2.1 Design
This qualitative study is part of an ongoing comprehensive project examining the process of improving the delivery of evidence-based PTSD care. This early work using grounded theory methodology15 explored prescribing clinicians’ perceptions of adherence to the CPG recommendation regarding prescribing of benzodiazepines to patients with PTSD through analysis of interviews with primary care (PC) and mental health (MH) clinicians. Our primary aims were to identify facilitators and barriers to PTSD guideline adherence regarding benzodiazepine prescribing to aid in the development of supportive strategies to de-implement this practice and improve the care of patients at increased risk of harms.
2.2 Participants and setting
To recruit participants, we selected VA Medical Centers based on hospital-level benzodiazepine data for patients with PTSD in fiscal years 1999- 2009 based on our larger work.5 We grouped all VA medical centers based on the degree of change in benzodiazepine prescribing rates in context to their initial frequency of prescribing into categories. Within each category 3 facilities were selected, considering geographic variance and number of veterans with PTSD treated. We had previously shown that rural patients treated at VA facilities had an increased likelihood of receiving benzodiazepines,16 so we selected at least one hospital in each group that was considered rural. We also noted geographic variation in prescribing benzodiazepines, so we selected hospitals to represent specific geographic regions. In total 12 VA Medical Centers and 2 smaller rural Community Based Outpatient Clinics were selected for inclusion in the study. Providers from 1 hospital did not respond to recruitment emails, resulting in 13 hospitals in the final sample. We asked MH leadership from selected hospitals to recommend the names of prescribing clinicians who treat PTSD in their general mental health (MH) and primary care (PC) clinics. We then randomly selected clinicians (1 from MH and 1 PC) from each hospital to participate. In total, we contacted 42 participants by email to recruit for participation, 8 failed to respond and 8 declined given time constraints, resulting in a total of 26 clinicians for a 68% participation rate.