A logistic regression analysis of predictor variables to successful outcome in employee assistance

This retrospective descriptive survey examined client records from two years of an internal medical campus Employee Assistance Program in order to analyze process variables related to successful and unsuccessful treatment outcomes. The treatment in question was brief mental health counseling of one to four sessions. Crisis intervention and brief solution-oriented therapy concepts along with EAP program information protocols were used to generate likely predictor variables. Specific client, referral, and problem related characteristics were identified as relating to counseling outcome. Logistic multiple regression analysis was employed to determine the discriminant and predictive capabilities of those characteristics.;Criteria for determining treatment success, the dependent variable, consisted of level of function or GAF upon termination, coping at work upon termination, and treatment goals accomplishment upon termination. Documented improvements in functioning and work coping plus treatment goal accomplishment earned a rating of "successful." Lack of documentation of such, or documentation of negative change, earned a rating of "unsuccessful." The dichotomous dependent variable of treatment outcome dictated the choice of logistic multiple regression analysis.;The single significant predictor of treatment outcome was "job status." For each one-point increase in job status, which was a function of educational level and professional achievement, the client was 45% more likely to be successful in short-term EAP counseling. Several other variables approached significance, including gender, social support, and referral source. The combination of descriptive survey and statistical analysis results from the study effectively described the contingencies operating within the targeted EAP program. Implications and recommendations were made for EAP program development, use of brief counseling interventions, and ongoing evaluation. This research should be repeated in other EAP settings using prospective data, to determine the applicability and wider usefulness of its results.