Patient-focused research: using patient outcome data to enhance treatment effects.

A program of research aimed at improving the quality of psychological interventions is described. Data from over 10,000 patients were analyzed to understand the association between number of treatment sessions and clinically significant improvement. In addition to a potential dose-response relationship, typical recovery curves were generated for patients at varying levels of disturbance and were used to identify patients whose progress was less than expected ("signal" cases). The consequences of passing this information along to therapists were reported. Analyses of dose-response data showed that 50% of patients required 21 sessions of treatment before they met criteria for clinically significant improvement. Seventy-five percent of patients were predicted to improve only after receiving more than 40 treatment sessions in conjunction with other routine contacts, including medication in some cases. Identification of signal cases (potential treatment failures) shows promise as a decision support tool, although further research is needed to elucidate the nature of helpful feedback. Outgrowths of this research include its possible contribution to social policy decisions, reductions in the need for case management, use in supervision, and possible effects on theories of change.