Homework “Dose,” Type, and Helpfulness as Predictors of Clinical Outcomes in Prolonged Exposure for PTSD

Abstract

Homework is often viewed as central to prolonged exposure (PE) for posttraumatic stress disorder (PTSD), but its relationship with treatment outcome is not well understood. We evaluated homework type, dose, and patients’ perceptions of helpfulness as predictors of symptom change and posttreatment outcomes in PE. Patients with chronic PTSD received PE in a randomized clinical trial. Independent evaluators assessed PTSD severity at pre- and posttreatment. Patients reported homework adherence and perceived helpfulness at the beginning of each session, separately for in vivo and imaginal exposure assignments. These variables were examined as predictors of change in PTSD symptoms, PTSD remission, and good end-state functioning (GESF; low PTSD, depression, and anxiety) at posttreatment. Higher imaginal homework adherence predicted greater symptom improvement between sessions and across treatment, as well as twice the odds of achieving remission and GESF. Patients who were at least moderately adherent to imaginal homework assignments (two or more times a week) reported more symptom gains than those who were least adherent but did not differ from those who were most adherent. In vivo adherence was not consistently associated with better outcome, perhaps due to heterogeneity in form and function of weekly assignments. Higher ratings of helpfulness of both types of homework predicted greater symptom improvement from pre- to posttreatment and between sessions. Overall, imaginal exposure homework may complement in-session exposures by enhancing key change processes, though perfect adherence is not necessary. Patients’ perceptions of helpfulness may reflect buy-in or perceived match between homework completion and functional impairment. Clinically, in addition to targeting adherence to homework assignments, querying about perceived helpfulness and adjusting assignments appropriately may help augment clinical gains.