African-American cocaine dependent patients, these authors found a paradoxically negative relationship between the reported number of available family and friends of the patient and relapse to cocaine use following treatment: the more friends and family available to the patient, the earlier the return to cocaine use. The authors' hypothesize that in this severely affected cohort of patients, the only available sources of social support may have been associates with whom the patients had previously used drugs.
Patient factors have been much more widely studied than treatment setting, modality, process, and service factors as predictors of outcome from addiction rehabilitation treatments. Perhaps the major reason for this is that while there have been many reliable and valid measures of various patient characteristics, there are still very few measures of treatment setting (Moos, 1974; Moos et al., 1990) or treatment services (McLellan et al., 1992a; Widman et al., 1997).
There is good news however, regarding the study of treatment factors in the substance abuse field. Recent developments in the psychotherapy field have led to the creation of manual-based treatments and with them, appropriate measures of treatment fidelity and integrity. Following on this progress, the multisite NIAAA study of patient treatment matching (Project MATCH, 1997) has provided the field with new manuals for the three Project MATCH treatments as well as additional measures of the nature and fidelity of each treatment. These are likely to improve the study of addiction treatment process in the years to come. Below we review several dimensions or characteristics of treatment that have been studied and that have shown some relationship with outcome following treatment.
Setting of Treatment—There have now been many studies investigating potential differences in outcome between various forms of inpatient and outpatient rehabilitation. For example, studies by McCrady et al. (1986) and Alterman et al. (1994) randomly assigned alcohol dependent patients to an equal length (28-30 days) of either inpatient or day-hospital rehabilitation, where the treatment elements were also designed to be similar. Both studies showed very similar findings. Patients in both the inpatient and outpatient arms of both these studies showed substantial and significant reductions in alcohol use, as well as improvements in many other areas of personal health and social function—suggesting that both settings of care were able to produce substantial benefits. At the same time, a wide range of outcome measures collected at six-month follow-up in both studies, showed essentially no statistically significant or clinically important differences be-