to be a patient problem, and little attention has been given to the way that program staff, administrative policies, and resources are used to respond to patient needs. In their review of the literature, they observe that:
“Less attention has been given to variations between treatment programs in how well they engage their patients. Poor therapeutic involvement usually has been assumed to be a problem of the individual, sometimes involving obstacles or deficits to overcome...before investing fully in the program. However, personal levels of motivation and involvement are likely to depend not only on individual experiences, but also on how well program staff and resources can respond to patient needs. The consistency of patient-level indicators of engagement within programs as well as their between-program variations therefore deserve closer study.” (p. 128)
Broome, Simpson and Joe used hierarchical linear modeling techniques to estimate the influence of patient characteristics, treatment experiences, and program-level factors on patients’ “therapeutic induction and involvement.” They acknowledged, however, that their program-level variables were limited to only a few measures of the service delivery approach and “program climate,” based on aggregated patient responses. These measures included average number of referred services, proportion of patients who knew general information about their personal treatment plans, average frequency of missed counseling sessions, average ratings of social pressure from staff and other patients to stay in the program, and diversity of needs expressed by patients. Their results suggested the importance of further investigation of probable interactions between patient and program attributes and their implications for patient and program outcomes.
Researchers have also used the SROS data to investigate the influence of program staff,