Search | Search by Center | Search by Source | Keywords in Title
Williams LS, Ofner S, Yu Z, Beyth RJ, Plue L, Damush T. Pre-post evaluation of automated reminders may improve detection and management of post-stroke depression. Journal of general internal medicine. 2011 Aug 1; 26(8):852-7.
BACKGROUND: Post-stroke depression (PSD) occurs in at least one-third of stroke survivors, is associated with worse functional outcomes and increased mortality, and is frequently underdiagnosed and undertreated. OBJECTIVE: To evaluate the effectiveness of an electronic medical record-based system intervention to improve the proportion of veterans screened and treated for PSD. DESIGN: Quasi-experimental study comparing PSD screening and treatment among veterans receiving post-stroke outpatient care one year prior to the intervention (the control group) to those receiving outpatient care during the intervention period (the intervention group); contemporaneous data from non-study sites included to assess temporal trends in depression diagnosis and treatment. PARTICIPANTS: Veterans hospitalized for ischemic stroke and/or receiving primary care (PC) or neurology outpatient follow-up within six months post-stroke at two (Veterans Affairs) VA Medical Centers. INTERVENTIONS: We formed clinical improvement teams at both sites. Teams developed PSD screening and treatment reminders and designed tailored implementation strategies for reminder use in PC and neurology clinics. MAIN MEASURES: Proportion screened for PSD within 6 months post-stroke; proportion screening positive for PSD who received an appropriate treatment action within 6 months post-stroke. KEY RESULTS: In unadjusted analyses, PSD screening was performed within 6 months for 85% of intervention (N = 278) vs. 50% of control (N = 374) patients (OR 6.2 , p < 0.001), and treatment action was received by 83% of intervention vs. 73% of control patients who screened positive (OR 1.8 p = 0.13). After adjusting for intervention, site and number of follow-up visits, intervention patients were more likely to be screened (OR 4.8, p < 0.001) and to receive a treatment action if screened positive (OR 2.45, p = 0.05). Analyses of temporal trends in non-study sites revealed no trend toward general increase in PSD detection or treatment. CONCLUSIONS: Automated depression screening in primary and specialty care can improve detection and treatment of PSD.