zyxwv zyxw zyxwvuts Families in Society: The Journal of Contemporary Human Services Copyright © 1991 Family Service America zyxwvu zyxw zyxw zyxwvuts zyxw Criteria for Terminating Treatment Anne E. Fortune, Bill Pearlingi, & Cherie D. Rochelle ABSTRACT: What criteria do social workers use to terminate treatment? Sixty-nine practitioners reported on their most recently terminated case. The most important criteria were improved coping in the environment, improved intrapsychicfunctioning, the client’s wish to terminate, meeting initial goals, and changes in therapeutic content. Practitioners rarely considered terminating trearment unless they perceived trearment as successful, although many different success indicators were used. T TERMINATION OF TREATMENT, like death and taxes, is inevitable. But how do client and practitioner know when to end it? This article examines the termination criteria that 69 experienced practitioners deemed important in a recently concluded case. Ideally, treatment ends when it is successful. But how is success defined? What does it look like? Although success usually includes t h e client’s improvement, the areas of improvement may vary from specific behaviors to intrapsychic functioning. Selecting a criterion may be difficult because total success is rarely possible. As a result, many practitioners define “success” as achievement of goals, under the assumption that goals are realistic and attainable. Others look for cues to success or to client readiness to terminate, such as changes in transference and dependency or the client expressing a desire to end treatment. O t h e r approaches to termination avoid defining success. In cost-benefit analysis, practitioners terminate when the costs of continuing treatment outweigh further gain; for example, time and effort may not be worth the minor improvement that may occur by continuing treatment. Other practitioners terminate treatment when no improvement is evident after a certain length of time, arguing that it is unethical to continue ineffective treatment. In time-limited treat- 366 ment, termination comes at the end of an agreedupon period. Other reasons to end treatment include client circumstances that “force” termination, practitioner circumstances such as pregnancy or promotion, and agency policy regarding length of treatment or client eligibility. W h e n should each criterion be invoked? Client age, outcome, treatment structure (such as time limits), practitioner status or experience, and theoretical orientation can affect selection of criteria. For example, family needs and desires are important if the client is a child (Barton & Marshall, 1986). Student practitioners terminate when they leave an agency, whereas inexperienced practitioners sometimes set goals too high and continue treatment too long (Gould, 1978; Greene, 1980). Theoretical orientation influences choice of intrapsychic, interpersonal, or behavioral criteria, whereas time limits or accomplishment of goals are relevant only if discussed at zy Anne E . Fortune is Associate Professor, School of Social Welfare, State University of New York at Albany, Albany, New York. Bill Pearlingi is Staff Therapist, Adolescent Program, Psychiatric Institute of Richmond, Richmond, Virginia. Cherie D. Rochelle is Program Therapist, Solomon Center for Girls, Hallan- dale. Florida. zy zy Terminating Treatment Fortune et al. the beginning of treatment (Kramer, 1982; Levinson, 1977). The criteria practitioners actually use to terminate treatment are rarely investigated. Glover’s (1955) study of 24 psychoanalysts found that most of them used intuition about when to end; onethird gave no criteria for terminating. Firestein ( 1978) examined the termination of eight psychoanalyses without looking at criteria, although he did mention that the therapist rather than the patient could introduce termination successfully. Kramer (1986) found social workers waiting for clients to introduce termination before consider, ing it. Their reasons for agreeing to terminate included intrapsychic improvement, observable external improvement, overall global improvement, cost-benefit analysis indicating that further gain would be marginal, financial cost, and unsatisfactory treatment (Kramer, 1982). Fortune’s ( 1985) study found that social workers frequently cited achievement of goals, improved behavior, and improved intrapsychic functioning for termination of treatment; they rarely cited lack of progress or cues such as missing sessions. Eldridge’s ( 1983) social workers stated that client behavior was a key outcome but they seldom measured it, whereas Richey, Blythe, and Berlin’s (1987) practitioners used outcome measurement least among research-based treatment techniques. Only one study looked at factors that influence choice among termination criteria. Criteria varied according to practitioner experience, case load, and familiarity with short-term treatment. Criteria did not differ by other important factors such as theory base, public or private auspice of the agency, or client characteristics (Fortune, 1985). However, that study did not look at actual cases, which might better explain factors that affect criteria. In the present article, practitioners were interviewed about the termination of recent cases. Practitioner reactions to termination are reported elsewhere ( F o r t u n e , Pearlingi, & Rochelle, in press). This article describes why the cases ended, addressing two areas: (1) the impor, tance of particular criteria in deciding to terminate and ( 2 ) the factors, such as experience or time limits. that influence choice of criteria. Methodology Master’s-level practitioners were recruited through word of mouth and listings in the local telephone book. Each practitioner reported on the individual client who had terminated most recently (but within the past six months) and who had been in treatment for at least 10 sessions. One hundred fifty practitioners in central Virginia were given a questionnaire. Sixty-nine (46%) responded. The nine-page questionnaire was adapted by the junior authors from a questionnaire used previously by the senior author (Fortune, 1985, 1987). Most questions were closed-ended, although a few open-ended questions such as presenting problems and diagnosis were used. Topics included client and practitioner demographic characteristics,case characteristics, criteria, and reactions to termination. This report focuses on the criteria for termination. zyxwvu zyx Practitioners and Clients Practitioners were mostly white, female, with an average of 10 years’ work experience in various public and private agencies. More than half said that a psychosocial treatment approach “very strongly” influenced the case upon which they were reporting. Two-thirds of the clients were female, most were white, and most were adults, although nearly 25% were younger than 21. Major problems included depression (36); family problems (14); other individual problems such as developmental, identity, or trust issues (4); and substance abuse ( 2 ) . Sixty percent suffered some form of affective disorder as diagnosed on the DSM-111-R Axis I (American Psychiatric Association, 1987). Treatment lasted from 2 to 72 months with an average of 13 months and 37 sessions. In most cases, length of treatment was open-ended, with no intentional use of time limits for treatment. Time limits were set in only 10% of cases. In most cases, the practitioner began preparing the client for termination during the last tenth of sessions rather than earlier, but some (7%) did not prepare the client at all. Usually, the client and practitioner raised the issue of ending treatment at more or less the same time; one of seven cases ended due to external circumstan-ces. Cases were usually successful as rated by practitioners, with an average outcome of 1.6, wherein 1 indicated “very successful”and 4 “not at all successful.” To summarize, respondents were experienced practitioners reporting on a variety of clients in successful long-term, open-ended treatment. Most terminations were by agreement, but preparation for termination did not occur until just before treatment ended. 367 zy zyxwvuts Families in Society June 1991 zyxwvut criteria included the client’s wish to terminate and a change in therapeutic content that signaled readiness to end. Rarely important were reaching planned time limits, interference from agency policy, increased transference, decreased transference, and increased dependency. TABLE 1. Criteria for termination: Means and standard deviations. Item M Goals set at or near beginning of treatment were met. 2.22 1.1 Time limit initially planned reached. 3.88 1.2 Client’s behavior or coping in the environment improved significantly. 1.69 1.2 Client’s intrapsychic functioning improved significantly. 2.03 1.2 Client had not improved and further treatment seemed futile. 3.86 1.4 Client had improved some but further treatment would yield no appreciable gains. 3.62 1.5 Client wished to terminate. 2.62 1.4 Change in the therapeutic content or issues signaled readiness to terminate. 2.66 1.5 Increased transference was noted. 4.14 1.1 Decreased transference was noted. 3.92 1.2 Increased dependency was noted. 4.10 1.1 Decreased dependency was noted. 3.21 1.4 Agency had policy regarding length of treatment or circumstances for 4.34 continuing treatment. 1.3 Client’s external circumstances made termination desirable (e.p., changes in schedule, change in job, family circumstances, 3.53 transportation problems). 1.6 SD Chracteristics Associated with Criteria To determine whether criteria were more zyxwvu likely to be used under some circumstances than under other, statistical analyses were conducted in two steps. First, each criterion was paired with each of 16 client, practitioner, or case characteristics. Twenty-four (1 1%) of these analyses were statistically significant (p < .05), more than twice that which could be expected by chance. Nine characteristics were not related to anything and thus were dropped: client gender, race, and age; practitioner gender; gender match; importance of psychosocial theory; agency setting; time limits; and case duration in months. Second, because many characteristics were associated with one another, all significant variables were included in a multivariate analysis for each criterion. The results indicated which characteristics were significant when the effects of other characteristics were taken into account. Four criteria did not vary under any circumstances: meeting treatment goals, reaching time limits, lack of improvement, and further major gain not expected. All other criteria were more important in one or more situations. T h e importance of improved behavior depended on who raised the issue of termination. It was more important if the client suggested ending or if both did, less important when the practitioner raised the issue, and least important when external factors forced termination. Behavior was also more important as a criterion when the prdCtitioner rated the treatment as more successful. Improved intrapsychic functioning also was more important in the more successful cases. T h e importance of the client’s wishes varied by outcome (the client’s desire to terminate was more important in successful cases) and by who raised the issue of termination (more important if the client suggested termination or both agreed, less important if the practitioner raised the issue). Change in therapeutic content was considered a signal to end treatment if outcome was favorable and if the client had an affective disorder. Both decreased transference and decreased dependency were more important if outcome was better, where- zyxwvuts zyxwvuts Note: 1 = “very important in termination decision,” 5 “not at all important.” N = 69. = Criteria for Termination Practitioners rated each of 14 possible reasons to terminate treatment on a scale ranging from one to five: 1 = “very important in the termination decision for this case,” 5 = “not at all important.” The most important criteria were improvement in the client’s behavior, improvement in intrapsychic functioning, and meeting treatment goals (Table 1). Moderately important 368 z zy zy Terminating Treatment Fortune et d. lack of improvement, and agency policy. Other authors ( Firestein, 1978; Levinson, 1977) have stated that changes in dependency and transference indicate readiness to end successful treatment; this appears to be accurate in that decreases in both were indeed more important when outcome was better. Still, such subtle criteria were far less important in decisions to end treatment than were real changes in the client’s functioning. The unimportance of lack of improvement and of the cost-benefit criteria might be explained by the overall success of the sample cases. Howev- as increased transference and dependency were more relevant if termination was difficult. Among external reasons for ending treatment, agency policy was more influential if outcome was poorer, whereas the client’s external circumstances were more important for less experienced practitioners. Overall, the most important characteristic for terminating treatment was outcome (7 significant associations). Case characteristics related to criteria included who suggested ending ( 2 significant associations), when the client was prepared for termination ( I ) , and difficulty terminating ( 2 ) . Client diagnosis (1) and practitioner experience (1) were the only important personal characteristics. zyxwvuts Discussion The three most important indicators for termination were improved client behavior, improved intrapsychic functioning, and achieving goals. Improved behavior and intrapsychic functioning were usually used together. Both were more important in successful cases but did not vary with theoretical orientation. In short, practitioners waited for success before terminating and used both internal and external functioning to determine success. Using several criteria avoids the difficulty of determining which criterion is “correct” and is appropriate if the emphasis is on both the individual and the environment. However, practitioners use other criteria besides measures of functioning. Goals met, the third most common criterion, is often considered a less ambitious substitute for improvement (Fox, Nelson, & Bolman, 1969; Hess & Hess, 1984). This criterion was a partial substitute but was not interchangeable with functioning ( i t correlated moderately with behavior and intrapsychic functioning and varied only somewhat with outcome). Other important criteria were the client’s desire to terminate and changes in treatment issues as a signal of readiness to terminate. Like the improvement criteria, both were more important when outcome was successful. Thus, practitioners consider many indicators from several theoretical perspectives in determining when to end treatment. Factors such as gender or theoretical orientation of practitioner do not interfere w i t h outcome, goal accomplishment, or the client’s wishes as primary criteria for termination. The least important criteria were changes in the relationship (transference and dependency), zyxw zy Practitioners waited for success before terminating and used both improved behavior und intrapsychic functioning to determine success. er, neither of these criteria varied with outcome; they were not applied even in unsuccessful cases in which they were most appropriate. O n the other hand, agency policy was more important if overall outcome was poorer. Perhaps practitioners invoked agency policy to avoid admitting to themselves that further treatment was futile. Perhaps policies were applied to get rid of more difficult cases. Or perhaps practitioners rated cases as less successful because agency policy interfered with their plans. It should be noted that private practice did not seem to affect this criterion; agency constraints were equally unimportant in agency-based and private-practice cases. Time limits for treatment were seldom set. However, reaching the time limit was not important even in those cases that did have time limits. Other studies also suggest that practitioners rarely set or adhere to time limits (Fortune, 1985; Rooney, 1985). Other than outcome, few patterns of criteria were important. Practitioner experience had a minor effect on termination of treatment in that less experienced practitioners placed more importance on the external circumstances of the client. Perhaps these practitioners had difficulty handling such circumstances or were more open to admitting their presence. Whether the practitioner or client raised the issue of termination affected two criteria: if the client suggested termination or both did, the client’s wishes and improved behavior were more important criteria. However, other characteristics such as client age or gender were zyxwv 369 Families in Society June 1991 zyxwvut irrelevant. O n the one hand, this finding suggests lack of bias or stereotypes, for example, not treating men and women differently. O n the other hand, some variation by theoretical orientation, when preparation for termination begins, should he expected, according to the literature on tennination. Perhaps the measures of success were so broadly worded that subtle differences could not be picked up. Or perhaps such subtle differences are not relevant to the realities of practice. This study has several limitations. First, the practitioners and clients are not representative. The practitioners were selected from one region. Their willingness to complete a long questionnaire suggests that they were especially interested in termination. They may have selected ideal cases: open-ended, long-term, with voluntary terminations. However, these are the types of cases upon which the literature on termination criteria focuses. Consequently, they are appropriate for this study. Another problem, shared by other studies (Firestein, 1978; Kramer, 1982, 1986), is that practitioners’ memories of the cases may be faulty and perhaps idealized. However, practition- ers’ conceptualizations of what they are doing are the building blocks for theory and are closer to “reality” than are the untested assumptions found in most discussion of termination. Although a descriptive study cannot determine whether the criteria were used appropriately, the results suggest an appropriately pragmatic approach by practitioners. While emphasizing environmental functioning (presumably the client’s primary interest) and intrapsychic functioning, practitioners sought confirmation of success in other indicators: goals met, client desire to terminate, and changes in therapeutic content. Waiting until several different criteria indicate that it is time to end treatment helps ensure that termination does not occur prematurely. The risk in seeking such perfection, of course, is that treatment may be continued too long, beyond a point of diminishing returns. Perhaps the cost-benefit criterion should be invoked more often. However, the emphasis on relatively overt, verifiable criteria that indicate the client is functioning outside the treatment relationship is appropriate for service aimed at restoring or maintaining social functioning. zyxwvut zyxwvut zyxwv zyx zyxwvut zyxwvuts zyxwvutsrqpo zyxwvutsrq REEERENCES American Psychiatric Association. ( 1987). h p s t i c and statistical manual of mental disorders, 3rd ed., rev. Washington, DC:Author. Barton, B. R., & Marshall, A. S.(1986). Pivotal partings: Forced termination with a sexually abused hay. Clinical Social Workloud, 14, 139-149. Eldridge, W. D. ( 1983). Practitioners and self-evaluation. Social Casework, 64,42&430. Firestein, S. K. (1978). Termination in psychoanalysis. New York: International Universities Press. Fortune, A. E. (1985). Planning duration and termination of treatment. Socd Service Rewiew, 59,647461. Fortune, A. E. (1987). Grief only! Clienc and social worker reactions to termination. Clinical Social Workloud, 15, 159-171. Fortune, A. E., Pearlingi, B., & Rochelle, C. D. (in press). Reactions to termination of individual treatment. Social Work. Fox, E., Nelson, M., & Rolman, W. (1969). The termination process: A neglected dimension in social work. Social Work, 14, 53-63. Glover, E. (1955). The technique ofpsychoanalysis.New York: International Universities Press. 370 Greene, L. R. (1980). On terminating psychotherapy: More evidence of sex-role countertransference. Psychology of Women Quarterly, 4, 548-557. Hess, H., & Hess, €? M. (1984). Termination in context. In B. Compton & B. Galaway (Eds.), Socd work processes, 3rd ed. (pp. 559-5701, Homewood, IL: Dorsey. Krarner, S. A. (1982). The hapist’s view of m i n u t i a in open-ended psychotherapy. Unpublished doctoral dissertation,University of Chicago. Krarner, S. A. (1986). The termination process in open-ended psychotherapy: Guidelines for clinical practice. Psychotherapy, 23, 526-531. Levinson, H. L. (1977). Termination of psychotherapy: Some salient issues. Social Casework, 58, 48C.489. Richey, C. A., Blythe, B. J., & Berlin, S. B. (1987). Do social workers evaluate their practice? Social Work Research and Absnacts, 23, 14-20. Rooney, R. H. (1985). Does inservice training make a difference?Results of a pilot study of task-centered dissemination in a public social service setting. l o u d of Social Service Research,8,33-50.