Therapeutic communities: what are the retention rates?

Abstract

The present research is the first in a systematic investigation of retention in therapeutic communities (TCs). Findings are reported from an interrelated set of analyses clarifying retention rates and the temporal pattern of dropout. Retention information was obtained on all admissions to a nonrandom sample of seven traditional (long-term) TCs during February 1-August 15, 1979. Results showed that: (a) Across the seven programs, the 12-month retention rates ranged from 4 to 21%. Program differences were generally nonsignificant. (b) Less than 17% of admissions were readmissions. Among these, however, the number of days before initial dropout was a significant predictor of long-term retention. (c) The temporal pattern of dropout was uniform across programs. Dropout was highest within the first 14 d and declined thereafter, indicating that practically all dropouts leave within 3 months. (d) Based upon rates adjusted for those who had left treatment ("survivor rates"), estimates indicated that the likelihood of continued retention increased significantly with longer stay in treatment. (e) The temporal pattern of dropout was remarkably stable, evident in both replicational and comparative analyses; the latter involved the entire residential drug-free modality in several data systems. The present findings establish that the temporal pattern of retention is orderly, predictable, and has utility for program management and treatment planning, although the phenomenon of retention itself remains to be explained.