Send email for updates
McLellan A.T., Woody G.E., Luborsky L. et al.
Journal of Nervous and Mental Disease: 1988, 176(7), p. 423–430.
The unexpected resignations of two counsellors at a US methadone clinic in early 1985 triggered a unique study of the influence of counsellors on their patients’ recovery. Its insight remains relevant today, and the study has been added to the Effectiveness Bank as a piece of ‘old gold’.
Summary There is a large and still growing literature devoted to evaluating the effectiveness of different types of counsellors. More than 50 studies have compared ‘professional’ counsellors (ie, those with master’s level education), ‘paraprofessional’ counsellors (ie, those with less than a master’s level education), and counsellors with prior histories of problem substance use who may or may not have a master’s level of education. Overall, it seems there is no significant difference in the “performance” of patients assigned to these different types of counsellors (eg, based on drug use during and after treatment, and retention in treatment), and that “paraprofessionals achieve clinical outcomes equal to or significantly better than those obtained by professionals”.
The unexpected resignations of two counsellors in a US methadone clinic triggered a unique study of the role of the individual counsellor.
Going by the trajectory of patients, and differences between the progress of each counsellor’s caseload, the ‘better counsellors’ were found to show consistent and professional patient management practices, enabling the patients to stabilise their lives and to use the other social and medical services available to address their problems.
These may be the minimal conditions necessary for continued patient improvement, even with the powerful effects of methadone.
The featured report examines the role of the individual counsellor in the recovery of male ex-military veterans in a US methadone maintenance programme in early 1985. The unexpected resignations of two counsellors at the Philadelphia Veterans Administration Medical Center provided an opportunity to examine the effects of assigning patients to a new counsellor, while keeping other conditions the same including medication philosophy, programme rules, and supplementary services. It was anticipated that this unusual situation would provide a relatively pure test of the contribution of the counsellor without the artificiality of an experimental setting, and without the influence of the multiple factors generally associated with assigning or reassigning patients in the normal clinical situation.
The clinic under study collected weekly records of patients’ substance use, methadone dose, and prescriptions for additional medications, as well as employment status and whether there were any arrests. This made it possible to compare the progress of patients during the six months before and six months after being assigned to new counsellors. The five measures used to evaluate performance were:
As a result of the large number of patients affected by the counsellor resignations (80), and the abruptness of the situation, the normal clinic policy of contacting individual patients and counsellors to develop mutually satisfactory ‘matches’ was not possible. Instead, the senior clinical supervisor reassigned patients over a three-day period, based entirely upon the size of the remaining counsellors’ caseloads. Although the reassignments were not entirely random, they were also not based on either counsellor choice or patient characteristics.
The caseloads were examined to see whether there were any differences in the demographic and background characteristics of patients. All comparisons showed non-significant differences.
Of the 80 patients reassigned, 68 who had been in treatment at least six months before transfer and who stayed in treatment six months after transfer were considered eligible for the study. The final sample then included the 61 patients assigned to one of the four counsellors deemed to have an adequate number of patients (at least 12). At issue was how the progress of these patients effectively reallocated at random differed between the four counsellors, known as counsellors 1, 3, 3 and 4.
Methadone dose. The transfer had a significant effect on the dose of methadone. During the post-transfer period, the caseload of counsellor 2 had a significantly lower average methadone dose than the other caseloads, which did not differ significantly from each other.
Additional medications. Overall, there was no change in the proportion of patients receiving medication for depression, sleep problems, anxiety, etc. However, there were significant differences between the caseloads. For example, counsellor 3 showed a modest increase in the proportion of patients receiving prescribed medication, whereas counsellor 2 showed a significant decrease.
Urine tests. There was a general decrease in the proportion of urine samples testing ‘positive’ for the use of the specified drugs after the transfer (from 37% to 29% overall). However, the changes were not uniform across the counsellors. Among the patients of counsellor 3 the proportion of positive urine tests increased, differing significantly from the patients of the remaining three counsellors, whose caseloads decreased, and did not differ significantly from each other.
Employment. There was an overall improvement in the proportion of patients in education or full- or part-time employment. However, the percentage of counsellor 3’s patients employed declined after transfer, and was significantly lower than the other caseloads.
Arrests. The proportion of patients arrested and charged with a crime was generally low (4%) throughout the period of study. Yet, there were still modest reductions in the arrest rates of patients in three of the four caseloads, though not to a statistically significant degree.
Counselling is an active ingredient in the recovery of patients with problem substance use. As seen in earlier studies of psychotherapists (1 2), there are significant differences in the effectiveness of counsellors; some promote rapid and sustained change in their caseloads, and others actually detract from the effectiveness of the other components of treatment.In the featured study there were statistically significant and clinically meaningful differences between the progress of counsellors’ caseloads, which, from the analyses conducted, were presumed not to be due to differences in the demographics or treatment backgrounds of the patients.
Counsellor 3 was the only one to average an increase in methadone dose among patients, but also to see an increase in ‘positive’ urine tests (indicating continued substance use) and unemployment, whereas counsellors 1, 2 and 4, who decreased patients’ methadone doses (albeit for counsellor 1 to a considerably smaller degree) saw a decrease in positive urine tests, unemployment, and arrests.
Although it is possible that there were important differences between the counsellors’ caseloads that were not measured or not detected, the comparisons conducted convinced the researchers that there were no explicit or overt differences among the caseloads at the time patients were reassigned, and the procedure for reassigning reassured them that patients were (as far as possible) effectively reallocated at random.
The four counsellors, who averaged more than eight years’ working in the field, took different routes to their professions. For example, counsellor 3 had a high-school degree and prior history of substance use, counsellors 1 and 4 had bachelor’s degrees in psychology, and counsellor 2 had a master’s degree in psychology and had received a state license to practice. While differences in the level of formal education could have had an effect, a large number of prior studies (see review) suggest that patient outcomes cannot generally be attributed to educational differences.
Due to the spontaneous nature of the study, it was not possible to tape counsellor–patient sessions as a means of examining differences. Instead, patient records were consulted, and revealed clear and consistent differences in the written work of counsellors.
For example, the written notes of counsellors 1 and 4 indicated that they performed their counselling duties in a highly organised manner. There was a formulated plan of recovery that had been worked out through consultation with the treatment team and the patient, and the initial plans were generally followed and documented at all points during the process. In contrast, the information from the records of counsellor 3 (whose patients progressed relatively poorly) indicated far less organisation in the treatment plans and far less detail in the treatment notes. There was little indication of consistency in the enforcement of programme rules, and very little documentation of the use of programme resources. Most striking was the general lack of documentation surrounding changes in methadone dose and prescription of additional medication. Finally, according to the notes patients were seen less frequently than the patients of the other counsellors.
The records of counsellor 2 indicated the same high level of organisation and consistency as counsellors 1 and 4, but a different counselling process – one of anticipating problems and discussing strategies to deal with them, thereby focusing recovery on the development of new behaviours and new ways of thinking by the patient.
Overall, counsellors 1 and 4 – considered the better drug counsellors – demonstrated consistent and professional patient management practices, enabling their patients to stabilise their lives and to use the other social and medical services available to address their problems. These patient management techniques appear to be the hallmark of the effective, professional counsellor and may be the minimal conditions necessary for continued patient improvement, even with the powerful effects of methadone. In the case of counsellor 2, the use of psychotherapy techniques in addition to the basic patient management skills may have enabled the patients to reduce their need for mood-stabilising medications (including methadone), additional medication, and street drugs. However, these conclusions must be considered speculative as they result from impressions and indications derived from an examination of patient records after the results of the study were already known.
commentary Answering the title question of the featured paper, the researchers concluded that ‘yes’, counselling is an active ingredient in the recovery of people with problem substance use. The study was triggered by the sudden and unexpected resignations of two counsellors in a US methadone clinic in early 1985, described elsewhere in the Effectiveness Bank as “one of those serendipitous accidents” exposing previously obscured influences.
Medication such as a maintenance dose of methadone can be a significant part of medical care, but is never all there is to this care. Relationships with staff can affect outcomes indirectly by enabling more or less appropriate prescribing, and directly through the provision of advice and psychosocial support. These are among the reasons why of medical care in general, it has been said that “ ‘talk is the main ingredient … and it is the fundamental instrument by which the doctor–patient relationship is crafted and by which therapeutic goals are achieved’. From this viewpoint, a good interpersonal relationship can be regarded as a prerequisite for optimal medical care.”
As the relationship-forming qualities of treatment staff are much less easily manipulated by researchers and less commonly researched than the intervention itself, studies usually have to rely on an association between relationship quality or staff attributes on the one hand, and retention or substance use on the other, rather than deliberately allocating patients to workers known to differ along these dimensions. This makes it difficult to be confident that relationship quality or staff attributes actually caused any differences in outcomes.
The featured study was able to overcome this limitation without the artificiality of a randomised trial. It gave researchers the unique opportunity to see whether four counsellors, allocated effectively at random at least a dozen patients each, altered the trajectory of patients’ progress from the six months before to the six months after. Not only did this permit the counsellors to be compared with each other, it also meant pre-transfer measures could be used to assess how patients changed, not just where they ended up.
Despite higher methadone doses and more of their patients being prescribed other medications, those transferred to the least effective counsellor worsened in their drug use, arrests and employment. In contrast, over a six-month period the most effective counsellor was able to bring patients to the point where their non-prescribed drug use was virtually zero and employment significantly improved, while at the same time substantially reducing their doses of methadone and the proportion prescribed ancillary medications.
See the Effectiveness Bank matrix bite to understand how other studies have investigated the influence of the practitioner in medical care, and what they have found, and skip to this section for further commentary on this virtually unique study.
Last revised 04 May 2018. First uploaded 17 April 2018
Comment/query to editor
Give us your feedback on the site (one-minute survey)
Open Effectiveness Bank home page
Add your name to the mailing list to be alerted to new studies and other site updates
STUDY 2008 Does following research-derived practice guidelines improve opiate-dependent patients’ outcomes under everyday practice conditions? Results of the Multisite Opiate Substitution Treatment study
STUDY 1999 Treatment staff matter as much as the drug
MATRIX CELL 2018 Drug Treatment Matrix cell B3: Practitioners; Medical treatment