Methods

Subjects were patients waiting to see their physicians at either a medical school family practice residency training center or one of six community family practice offices. Trained medical student research assistants requested the patients’ participation in a research study about their feelings and health and answered questions related to the project before obtaining the patient’s signed authorization on a consent form. The research protocol was approved by the Institutional Review Board of our affiliated institution. All patients age 18 and over who could communicate verbally were eligible to enter the study. Patients were introduced to the study in the examination room at the medical training center and in the waiting room at the community practices. Three different protocols, based on site specifications, were used so the research assistants could explain the research project to as many patients as possible without interfering with the normal office flow of patients to see their physician. Patients were apprised that only the researchers and their doctor would have access to their information. In addition to the MOS screener (Table 1), subjects also completed a two-page questionnaire about demographics (Table 2) and health habits. A total of 1,023 patients completed the MOS screener, 566 from the training center and 457 from the community sites. Because the MOS depression screener was developed using logistic regression to better predict depressive disorders, scoring usually requires using a calculator to compute the differential weighting, in decimals, for each of the eight items. However, a scoring strategy that transformed the decimals into integer scores for each item was developed, and a scoring template was devised by one of the authors to quickly score each patient’s responses and corresponding depression status. Using depression status from the MOS, all patients scoring above the recommended cutoff and a random sample of 25% of those scoring below the cutoff were interviewed with the DIS. The medical student research assistants, trained to administer the DIS over the phone, were provided a randomized listing of patients to be called. They called each subject and administered a computerized version of the DIS to detect current depressive disorders.