Premature termination of psychotherapy is a pervasive problem-30 to 60% of psychotherapy clients drop out of treatment before its completion. This review summarizes 3 decades of research on the topic. Client, therapist, and administrative variables have been extensively investigated. Because of a variety of methodological problems, this literature is highly contradictory, and results are difficult to reconcile, with only socioeconomic status (SES) and ethnicity emerging as consistent predictors of dropout. Research looking at interactive and multidimensional factors such as working alliance, satisfaction, patient likability, and expectations has proven more useful than research on client, therapist, and administrative variables. Findings suggest that dropouts might be minimized if differences between therapists' and patients' perspectives on the therapeutic enterprise are

Practicing clinicians realize that patients' dropping out of therapy is detrimental to treatment outcome and can prove costly to psychotherapists in terms of financial and personal consequences. Two procedures to prevent therapy dropout were tested in the "real-world," naturalistic environment of a health maintenance organization (HMO). Whereas video preparation significantly reduced dropout, opportunity to estimate treatment duration did not. Results were obtained from 125 randomly assigned adult outpatients. Findings suggest that psychologists in clinical and administrative positions may experience reduced dropout rates by providing new patients with videotaped instructional material about what they might expect in the psychotherapy process.

The Hopkins Symptom Checklist (HSCL-25; Mattsson, Williams, Rickels, Lipman, & Uhlenhuth, 1969) was translated into the Hmong language and administered to 159 Hmong adults, 73 nonclinical and 86 mental health clients. The instrument demonstrated internal consistency of .97 and had a split-half coefficient of .92 and test-retest reliability of .90. Mental health clients produced scores that were significantly higher than those of nonclinical participants on the Anxiety, Depression, and Total scores. Consistent with expectations, Hmong more intensely affected by the casualties of war, those currently unemployed, those older, and those with less education tended to report more symptoms of anxiety and depression. The Hmong version of the HSCL-25 provided a sensitivity of 100%, specificity of 78%, and overall accuracy of 89%, demonstrating that it is a useful screening tool for assessing general distress and anxiety problems in Hmong people.

We developed the Hmong Adaptation of the Beck Depression Inventory (HABDI) and evaluated the instrument's psychometric characteristics. Also examined was the relationship between depression and demographic variables such as age, sex, length of stay in America, English-speaking ability, and social support in Hmong refugees. One hundred twenty-three Hmong living in Fresno County, between the ages of 18 and 66, participated in the study. The new measure demonstrated a high coefficient alpha (.93), and test-retest reliability (.92), and a significant mean score difference between the nondepressed and the depressed groups. Individual items were distributed evenly and correlated highly with the total depression score. The HABDI correctly identified 94% of depressed and 78% of nondepressed in the Hmong sample. The results suggest that quality of social support and years of education play important roles in buffering Hmong refugees against depression, whereas length of stay in America and number of social supports do not.

No abstract