Therapeutic Alliance and Treatment Outcome

I've been reading about different treatment modalities on the various forums here at Behavior OnLine. I want to offer the following brief abstract from research I am doing on the role of therapeutic alliance in treatment outcome. I'd be interested in further discussion about the abstract and/or the questions which follow.

Brian O'Neill, editor Gestalt Therapy Forum

The literature on therapeutic alliance has in the last decade included work with suicide , major depression, personality disorder , affective disorder and schizophrenia and a number of studies link therapeutic alliance with positive outcome in treating serious mental illness , in particular schizophrenia .

Authors have defined a relationship aspect of alliance (as described by the psychoanalytic and client-centred therapy schools), and a task aspect (as found in behaviour therapy literature). Others define therapeutic alliance as comprising both aspects, task and relationship, where alliance is the bond between therapist and client (relationship) and the agreement between them concerning the goals and tasks of the therapeutic endeavour (task).

Studies have considered the relationship between therapy approaches, therapeutic alliance and successful treatment outcome. Frank & Gunderson (1990) found no significant difference between two types of psychotherapy in relation to outcome, but did find a strong relationship between therapeutic alliance and outcome. Salvio (1992) compared three therapy approaches, cognitive-behaviour therapy, Gestalt therapy and client-centred counselling and found no significant difference between therapies and the strength of the alliance.

However Hansson and Berglund (1992) in a study of outcome in short term in-patient psychiatric care, found that patients who improved their alliance were more often subject to psychopharmocological treatment. Raue (1992) found higher alliance scores for cognitive-behavioural therapy as compared to interpersonal psychotherapy, while Marmar (1989) in comparing the effects of alliance on treatment of depression across behavioural, cognitive and brief psycho dynamic therapies found a significant relationship between outcome and alliance only for cognitive therapy.

In a meta-analysis of 24 studies, Horvath and Symonds (1991) found a correlation between therapeutic alliance and outcome across all types of measure of outcome, which was not a function of the type of therapy practiced. The measure which was most predictive of treatment outcome was that based on the clients' assessments of alliance, while ratings by therapists were less predictive and ratings done by observers least predictive of outcome. Other studies show little correlation between staff and client perceptions of alliance, with the client perception as the critical factor in predicting outcome.

Two Questions:

Is therapeutic alliance a key mediator between therapeutic approaches and treatment outcome?

Perhaps we need to consider not which therapy works best with which disorder but how does each therapy contribute to developing therapeutic alliance, and hence better treatment outcome?

References and/or more detailed abstract supplied on request.

• No Statistical Proof Needed! I'm Convinced by Experience! by c.h., 2/1/97
• The Role of the Therapeutic Alliance in Affecting Treatment Outcome by Bruno G. Just, 4/23/97