resources within the environment rather than on support persons, per se (e.g., sheltered work and living settings, and special discharge programs). The distinction between supportive persons and supportive settings simply highlights the different ways in which environmental modifications occur. In practice, these modifications often occur simultaneously. The main identifying feature of both types of supportive interventions, as distinguished from skilldevelopment interventions, is that they do not attempt to change the patient's behavior systematically and directly. Rather, the attempt is simply to support and accommodate the patient's present level of functioning. The early studies by Katkin (1971, 1973) and a more recent study by Cannady (1982) have clearly demonstrated the positive impact on patient outcome of a support person. Cannady (1982) used citizens from the discharged patients' rural neighborhood to function as "supportive case workers." Over a 12-month period, inpatient days were decreased by as much as 92 percent. Witheridge, Dincin, and Appleby (1982) have reported on the use of a support team for psychiatrically disabled individuals at high risk for readmission. Working out of the patients' homes and neighborhoods, this team had as its goal to develop an individualized support system for each patient. Of the original 50 participants, 41 remained in the program. One-year followup data indicated that average days hospitalized per year decreased from 87.1 to 36.6. Even better results were obtained in a similar program which has been a model for assertive outreach and advocacy for chronic mentally ill persons (Stein and Test 1978). A study by Stickney, Hall, and Gardner (1980) investigated the effects of introducing both a support person and a more supportive environment, separately and in combination, at the time of the person's hospital discharge. They studied the impact of four predischarge formats, which differed in level of personal and environmental support. The goal of the discharge plans was to increase use of the CMHC and to decrease recidivism for 400 patients discharged from a State mental hospital. The results of the study demonstrated the impact of increasing personal and environmental support on patient compliance with referral and 1-year hospital recidivism rates. With minimal support, the respective referral compliance and recidivism percentages were 22 percent and 68 percent; with increased personal support, 36 percent and 39 percent; with increased environmental support, 68 percent and 31 percent; and with both increased personal and environmental support, 75 percent and 28 percent. Thus, whenever an added element of support was introduced, the referral compliance percentage increased and the recidivism rate decreased. Ballantyne (1983) and Wasylenski et al. (1985) have reported results from an environmental intervention study based directly on the principles of psychiatric rehabilitation (Anthony 1979). Community rehabilitation workers were trained to provide interpersonal support and case management to patients being discharged from the hospital. Compared to a matched group of patients receiving traditional discharge planning and aftercare, patients in the experimental group were discharged an average of 21 days sooner and participated in approximately twice as many aftercare services. Patient perception of the program indicated high levels of involvement and satisfaction. At 6-month followup, over 75 percent of the patients were still involved in the program; 87 percent reported "feeling understood"; and 88 percent reported "feeling like a partner" in the rehabilitation planning process. In addition, there were significant differences in instrumental role functioning as student, worker, or homemaker. Weinman and Kleiner (1978) compared the effectiveness of community-based "enablers" with two hospital-based conditions: sorioenvironmental therapy and traditional hospital treatment. The enablers' major roles were to assist their patients in coping with community life and escort them to various community resources. Personal support proved superior to each of the hospital-based treatment approaches in