Send email for updates
Journal of Studies on Alcohol: 1981, 42(9), p. 783–790.
Unable to obtain a copy by clicking title? Try this alternative source.
Seminal US study which found that the therapy-related social skills of alcohol counsellors were strongly related to how many of their patients relapsed in the two years after leaving inpatient treatment.
Summary At a US hospital-based alcoholism treatment unit 247 inpatients admitted for the first time were randomly assigned to one of eight counsellors. Six of the counsellors were men. All eight were recovered alcoholics who had attended a school of alcohol studies or had equivalent training experience. Most patients were men, nearly all were employed, and about two thirds married.
I wish everybody would stop talking to me about drinking. So I have a few too many drinks once in a while. If my wife would stop nagging me about the bills and the kids and everything else, I wouldn't have to drink. If she would just shut up.
Please tell me what to do to get my sister to stop drinking. I have tried everything but nothing works. What am I going to do?
I know I love him, why, I remember when we would spend all weekend together fishing or building something in the backyard. But he's not like that any more and when I see him on the streets, I try to avoid him. Even my friends are making jokes about him. Mom says he's sick, but if he's sick, why isn't he in the hospital? I just don't know what to say to him, or how I should feel. After all, he is my father.
Lately, it seems we're always leaving parties early and I find myself worrying about what my wife will do or say instead of enjoying my friends. Then we talk about it the next day and she always says she doesn't remember. She's really been acting strange lately ... and I wonder, well, do you think she could have a drinking problem? I know she's not an alcoholic because she can go without a drink for weeks, but when she does drink, well ...
I can't do it, I can't stay sober. No matter how hard I try, I can only get a few weeks together. I might just as well admit I'm a no good bum.
Reproduced from: Saarnio P. et al. "Rating therapists who treat substance abusers." International Social Work: 2002, 45(2), p. 167–183.
Treatment over an average 12-day stay consisted of individual and group counselling, lectures, Alcoholics Anonymous meetings, psychological evaluation, recreational therapy, and daily consultations with doctors. Counsellors were the primary therapists, responsible for coordinating all services for patients and had the most contact with them.
The counsellors were asked to submit written responses to several written scenarios intended to approximate actual interactions between counsellors and their patients or their patients' relatives panel. Two trained raters then scored these responses for the degree to which they exhibited:
• Empathy: ability to respond to the feelings and reasons for the feelings the patient is experiencing in a manner which communicates understanding.
• Genuineness: degree of sincerity manifested in the helping relationship; extent to which counsellor exemplifies a manner free of roles and not 'phony'.
• Respect for the client: also known as positive regard; the ability to convey to patients that they are persons of worth, best communicated through warmth and understanding; crucial in establishing a basis for empathy.
• Concreteness: ability to be specific and direct in expression of feelings and experiences; serves to ensure the counsellor's responses stay on target and are accurate in terms of the patient's feelings and experiences.
Scores on these four dimensions were combined to arrive at a rating of each counsellor's overall level of interpersonal functioning. At the lowest level the counsellor did not respond helpfully either or both of the content of the patient's communication or its emotional tone; medium level counsellors responded accurately to both; higher level counsellors did this and more, offering perceptive personalised responses and at the highest level also personalised treatment goals and plans. This tripartite grading of counsellors was then related to their patients' progress in the six months to two years after they had left the unit, assessed for all patients via hospital records and by a survey of patients at the final follow-up, to which two thirds responded.
The level of interpersonal functioning of the eight counsellors stretched across the range from low to high. In general, the higher it was, the fewer of their patients were known from hospital records to have relapsed, and the less extended and fewer the relapses.
For example, over the entire two years ( chart) 17% of the patients of low functioning counsellors were known to have relapsed at least twice compared to just 3% of those of high functioning counsellors. Corresponding figures for any relapse were 38% and 18%, and for cumulative days in relapse eight versus two days. On all these measures the progress of patients of medium functioning counsellors was intermediate, meaning the relapse indicators improved as patients had been assigned to increasingly well functioning counsellors.
However, this pattern did not hold for patients' accounts of whether they had drunk at all over the two years. Again the worst figure was for the patients of low functioning counsellors, 79% of whom had drunk. But at 51% the best figure was for the medium functioning counsellors. In between at 64% were the high functioning counsellors.
Of the battery of 22 measures of the quality of personal and work life and of drinking and alcohol-related adverse consequences completed by patients two years after leaving treatment, it was only the percentage who had at some time drank after leaving treatment which differed to a statistically significant degree depending on their counsellors' functioning. Neither was retention in the initial treatment related to counsellor functioning.
The findings suggest that the better the interpersonal skills of alcoholism counsellors, the better the drinking outcomes for their patients. These results have implications for the process of determining what constitutes an effective alcoholism counsellor and in turn for the certification of counsellors. They indicate that one measure of counselling quality – interpersonal skills – is quite strongly related to treatment outcome, so may prove worthy of consideration as a criterion for certification of individual counsellors and accreditation of counsellor training programmes. Emphasising quality by assessing the interpersonal skills of counsellors and providing related training is one way to improve the quality of alcoholism services.
commentary Such studies stand in contrast to generally negative findings on the relationship between how their clients do and clinicians' professional background characteristics, such as years experience and training and whether they are themselves former problem substance users. When their impacts are allowed to emerge, studies commonly find that clinicians make a big difference to outcomes, but rarely are these differences related to the 'hard' variables of qualifications or sex/race match with patients. Though rarely studied, social skills, and whether the counsellor's way of relating to other people suits the patients, are more promising potential causes of differences in counsellor performance.
Considered notable for its large sample size and random assignment of patients to counsellors, the featured study also had the benefit of predating the trend to test treatments so highly standardised and delivered by therapists so highly selected, trained and supervised, that the impact of counsellor quality (if assessed at all) is minimised, though not always entirely ironed out. In many such studies the lowest level of competence scored by the featured study would have led those counsellors to be eliminated from the study or subject to further training and supervision until they conformed to the study's quality standard, in order that the interventions being tested were delivered as intended.
In this way studies have risked eliminating what matters (the quality of the therapeutic relationship) in order to highlight what often matters not at all (the type of psychosocial therapy). The featured study instead took 'run of the mill' counsellors not subject to special training or supervision and tested what happens when the variation in patients is eliminated by randomly allocating them to the counsellors. This study design exposed the variation in counsellor/therapist effectiveness obscured in more highly controlled studies. It validates the (when these questions are asked in studies) common attribution by patients/clients of part of the impetus for their recovery to their therapist's qualities and how they related to them, and the common understanding of staff and patients that some practitioners are highly regarded whereas others are avoided.
With just two counsellors each in two of the skill categories the study's results were vulnerable to the particular characteristics of those individuals. It would also have benefited from a more detailed account of how it was decided that someone had relapsed and the duration of that relapse. It seems this was decided by their return to the clinic for further treatment, yet this may reflect willingness to return in response to a relapse rather than the frequency of relapse. It seems possible that people who found their counsellor sympathetic and non-judgemental would be more willing to admit to them that things had gone wrong and seek further help from the same source, while others who had also relapsed might not seek help or go elsewhere.
Arguing against this, however, is the puzzling reversal in the proportion who had drunk at all, at its lowest not among patients of the 'best' counsellors, but of the intermediate ones. It meant that of the patients who had drank, just 29% who had seen the most skilled counsellors were known to have relapsed compared to 48% each of those who had seen either medium or less skilled counsellors. This pattern does not suggest patients of the higher rating counsellors were more willing to come back, rather that they were more able to drink without major problems.
Another conundrum is why retention was not improved by the higher rating counsellors. Typically patients who feel positive and comfortable about their therapists stay longer in treatment than other patients, even if they fare no better in terms of substance use. In the featured study the reverse was the case. One explanation may be the inpatient setting. The constraints on mainly married and employed men on how long they could stay away from home and work, the atmosphere at the unit, and its environment, were possibly the determining factors, overwhelming any impact of the counsellor.
Over two decades later a similar study was conducted in Finland where four counsellors were rated (this time by social workers and social work students not trained in the rating method) on the same dimensions used by the featured study and using the same written scenarios. Counsellors tended to rate high or low on all four dimensions and (as with the trained raters in the featured study) the raters largely agreed in their assessments. The Finnish study was conducted at an outpatient substance misuse clinic, and this time the ratings were related to retention.
Over five scheduled sessions, patients attending the service for the first time who dropped out of treatment were compared with those who had stayed or left by mutual agreement. In this study there was less variation between the professional therapists than between the ex-alcoholics in the US study; all scored as high functioning according to the US study's cut-off point. Nevertheless 63% of one therapist's clients dropped out ranging down to just 11% for another. This variation was largely predictable from the ratings made of their interpersonal skills. It was also predictable from the post-session ratings therapists and clients made of their 'rapport' with the other; when client and therapist agreed that their rapport was good (that is, in the top half of the distribution) only a fifth of clients dropped out; when both saw it as poor, half did so.
Last revised 11 April 2013. First uploaded 11 April 2013
Give us your feedback on the site (two-minute survey)
Open Effectiveness Bank home page
Add your name to the mailing list to be alerted to new studies and other site updates
DOCUMENT 2013 Sometimes best to break the rules