analysis of
the interest in autonomy within the related but divergent framework of
redress for professional incompetence. Although courts sometimes provide
greater vindication for patient choice where there is a conflict of interest
on the part of the doctor or heightened electiveness on the part of the
patient, these exceptions are unpredictable and inadequately generalized.
Protection of patient autonomy remains derivative rather than direct, episodic rather than systematic. As a result, significant harms to patients'
interest in choice go unredressed.
The subtlety of power-sharing in an ideal relationship between doctor
and patient must be acknowledged.35 0 Even patients who are clearly competent to make decisions will suffer confusion and ambivalence. They will
need guidance and support of professionals and loved ones. Moreover,
professionals are to a significant degree motivated by caring for others; we
need them to continue to be. Respect for patients' autonomy should not
cause doctors either to abandon compassion or to shed their responsibility
for advising and caring for patients. But medical decisions depend upon
moral values, economic considerations and risk preferences, as well as on
medical expertise. Because health care decisions affect the patient more
directly than anyone else, the patient's choices, educated but not preempted by the doctor's expertise, should be controlling.
The law is not the only relevant tool for achieving such a relationship
between doctor and patient. But ultimately the law is about line-drawing,
and some basic division of authority is essential both for purposes of
norm-setting and of dispute resolution. The fact that practice, time and
complexity will embroider nuance and qualification upon the basic structure does not alter the need for such a framework. Patient autonomy
should be recognized and protected as a distinct legal interest.
350. For a thoughtful discussion of the difficulties and limitations of sharp role delineation both
for doctors and patients, see R. BuRT, supra note 5.
HeinOnline -- 95 Yale L.J. 299 1985-1986
Keep reading this paper — and 50 million others — with a free Academia account
Used by leading Academics
Linda MacDonald Glenn
University of California, Santa Cruz
Petros Bouras-Vallianatos
National & Kapodistrian University of Athens
Samia Hurst
Université de Genève
Sergio Rego
Fundação Oswaldo Cruz
