See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/8471669

The Bipolar Affective Disorder Dimension Scale (BADDS) - A dimensional scale

for rating lifetime psychopathology in Bipolar spectrum disorders

Article  in  BMC Psychiatry · February 2004

DOI: 10.1186/1471-244X-4-19 · Source: PubMed

CITATIONS READS
82 627

4 authors, including:

Ian Richard Jones Lisa Jones

Cardiff University University of Melbourne
497 PUBLICATIONS   32,470 CITATIONS    125 PUBLICATIONS   16,053 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Bipolar Education Programme Cymru View project

Bipolar Disorder Research Network View project

All content following this page was uploaded by Ian Richard Jones on 28 May 2014.

The user has requested enhancement of the downloaded file.

BMC Psychiatry BioMed Central

Research article Open Access

The Bipolar Affective Disorder Dimension Scale (BADDS) – a
dimensional scale for rating lifetime psychopathology in Bipolar
spectrum disorders
Nick Craddock*1,2, Ian Jones1,2, George Kirov1 and Lisa Jones1,2

Address: 1Department of Psychological Medicine, University of Wales College of Medicine, Heath Park, Cardiff CF14 4XN, UK and 2Department
of Psychiatry, University of Birmingham, Queen Elizabeth Psychiatric Hospital, Edgbaston, Birmingham B15 2QZ, UK
Email: Nick Craddock* - craddockn@cardiff.ac.uk; Ian Jones - JonesIR1@cardiff.ac.uk; George Kirov - kirov@cardiff.ac.uk;
Lisa Jones - l.a.jones@bham.ac.uk
* Corresponding author

Published: 05 July 2004 Received: 23 September 2003

Accepted: 05 July 2004
BMC Psychiatry 2004, 4:19 doi:10.1186/1471-244X-4-19
This article is available from: http://www.biomedcentral.com/1471-244X/4/19
© 2004 Craddock et al; licensee BioMed Central Ltd. This is an Open Access article: verbatim copying and redistribution of this article are permitted in all
media for any purpose, provided this notice is preserved along with the article's original URL.

Abstract
Background: Current operational diagnostic systems have substantial limitations for lifetime diagnostic
classification of bipolar spectrum disorders. Issues include: (1) It is difficult to operationalize the integration
of diverse episodes of psychopathology, (2) Hierarchies lead to loss of information, (3) Boundaries
between diagnostic categories are often arbitrary, (4) Boundaries between categories usually require a
major element of subjective interpretation, (5) Available diagnostic categories are relatively unhelpful in
distinguishing severity, (6) "Not Otherwise Specified (NOS)" categories are highly heterogeneous, (7)
Subclinical cases are not accommodated usefully within the current diagnostic categories. This latter
limitation is particularly pertinent in the context of the increasing evidence for the existence of a broader
bipolar spectrum than has been acknowledged within existing classifications.
Method: We have developed a numerical rating system, the Bipolar Affective Disorder Dimension Scale,
BADDS, that can be used as an adjunct to conventional best-estimate lifetime diagnostic procedures. The
scale definitions were informed by (a) the current concepts of mood syndrome recognized within DSMIV
and ICD10, (b) the literature regarding severity of episodes, and (c) our own clinical experience. We
undertook an iterative process in which we initially agreed scale definitions, piloted their use on sets of
cases and made modifications to improve utility and reliability.
Results: BADDS has four dimensions, each rated as an integer on a 0 – 100 scale, that measure four key
domains of lifetime psychopathology: Mania (M), Depression (D), Psychosis (P) and Incongruence (I). In
our experience it is easy to learn, straightforward to use, has excellent inter-rater reliability and retains
the key information required to make diagnoses according to DSMIV and ICD10.
Conclusions: Use of BADDS as an adjunct to conventional categorical diagnosis provides a richer
description of lifetime psychopathology that (a) can accommodate sub-clinical features, (b) discriminate
between illness severity amongst individuals within a single conventional diagnostic category, and (c)
demonstrate the similarity between the illness experience of individuals who have been classified into
different disease categories but whose illnesses both fall near the boundaries between the two categories.
BADDS may be useful for researchers and clinicians who are interested in description and classification of
lifetime psychopathology of individuals with disorders lying on the bipolar spectrum.

Page 1 of 10
(page number not for citation purposes)
BMC Psychiatry 2004, 4:19 http://www.biomedcentral.com/1471-244X/4/19

Background which certain symptoms "trump" others. For example, an

During the course of our family-genetic studies of Bipolar individual can have a diagnosis of DSMIV Schizophrenia
Disorder we became aware of the need for a relatively sim- despite having had more episodes of mania during his or
ple dimensional rating scheme that can be used to provide her lifetime than another individual with a diagnosis of
summary measures of several key areas of lifetime psycho- Bipolar I Disorder. For those interested in Bipolarity, this
pathology relevant to characterization of individuals with results in a serious loss of important information.
Bipolar spectrum illness.
3) Boundaries between diagnostic categories are often arbitrary
The operational diagnostic systems, such as RDC [1], – Although there is usually a plausible evidence base and/
DSMIV [2] and ICD10 [3], that have been developed over or conceptual basis to support the separation into distinct
the last 30 years are widely used by clinicians and diagnostic categories, the criteria used to define the
researchers and have been an important methodological boundaries between categories is almost always arbitrary.
advance over earlier, non-structured approaches [4]. For example, the level of impairment defines the bound-
Although open to a variety of criticisms and unlikely to ary between Bipolar I and II Disorders – it is most implau-
map directly onto the pathophysiology of the disorders, sible that any specific level of impairment could neatly
the operational approach is relatively simple, provides carve the boundary between distinct disorders. Similarly,
acceptable levels of reliability and is useful for communi- in DSMIV the boundary between Bipolar I Disorder and
cation and decision making regarding management, Schizoaffective Disorder, Bipolar Type is defined by the
research and service provision. In most cases the catego- precise timing of occurrence of psychotic symptoms out-
ries defined are informed by a broad range of research side of an affective episode. Table 1 lists some key diag-
data and are revised at regular intervals to take account of nostic boundaries relating to bipolar spectrum disorders
new findings and concepts as they emerge. and the criteria used in making diagnostic decisions at
these boundaries.
However, for clinicians and researchers, such as ourselves,
interested in disorders lying within the Bipolar spectrum 4) Boundaries between categories usually require a major ele-
there are several problems in using the current systems for ment of subjective interpretation – Most of the key bounda-
lifetime diagnosis. These include: ries for diagnoses within the Bipolar spectrum require
judgements about severity and/or the balance of symp-
1) It is difficult to operationalize the integration of diverse epi- tomatology. This substantial subjective element reduces
sodes of psychopathology – The operational systems perform reliability and, as commonsense suggests and everyone
best for categorizing a discrete, well-delineated single epi- who has ever participated in formal reliability exercises
sode of psychopathology – for example, there are clear-cut knows, cases lying near diagnostic boundaries contribute
criteria to define episodes of mania and major depression. most of the diagnostic disagreements.
However, a lifetime diagnosis requires integration of the
lifetime experience of psychopathology and the criteria 5) Available diagnostic categories are relatively unhelpful in
used are much less easy to operationalize, typically requir- distinguishing severity – No distinction is made between an
ing judgements about the balance between different types individual that just meets the threshold for a specific cat-
of episode [4]. egory and another individual that has had multiple severe
episodes.
2) Hierarchies lead to loss of information – The existence of
explicit, or implicit, hierarchies creates the situation in

Table 1: Some key boundaries in Bipolar Spectrum Disorders. The table lists some of the important diagnostic boundaries in bipolar
spectrum disorders and the criteria used in making diagnostic decisions at these boundaries.

Boundary Criteria on which decision based

Major Depression (with Sub-clinical hypomanias) v. Bipolar II Disorder Number and duration of hypomanic-like symptoms
Bipolar I v. Bipolar II Disorder Impairment and duration
Bipolar I v. Schizoaffective Disorder, Bipolar Type Occurrence and timing of psychotic symptoms
Schizophrenia v. Schizoaffective Disorder, Bipolar Type Balance of psychotic and affective symptoms

Page 2 of 10
(page number not for citation purposes)
BMC Psychiatry 2004, 4:19 http://www.biomedcentral.com/1471-244X/4/19

6) Subclinical cases are not accommodated usefully within the tions for diagnosis of individuals within the Bipolar spec-
current diagnostic categories – An individual who has a trum, the following key issues are considered within the
DSMIV Diagnosis of Major Depressive Disorder may have decision-making process:
experienced multiple mild sub-hypomanic episodes.
Indeed, it is being increasingly recognized that the bipolar 1) The presence and severity of manic syndromes.
spectrum extends well beyond the traditional Bipolar I
and II categories and includes many individuals with for- 2) The presence and severity of depressive syndromes.
mal diagnoses of unipolar major depression under cur-
rent operational systems [5,6]. For those interested in 3) The presence of psychotic symptoms and the balance of
bipolarity this is wasteful of information. mood and psychotic symptomatology.

7) "Not Otherwise Specified (NOS)" categories are highly het- 4) The mood congruence of psychotic symptoms and the
erogeneous – Because of the need for a catch-all category temporal relationship between affective and psychotic
that can accommodate the set of cases not fitting the other symptomatology.
operational criteria, the NOS categories in operational
systems may include a wide range of types of case ranging In order to capture information relating to the 4 key issues
between the mild and the severe. Thus, such categories above, we, therefore, chose to use 4 dimensions, one for
provide little information about the lifetime psychopa- each issue. Each dimension was set up to provide an
thology of individuals having the diagnosis. ordered (not necessarily linear) measure of the relevant
lifetime experience of psychopathology for the individual
Dimensional classifications offer an alternative to the such that those scoring higher on the scale would have
conventional categorical approach and have the potential experienced more clinically important and convincing
to address many of the issues listed above [7]. Dimen- psychopathology – typically a mix of severity and fre-
sional classifications are not a new idea but have not in quency/duration. Ranges and anchor points in the scales
the past been widely adopted by either the clinical or were initially decided after discussion by the senior inves-
research community. Some of the disadvantages, which tigators and informed by (a) the current concepts of sever-
have impeded widespread use in clinical and research set- ity and type of mood syndrome recognized within DSMIV
tings in psychiatry, stem from their relative complexity – and ICD10, (b) the literature regarding severity of epi-
leading to difficulty in use and interpretation in areas such sodes [8-10], and (c) our own experience in clinical work
as communication and decisions regarding management and research with patients with bipolar spectrum disor-
and services. ders. We undertook an iterative process in which we ini-
tially agreed scale definitions and rating guidelines,
No suitable dimensional instrument was already available piloted their use on sets of cases and modified the scale
for us to use within our own research on bipolar spectrum and guidelines to improve utility and reliability. The scale
disorders. During the course of our ongoing family-based – the Bipolar Affective Disorders Dimension Scale,
clinical research projects that involved assessment and BADDS – has been under development by our group since
classification of the lifetime experience of psychopathol- 1996 and has gone through several iterations. We describe
ogy of Bipolar probands and their relatives we, therefore, the most recent iteration, Version 3.0 which has been used
developed and piloted a simple dimensional rating by our group and collaborators since 1999. All individuals
scheme that was informed by the beneficial aspects of cur- who were assessed with BADDS as part of the diagnostic
rent operational categorical systems but addressed several assessment provided written informed consent to partici-
of the limitations inherent in the use of discrete categories pate in family-genetic studies of mood disorder and our
and provides a richer description of each individual's life- protocols received approval from relevant ethical review
time experience of psychopathology. We describe the committees.
development, structure and characteristics of this system
within the current paper. Results
Basic structure of BADDS
Methods BADDS comprises four dimensions that provide a quanti-
In developing a dimensional scheme our aim was to use a tative measure of lifetime experience of psychopathology
small number of numerical measures that would usefully in each of four domains: Manic-like episodes (the Mania
extend the existing diagnostic schemes – specifically, to dimension, M), Depression-like episodes (the Depression
retain the key information required to make diagnoses dimension, D), Psychotic symptomatology (the Psychosis
whilst maximizing the richness of the additional descrip- dimension, P) and the relationship (in both congruence
tive information and minimizing the problems inherent of content and in timing) between psychotic features (if
in the categorical approach. In using the current classifica- present) and mood episodes (the Incongruence dimen-

Page 3 of 10
(page number not for citation purposes)
BMC Psychiatry 2004, 4:19 http://www.biomedcentral.com/1471-244X/4/19

sion, I) (see supporting material, appendix A for the hypomanic episodes). An individual who has experienced
BADDS rating guidelines). Each dimension provides a 2 episodes of mania (none incapacitating) and 10 epi-
composite measure that takes account of both severity sodes of near-hypomania would receive a rating of M = 62
and frequency of relevant psychopathology and is rated or 63 depending upon the judgement of the rater as to the
using integers in the range 0 to 100, inclusive. As with importance of the near-hypomanic episodes (the range is
conventional lifetime best-estimate categorical diagnosis, 60–79; the worst episode of mania provides a starting
the dimensional ratings are made on the basis of all avail- score of 60 to which is added 2 points for the second
able information – which typically would include semi- manic episode and up to one point for the 10 sub-
structured lifetime psychiatric interview and review of psy- hypomanic episodes).
chiatric case notes. The criteria used for bipolar spectrum
diagnoses is similar within DSMIV and ICD10. We have Depression dimension, D
used ICD10 as the primary source for episode definitions The principles for this dimension follow closely those for
because it provides a clearer differentiation of severity for the M dimension. The severity of the lifetime worst (ie.
depressive episodes. Basic background information and most severe) episode of depression spectrum psychopa-
general rating guidelines are provided in pages 1 and 2 of thology identifies a range of scores on the D dimension to
the BADDS rating guidelines (see Appendix A). The spe- be considered (see table 3 and the rating guidelines for the
cific characteristics of each dimension are described scale: Appendix A). The lifetime "amount" of depression
below. spectrum psychopathology experienced then determines
the score within the range according to clear guidelines
Mania dimension, M that attach weight to the number and severity of episodes
The severity of the lifetime worst (ie. most severe) episode and allows sufficient flexibility that ratings can take
of manic spectrum psychopathology identifies a range of account of other factors where appropriate (such as length
scores on the M dimension to be considered (see table 2 of episodes).
and the rating guidelines for the scale: Appendix A). The
lifetime "amount" of manic spectrum psychopathology Once the range for rating has been specified by consider-
experienced then determines the score within the range ing the severity of the worst episode, the score is deter-
according to clear guidelines that attach weight to the mined by adding 1 point for each additional episode of
number and severity of episodes but allow sufficient flex- equal severity for the mild and moderate depression
ibility that ratings can take account of other factors where ranges (each of which spans 10 points) and 2 points for
appropriate (such as length of episodes). each additional episode of equal severity for the other
ranges (each of which spans 20 points). Thus, an individ-
Once the range for rating has been decided by considering ual who has experienced 11 or more episodes of incapac-
the severity of the worst episode, the score is determined itating depression would be rated D = 100 (the range is
by starting at the lowest score in the range and adding 2 80–100 with 20 points being added to the initial 80
points for each additional episode of equal severity, up to, because there have been 10 or more incapacitating epi-
but not exceeding, the maximum score in the range. Thus, sodes over and above the worst episode that identified the
an individual who has experienced 7 episodes of incapac- range). An individual who has experienced 3 episodes of
itating mania would be rated M = 92 (the range is 80–100 moderate depression would be rated as D = 52 (the range
with 12 points being added to the initial 80 because there is 50–59 with 2 points being added to the initial 50
have been 6 incapacitating episodes over and above the because there have been 2 episodes over and above the
worst episode that identified the range). Similarly an indi- worst episode that identified the range). As with the M
vidual who has experienced 3 episodes of hypomania dimension, for individuals who have also experienced
would be rated as M = 44. An individual who has experi- episodes of lower severity than the worst ever episode,
enced 50 near-hypomanic episodes (ie. episodes that points can be added but with a substantially lower weight-
closely approach, but do not meet, criteria for hypoma- ing than for additional episodes at the same severity.
nia) would be rated as M = 39. For individuals who (as is
common) have also experienced episodes of lower sever- Psychosis dimension, P
ity than the worst ever episode, points can be added – but This dimension is concerned with lifetime occurrence of
with a substantially lower weighting than for additional psychotic and near-psychotic features. It provides a meas-
episodes at the same severity. Thus, an individual who has ure of the proportion of functional psychotic illness in
experienced 2 episodes of mania (none incapacitating) which psychotic symptoms (delusions, hallucinations,
and 10 episodes of hypomania would receive a rating of positive formal thought disorder, catatonia or grossly dis-
M = 67 (the range is 60–79; the worst episode of mania organized behaviour) have been present. The rating takes
provides a starting score of 60 to which is added 2 points account of both the number and duration of episodes
for the second manic episode and 5 points for the 10 with and without psychotic features (see table 4 and the

Page 4 of 10
(page number not for citation purposes)
BMC Psychiatry 2004, 4:19 http://www.biomedcentral.com/1471-244X/4/19

Table 2: Outline of Mania dimension scale. Table shows key points and ranges on the M dimensions together with the criteria defining
the ranges. More details including explicit guidelines for ratings can be found in page 3 of the BADDS rating guidelines (Appendix A).

Range on M dimension Criterion defining range

0 No evidence of manic features during lifetime

1–19 Elation or irritability and 1+ associated manic symptoms for a distinct period
20–39 Elation or irritability and 3+ associated manic symptoms for at least 1 day
40–59 At least one hypomanic episode
60–79 At least one manic episode but never experienced a manic episode meeting criteria for "incapacitating
mania" (as defined in BADDS guidelines)
80–100 At least one manic episode meeting criteria for "incapacitating mania" (as defined in BADDS guidelines)

Table 3: Outline of Depression dimension scale. Table shows key points and ranges on the D dimensions together with the criteria
defining the ranges. More details including explicit guidelines for ratings can be found in page 4 of the BADDS rating guidelines
(Appendix A).

Range on D dimension Criterion defining range

0 No evidence of depressive features during lifetime

1 – 19 At least one sub-minor depression episode (as defined in BADDS guidelines)
20–39 At least one minor depression episode (as defined in BADDS guidelines)
40–49 At least one major depression episode of mild severity (as defined in BADDS guidelines)
50–59 At least one major depression episode of moderate severity (as defined in BADDS guidelines)
60–79 At least one major depression episode of severe severity (as defined in BADDS guidelines) but never
experienced incapacitating depressive episode (as defined in BADDS guidelines).
80–100 At least one incapacitating major depression episode (as defined in BADDS guidelines)

rating guidelines for the scale: Appendix A). Near-psy- non-prominent would be rated P = 20. An individual who
chotic schizotypal features (specifically the following has never experienced clear-cut psychotic features but has
DSMIV schizotypal items: ideas of reference; odd beliefs had frequent near-psychotic features would be rated P = 9,
or magical thinking that influences behaviour and is and such a person who has had only occasional near-psy-
inconsistent with sub-cultural norms; unusual perceptual chotic features would be rated P = 2.
experiences including bodily illusions; odd thinking and
speech; suspiciousness or paranoid ideation; behaviour or Incongruence dimension, I
appearance that is odd eccentric or peculiar), in the This dimension is the most complex and provides lifetime
absence of clear-cut psychotic features, can be rated in the information about the relationship between psychotic
lowest range of the dimension if there have been no clear- and affective psychopathology, specifically in three areas:
cut psychotic features. (a) the mood congruence of any psychotic features that
occur, (b) the occurrence of specific symptoms that have
Rating of 0 and 1 have specific definitions. The severity special diagnostic weight in the diagnosis of schizophre-
and amount of relevant psychopathology are rated within nia and schizoaffective disorder within current opera-
the ranges 2–9 (for near-psychotic features), 10–20 (for tional classifications (which we denote for convenience,
relatively brief single or multiple psychotic features) and the "S set": thought echo, insertion, withdrawal or broad-
21–100 (for individuals having multiple episodes where casting; passivity experiences; hallucinatory voices giving
psychotic symptoms are a prominent feature). Thus, an running commentary, discussing subject in third person
individual for whom psychotic features have been present or originating in some part of the body; bizarre delusions;
and prominent in each episode of illness would be rated catatonia), and (c) the temporal relationship between
P = 100 (whether this is a single episode or 20 episodes). mood and psychotic psychopathology (see table 5 and the
An individual for whom psychotic features have been rating guidelines for the scale: Appendix A). The dimen-
present and prominent in one third of episodes of illness sion is rated only if the P dimension has been rated at P >
would be rated P = 33. An individual for whom multiple 9 (ie. occurrence of definite psychotic symptoms at some
psychotic features have been present but were brief and time during lifetime); otherwise, it is left blank.

Page 5 of 10
(page number not for citation purposes)
BMC Psychiatry 2004, 4:19 http://www.biomedcentral.com/1471-244X/4/19

Table 4: Outline of Psychosis dimension scale. Table shows key points and ranges on the P dimension together with the criteria defining
the ranges. More details including explicit guidelines for ratings can be found on page 5 of the BADDS rating guidelines (Appendix A).

Range on P dimension Criterion defining range

0 No evidence of psychotic or near-psychotic schizotypal features

1 Uncertainty about presence of psychotic-spectrum symptoms (ie. Suspected but not definite)
2–9 Presence of near-psychotic schizotypal features but never any clear-cut psychotic symptoms
10–20 Brief, clear-cut psychotic symptoms that are not a prominent feature of the illness
21 – 100 Clear-cut psychotic symptoms that are a prominent feature of one or more episodes of illness

Table 5: Outline of Incongruence dimension scale. Table shows key points and ranges on the I dimension together with the criteria
defining the ranges. The S set of psychotic symptoms are those recognized as having special weight in the diagnosis of schizophrenia
and schizoaffective disorder (thought echo, insertion, withdrawal or broadcasting; passivity experiences; hallucinatory voices giving
running commentary, discussing subject in third person or originating in some part of the body; bizarre delusions; catatonia). More
details including explicit guidelines for ratings can be found on page 6 of the BADDS rating guidelines (Appendix A).

Range on I dimension Criterion defining range

0–40 Psychotic symptoms occur only during affective episodes and do not include any of the S set.
Rating 0 – virtually completely mood congruent.
Rating 20 – approximate balance between mood congruent and incongruent.
Rating 40 – virtually completely mood incongruent
43 Psychotic symptoms occur only during affective episodes and include one or more of the S set which have not
definitely been present for 2 weeks.
47 Psychotic symptoms occur only during affective episodes and include one or more of the S set which have definitely
been present for 2 weeks.
50–59 Psychotic symptoms probably present for at least 2 weeks either side of an affective episode.
Rating 50 – on at least one occasion.
Ratings of 51–59 used to reflect recurrence and/or certainty.
60–100 Psychotic symptoms definitely present for at least 2 weeks either side of an affective episode.
Rating 60 – on at least one occasion.
Rating 80 – on many occasions.
Rating 100 – Psychotic symptoms predominate illness and occur chronically outside (or in absence of) affective
episodes.

Thus, an individual who has psychotic features only dur- have found it to be user-friendly, simple to learn and
ing affective episodes and for whom the psychotic symp- straightforward to incorporate within the usual lifetime
toms are mainly, but not exclusively mood congruent, best-estimate consensus procedures.
would be rated I = 10. An individual who has psychotic
features only during affective episodes and for whom the Reliability studies of BADDS
psychotic symptoms are exclusively mood incongruent, In order to examine the reliability of BADDS within the
would be rated I = 40. An individual who has psychotic context of our typical spectrum of cases we undertook a
features including passivity lasting at least 2 weeks at reliability study using written case vignettes containing
some time during the illness and only during affective epi- interview and case notes data for 20 cases selected as a rep-
sodes would be rated I = 47. An individual who has had resentative sample with a mix of diagnoses from our
psychotic features for at least 2 weeks outside of an affec- ongoing studies of mood disorder. Nine raters with expe-
tive episode on several occasions would be rated I = 70. rience of the conventional lifetime diagnostic process par-
ticipated (2 psychiatrists and 7 psychologists) and made
Utility of BADDS ratings on the BADDS dimensions. A meeting of all raters
Our group has substantial experience in use of BADDS was then held to agree a consensus for each rating that was
within the context of lifetime psychiatric assessment in then used as the gold standard against which agreements
family-genetic studies of Bipolar Disorder, unipolar were measured. Mean agreements, measured by intraclass
depression and puerperal psychosis. It has been used as correlations, were excellent for all dimensions (M: 0.96;
part of the diagnostic procedure in over 1100 patients. D: 0.90; P: 0.86; I: 0.89).
Within our group it has been used by 16 researchers
including psychiatrists, psychologists and sociologists. We

Page 6 of 10
(page number not for citation purposes)
BMC Psychiatry 2004, 4:19 http://www.biomedcentral.com/1471-244X/4/19

A second exercise was undertaken in order to examine the ple diagnostic process. In particular it avoids hierarchical
performance of BADDS with diagnostically difficult cases. loss of information; it retains a measure of severity; it
A different, non-overlapping, set of seven raters (4 psychi- accommodates sub-clinical cases. We have demonstrated
atrists and 3 psychologists) with experience of the conven- that it is straightforward to learn and incorporate within
tional lifetime diagnostic process undertook a reliability the usual lifetime diagnostic procedures for use by a range
study using written case vignettes containing interview of researchers including those from psychiatry and psy-
and case notes data for 20 cases selected from those chology backgrounds. We have demonstrated excellent
recruited for our on going studies of functional psychosis levels of inter-rater agreement even with diagnostically
over-represented with diagnostically challenging cases challenging sets of cases. Further, we have shown that the
representing a mix of diagnoses. Raters made lifetime best key information required for correct diagnostic decisions
estimate diagnoses according to DSMIV and ICD10 and according to DSMIV and ICD10 is retained within the
ratings on the BADDS dimensions. A meeting of all raters dimensional ratings.
was then held to agree a consensus for each rating that was
then used as the gold standard against which agreements Our group and our collaborators have extensive experi-
were measured. DSMIV consensus diagnoses were Bipolar ence of use of BADDS as an adjunct to conventional oper-
I Disorder – 6; Bipolar II Disorder – 3; Bipolar Disorder ational diagnosis and it has been part of our standard
Not Otherwise Specified – 3; Schizoaffective Disorder, assessment approach for over 5 years. We have found that
Bipolar Type – 2; Recurrent major Depression – 2; Single it is straightforward to use and adds little to the time taken
episode major depression – 1; Depression Not Otherwise to complete the consensus diagnostic process.
Specified – 2; Schizophrenia – 1. Mean categorical agree-
ment across raters, measured by Cohen's κ was 0.68 for For researchers, such as ourselves, wishing to establish a
DSMIV and 0.62 for ICD10. Mean agreements on the measure of "caseness" BADDS can easily be used to define
BADDS dimensions, measured by intraclass correlations thresholds – for example, a study of mania might require
were: M: 0.91; D: 0.86; P: 0.96; I: 0.78. that cases be included only for M > 64. This would allow
inclusion of all cases with the equivalent of 3 of more epi-
Prediction of DSMIV and ICD10 diagnoses from dimension sodes of mania, irrespective of diagnosis. In a study of psy-
scores chotic Bipolar spectrum illness it might be important to
In order to test the assumption that most of the categorical distinguish between cases in which psychotic features
diagnostic information is preserved within the dimen- were a prominent, recurrent feature of illness (rather than
sions, one rater (NC), experienced in use of BADDS, pre- an occasional relatively minor feature). Such individuals
dicted diagnoses according to DSMIV and ICD10 based could be selected using BADDS as having P > 50, together
soley upon the dimension scores on BADDS for 50 repre- with M > 60. BADDS can also easily be used in conjunc-
sentative cases selected by LJ from amongst those tion with categorical diagnoses for case selection.
recruited for our genetic studies of Bipolar spectrum
mood disorders (both categorical diagnoses and dimen- BADDS was developed within the context of family stud-
sion scores for each case had been made by 2 independent ies and it lends itself to providing a substantially more
raters according to best estimate lifetime procedures on useful description of the milder ("sub-clinical") end of the
the basis of semi-structured interview and psychiatric case Bipolar spectrum which is frequently encountered within
note data). The sample comprised a mix of cases with members of families of probands with full-blown Bipolar
DSMIV categorical diagnoses of Bipolar I Disorder (10), illness. Conventional categorical approaches often lead to
Bipolar II Disorder (10), Bipolar Disorder Not Otherwise unsatisfactory diagnoses such as "Never ill", "Major
specified (10), Schizoaffective Disorder (10) and Recur- Depressive Disorder" or some form of mild "Not Other-
rent Major Depressive Disorder (10). Correct prediction wise Specified" category when it is clear that there is some
was made for both DSMIV and ICD10 for all cases. definite, albeit mild, degree of bipolarity. Within the con-
text of family studies it is extremely wasteful to discard
Discussion such quantitative information about the presence and
We have described a new dimensional rating scheme that extent of bipolar features and BADDS provides a simple
can be used as an adjunct to conventional categorical approach to making simple but efficient use of such data.
diagnosis in order to provide a richer description of some
of the basic features of an individual's lifetime experience Directly related to this issue, there is currently great inter-
of psychopathology relevant to the bipolar spectrum. The est in delineating the breadth and frequency of expression
scheme uses the same data sources as conventional best- of the bipolar illness spectrum in the population. Recent
estimate lifetime diagnosis and is straightforward to use at research, championed by Akiskal and Angst, provides evi-
the same time as the conventional procedure. It retains dence that many cases that have been regarded as being
several key pieces of information that are lost in the sim- "unipolar major depression" actually have subtle (or not

Page 7 of 10
(page number not for citation purposes)
BMC Psychiatry 2004, 4:19 http://www.biomedcentral.com/1471-244X/4/19

so subtle) bipolar features [5,6] and classifications have will inevitably lead to poor dimensional ratings as well as
been suggested that recognize several categories of milder poor categorical diagnoses. It is essential that multiple
bipolarity in addition to the conventional DSMIV catego- data sources are used whenever possible that provide ade-
ries of Bipolar I and II Disorders [eg. [5]]. BADDS provides quate description of an individual's lifetime experience of
the capability to capture information about this milder psychopathology (not just one or two representative epi-
degree of bipolarity – a substantial part of the M dimen- sodes). As for any type of rating, poor data would be
sions (the range 0 – 39) is available for rating sub-clinical expected to affect both the validity and reliability of rat-
hypomanic features. ings. Second, ratings can only reflect what is known of the
lifetime experience of psychopathology up to the time the
The dimensional approach is particularly beneficial for ratings are made. In the light of new episodes of illness
cases close to diagnostic boundaries. As any researcher or scores on the M and D dimensions may increase; those on
clinician knows who has undertaken formal diagnostic the P and I dimensions may increase or decrease. Third,
assignment using operational classifications such cases subjective judgments are required in integrating multiple
may be associated with a substantial investment of time in data sources and matching data to the criteria within the
order to make a finely balanced decision between two (or guidelines. Within the context of our current approaches
occasionally more) discrete diagnostic groups. It is com- to psychiatric classification this is inevitable. Judgements
mon that different raters come down on different sides of must still be made about the range for a rating – this can
the finely balanced decision process leading to a split of be equivalent to making a categorical judgement, except
diagnoses with eventual agreement on a consensus but that the different categories lie contiguous with one
often with further agreement that it is a "difficult case" another on an ordered dimension. Fourth, there are fea-
and that the single category chosen does not quite do jus- tures of Bipolar spectrum illness that BADDS was not
tice to the complexity of the case. In contrast, the dimen- designed to capture – examples include the presence and
sional approach of BADDS provides a scheme which can extent of rapid cycling and the extent of mixed episodes
reflect that different ratings of such cases are relatively (although if all manic episodes are mixed this is denoted
close on the quantitative scale. An example is provided by in BADDS by adding an "m" qualifier to the M dimension
a case considered in the formal reliability exercise in – see rating guidelines in Appendix A). It is possible for
which the subject experienced several severe (but not inca- additional dimensions to be added to capture additional
pacitating) major depressive episodes and also mild recur- features. Fifth, BADDS was not developed for use in the
rent sub-hypomanic episodes. Of the 7 raters, 4 made a general population. It was designed for use in clinical
diagnosis of DSMIV Bipolar Disorder, Not Otherwise populations likely to contain patients with Bipolar spec-
Specified (the consensus) and 3 a diagnosis of Recurrent trum diagnoses. The dimensions have meaning in provid-
major Depressive Disorder. In contrast the dimensional ing an ordered measure of specific domains of
ratings were very similar across all raters (means for those psychopathology. The distributions remain to be tested in
raters making diagnosis of Bipolar Disorder Not Other- non-clinical populations but will certainly not conform to
wise Specified: M 27.3; D 70.3; P 0; I blank; means for normal distribution. Sixth, for the M and D dimensions
those raters making diagnosis of Recurrent Major Depres- there is a ceiling effect in that these dimensions do not
sion: M 23; D 74.7; P 0; I blank). allow discrimination between individuals having more
than 11 episodes of incapacitating mania, or depression,
The primary purpose in developing BADDS was to use it respectively. In practice, however, for the populations of
as an adjunct to better describe some key features of cases patients that we have studied relatively few patients score
and provide a simple mechanism for case selection on the M = 100 or D = 100. Seventh, BADDS is relatively poor at
basis of these features. BADDS has already been used characterizing cases where the majority of episodes are at
within family-based studies to investigate intra-familial a lower level of severity than the most severe.
resemblance for lifetime experience of mania and psycho-
sis [11] as well as investigating the relationship between Our justification for developing BADSS was that no
smoking and psychosis in Bipolar Disorder [12]. We are dimensional scale was already available that adequately
currently using BADDS to explore genotype-phenotype addressed the issues (1) – (7) discussed in the background
correlations within the context of both classical and section. However, several researchers have described
molecular genetic studies of large samples of patients with approaches relevant to dimensional ratings of psychopa-
functional psychosis and mood disorder. thology including bipolar features. Depue has described a
quantitative scale for screening for Bipolar and Unipolar
There are several limitations in use of BADDS, most of disorders within a non-clinical university population
which are common to other lifetime diagnostic proce- [13]. This derived a bipolar and a unipolar dimension
dures. First, and most obvious, is that the ratings are from a modified version of the General behaviour Inven-
entirely dependent upon the quality of the data. Poor data tory [14] and focused on screening for affective psychopa-

Page 8 of 10
(page number not for citation purposes)
BMC Psychiatry 2004, 4:19 http://www.biomedcentral.com/1471-244X/4/19

thology at the milder end of the spectrum. Brockington A further useful approach in characterization of episodic
and colleagues have described a complex procedure for disorders such as Bipolar Disorder is the life chart method
lifetime psychopathological assessment that includes a [23] which provides a visual schematic representation of
detailed interview schedule and case note review (taking 9 illness using a time-line on which is charted key events, ill-
hours per patient) and produces lifetime summary scores ness episodes and treatments during an individual's life.
on 30 scales covering a wide range of psychopathology We find this an invaluable component of our own assess-
[15]. One popular approach to lifetime rating of psycho- ments but it in general it is necessary for quantitative and
pathology for functional psychosis is provided by qualitative information about illness type, frequency and
OPCRIT, a computer-based 92 item checklist that includes severity to abstracted from the life chart for use in research
symptoms over a range of domains including positive, or clinical settings. BADDS clearly does not provide the
negative and disorganized psychotic symptoms, course full richness of individual description of the life chart
variables, depressive symptoms and manic symptoms method but is designed to capture some of the important
[16]. OPCRIT can be used in a variety of ways but was pri- features of an individual's lifetime experience of illness.
marily developed as a diagnostic system. It performs best
for schizophrenia spectrum disorders, although it can be Finally, it is important to emphasize that BADDS is a
used satisfactorily in diagnosis of Bipolar Disorder [17]. dimensional system developed on the basis of existing
However, OPCRIT does not provide a dimensional meas- data about the nosology of bipolar spectrum disorders in
ure of severity or frequency/duration of the domains of order to provide a description of domains recognized as
psychopathology and, in its unmodified form, is much important in classification. This is an entirely distinct
less satisfactory for use with disorders having a predomi- approach to that of researchers who have undertaken fac-
nantly episodic course. However, within these constraints, tor analyses of symptoms during acute episodes of func-
OPCRIT has been used by several groups for investigating tional psychotic illness – generally identifying factors or
factor structures of patient sets with functional psychotic clusters that represent the features of episodes (mania,
illness [eg. [18,19]]. Several groups working on the genet- depression etc) [eg. [24-26]].
ics of psychosis have described dimensional approaches
that focus on the psychotic domains of psychopathology. Conclusions
Maziade et al have examined lifetime ratings of psychotic Current operational diagnostic systems have substantial
symptom dimensions in patients with schizophrenia and limitations for classification of bipolar spectrum disor-
Bipolar Disorder [20]. There was no coverage of mood ders. We have described a relatively simple dimensional
symptoms and assessments were confined to predomi- rating system, BADDS, that can be used as an adjunct to
nant symptoms in acute episodes and predominant symp- conventional best-estimate lifetime diagnostic procedures
toms "between" episodes. Kendler et al [21] have used [27] in order to provide information about four key
clinical judgement to make ratings on a 4 point scale that domains of lifetime psychopathology: Mania (M),
reflected severity and duration for each of 9 symptom and Depression (D), Psychosis (P) and Incongruence (I). In
2 course variables which included depressive symptoms our experience BADDS is easy to learn, straightforward to
and manic symptoms. Levinson and colleagues [22] have use, has excellent inter-rater reliability and retains the key
recently described a lifetime dimension scale for use in information required to make diagnoses according to
psychosis research, the Lifetime Dimensions of Psychosis DSMIV and ICD10. Use of BADDS as an adjunct to con-
Scale (LDPS). This was developed within the context of ventional categorical diagnosis provides a richer descrip-
family-genetic studies of schizophrenia and motivated by tion of lifetime psychopathology that (a) can
several of the same concerns and aims that motivated us accommodate sub-clinical features, (b) discriminate
in developing BADDS. Ratings are made on a 39 item between illness severity amongst individuals within a sin-
scale that reflect severity (on a 5 point scale) and duration gle conventional diagnostic category, and (c) demonstrate
(on a 5 point scale) for lifetime occurrence of a range of the similarity between the illness experience of individu-
psychotic features encompassing positive, bizarre, nega- als who have been classified into different disease catego-
tive and disorganized domains plus depressive and manic ries but whose illnesses both fall near the boundaries
syndromes. As with the approach taken by Maziade and between the two categories. BADDS may be useful for
Kendler, the focus of LDPS is schizophrenia spectrum researchers and clinicians who are interested in descrip-
disorders and a chronic course. There is relatively little tion and classification of lifetime psychopathology of
attention to the milder mood psychopathology, episodic individuals with disorders lying on the bipolar spectrum.
course and to the relationship between mood and psy-
chotic symptomatology. These are all issues of key impor- Competing interests
tance to study of bipolar spectrum illness and are a focus None declared.
of BADDS.

Page 9 of 10
(page number not for citation purposes)
 BMC Psychiatry 2004, 4:19 http://www.biomedcentral.com/1471-244X/4/19

Authors' contributions 14. Depue RA, Klein D: Identification of unipolar and bipolar affec-
tive conditions by the general Behavior Inventory. In Relatives
NC was the overall principal investigator for the work and at risk for mental disorder Edited by: Gershon E, Barrett J. New York:
had primary responsibility for development of BADDS Raven Press; 1988:257-282.
and writing the manuscript. IJ and GK were closely 15. Brockington I, Roper A, Edmunds E, Kaufman C, Meltzer HY: A lon-
gitudinal psychopathological schedule. Psychological Medicine
involved in development and piloting of BADDS and con- 1992, 22:1035-1043.
tributed to writing the manuscript. LJ was involved in 16. McGuffin P, Farmer A, Harvey I: A polydiagnostic application of
supervision of researchers using BADDS, data analysis and operational criteria in studies of psychotic illness. Develop-
ment and reliability of the OPCRIT system. Archives of General
contributed to writing the manuscript. Psychiatry 1991, 48:764-770.
17. Craddock N, Asherson P, Owen MJ, Williams J, McGuffin P, Farmer
AE: Concurrent validity of the OPCRIT diagnostic system:
Additional material comparison of OPCRIT diagnoses with consensus best-esti-
mate lifetime diagnoses. British Journal of Psychiatry 1996,
169:58-63.
Additional File 1 18. Van Os J, Gilvarry C, Bale R, Van Horn E, Tattan T, White I, Murray
A Word file is provided that includes Version 3.0 of the BADDS rating RM: A comparison of the utility of dimensional and categori-
cal representations of psychosis. Psychological Medicine 1999,
guidelines as an appendix. 29:595-606.
Click here for file 19. Cardno AG, Jones LA, Murphy KC, Sanders RD, Asherson P, Owen
[http://www.biomedcentral.com/content/supplementary/1471- MJ, McGuffin P: Dimensions of psychosis in affected sibling
244X-4-19-S1.doc] pairs. Schizophr Bull 1999, 25:841-50.
20. Maziade M, Roy MA, Martinez M, Cliche D, Fournier JP, Garneau Y,
Nicole L, Montgrain N, Dion C, Ponton AM: Negative, Psychoti-
cism, and Disorganized Dimensions in Patients With Famil-
ial Schizophrenia or Bipolar Disorder: Continuity and
Discontinuity Between the Major Psychoses. American Journal of
Acknowledgements Psychiatry 1995, 152:1458-1463.
We are grateful to Fiona McCandless, Alastair Cardno, Michael Gill, Ed 21. Kendler KS, Glazer WM, Morgenstern H: Dimensions of Delu-
O'Mahony, Aiden Corvin, Jess Heron and Emma Robertson for comments sional Experience. American Journal of Psychiatry 1983, 140:466-469.
and contributions to piloting the use of BADDS. Funding for this work was 22. Levinson DF, Mowry BJ, Escamila MA, Faraone SV: The lifetime
dimension of psychosis scale (LDPS): description and iter-
provided by the Wellcome Trust. We are grateful to all individuals who rater reliability. Schizophr Bull 2002, 28:683-95.
have participated in our studies of mood disorder. 23. Denicoff KD, Leverich GS, Nolen WA, Rush AJ, McElroy SL, Keck PE,
Suppes T, Altshuler LL, Kupka R, Frye MA, Hatef J, Brotman MA, Post
References RM: Validation of the prospective NIMH-Life-Chart Method
(NIMH-LCM-p) for longitudinal assessment of bipolar illness.
1. Spitzer RL, Endicott J, Robins E: Research Diagnostic Criteria. Psychological Medicine 2000, 30:1391-7.
Rationale and reliability. Archives of General Psychiatry 1978, 24. Cuesta MJ, Peralta V: Integrating psychopathological dimen-
35:773-782. sions in functional psychoses: a hierarchical approach. Schizo-
2. APA: Diagnostic and statistical manual of mental disorders, (DSM-IV) phrenia Research 2001, 52:215-229.
Fourth edition. Washington, DC: American Psychiatric Association; 25. Cassidy F, Forest K, Murry E, Carroll BJ: A factor analysis of the
1994. signs and symptoms of mania. Arch Gen Psychiatry 1998, 55:27-32.
3. WHO: The ICD10 classification of mental and behavioural disorders. Diag- 26. Kitamura T, Okazaki Y, Fujinawa A, Yoshino M, Kasahara : Symp-
nostic criteria for research Geneva: World Health Organization; 1993. toms of Psychoses, A factor – analytic study. British Journal of
4. Farmer AE, Williams J, Jones I: Phenotypic definitions of psy- Psychiatry 1995, 166:236-240.
chotic illness for molecular genetic research. American Journal 27. Leckman JF, Sholomskas D, Thompson D, Belanger A, Weissman MM:
of Medicial Genetics (Neuropsychiatric Genetics) 1994, 54:365-371. Best estimate lifetime psychiatric diagnosis. A methodolog-
5. Akiskal H, Pinto O: The evolving bipolar spectrum. Prototypes ical study. Archives of General Psychiatry 1982, 39:879-883.
I, II, II, and IV. The Psychiatric Clinics of North America 1999,
22:517-534.
6. Angst J, Gamma A, Benazzi F, Ajdacic V, Eich D, Rossler W: Toward Pre-publication history
a re-definition of sub-threshold bipolarity: epidemiology and The pre-publication history for this paper can be accessed
proposed criteria for bipolar-II, minor bipolar disorders and
hypomania. J Affective Disorders 2003, 73:133-146. here:
7. Kendell RE: Diagnosis and classification of functional
psychosis. Br Med Bull 1987, 43:499-513.
8. Goodwin FK, Jamison KR: Manic_depressive illness New York, Oxford
http://www.biomedcentral.com/1471-244X/4/19/pre
University Press; 1990. pub
9. Endicott NA: Psychophysiological correlates of "bipolarity".
Arch Gen Psychiatry 1989, 17:47-56.
10. Carlson GA, Goodwin FK: The stages of mania. A longitudinal
analysis of the manic episode. Arch Gen Psychiatry 1973, 28:221-8.
11. O'Mahony E, Corvin A, O'Connell R, Comerford C, Larsen B, Jones
R, McCandless F, Kirov G, Cardno AG, Craddock N, Gill M: Sibling
pairs with affective disorders: resemblance of demographic
and clinical features. Psychological Medicine 2002, 32:55-61.
12. Corvin A, O'Mahoney ED, O'Regan M, Comerford C, O'Connell R,
Craddock N, Gill M: Cigarette smoking and psychotic symp-
toms in bipolar affective disorder. British Journal of Psychiatry
2001, 179:35-38.
13. Depue RA, Krauss S, Spoont MR, Arbisi P: General behavior
inventory identification of unipolar and bipolar affective con-
ditions in a nonclinical university population. Journal of Abnornal
Psychology 1989, 98:117-126.

Page 10 of 10
(page number not for citation purposes)
View publication stats