1023 — Results from a cluster randomized trial of two implementation strategies to deliver audit and feedback in the EQUIPPED medication safety program
Lead/Presenter: E. Camille Vaughan,
All Authors: Vaughan EC (Birmingham/Atlanta GRECC, Emory University), Burningham Z (Salt Lake City VA IDEAS COIN, University of Utah) Kelleher JL (Birmingham/Atlanta GRECC) McGwin G (Birmingham/Atlanta GRECC, University of Alabama at Birmingham) Jasien C (Atlanta VA) Hastings SN (Durham VA ADAPT COIN, Duke University) Stevens MB (Birmingham/Atlanta GRECC, Emory University) Morris I (Durham VA ADAPT COIN) Jackson GL (Durham VA ADAPT COIN, Duke University) EQUIPPED VA implementation QI group
EQUIPPED is a quality improvement program focused on enhancing prescribing safety toward older adults at the time of ED discharge. EQUIPPED implementation involves three core components including provider education, clinical decision support, and monthly prescribing audit and feedback using the American Geriatrics Society Beers Criteria to determine prescribing rates of potentially inappropriate medications (PIMs). This study evaluated whether implementation of audit and feedback through a centralized informatics-based dashboard was associated with improvement in prescribing as the personnel intensive academic detailing delivered one-on-one by an EQUIPPED champion.
In a cluster randomized implementation study (October 2019-September 2021), eight VA EDs were randomized to either the Academic Detailing (n = 4) or Dashboard (n = 4) strategy for the audit and feedback component of EQUIPPED. One of the eight sites withdrew during the COVID-19 pandemic and another site was recruited that implemented EQUIPPED October 2020â€“July 2022. The primary outcome was the monthly proportion of PIMs prescribed to Veterans 65 years or older and discharged from the ED. Poisson regression was used to evaluate the number of PIMs prescribed 6 months prior to EQUIPPED implementation compared to 12 months following the EQUIPPED intervention. Group assignment and the total number of prescriptions served as the offset term in the model and data are presented as proportions.
Eight VA ED sites successfully implemented provider education, at least 3 EQUIPPED order sets, and monthly provider feedback for 12 months. During the 6-month baseline period, the Academic Detailing and Dashboard sites had similar PIM prescribing rates [(7.96% (1,511 PIMs/18,990 prescriptions, Academic Detailing) vs. 8.04% (2,166 PIMs/26,936 prescriptions, Dashboard) (p = 0.75)]. Comparing 12 months of prescribing data after implementing EQUIPPED, the Academic Detailing group significantly improved PIM prescribing (7.10%, 1,605 PIMs/22,597 prescriptions) compared to the Dashboard group, (8.10%, 2,979 PIMs/36,795 prescriptions) [p = < 0.0001]. Within the groups, 3 of the 4 Academic Detailing sites demonstrated statistically significant reductions in PIM prescribing. One of the 4 Dashboard sites achieved nearly 50% relative reduction in PIM prescribing and this site also had the highest proportion of prescribers access their individualized dashboard report.
Eight VA EDâ€™s were able to successfully implement the core components of the EQUIPPED program amid the unprecedented challenges posed by the COVID-19 pandemic. The academic detailing approach to audit and feedback was more effective than the dashboard-based approach; however, the EQUIPPED dashboard site with the highest provider dashboard engagement showed substantial reduction in PIM prescribing toward older Veterans at the time of ED discharge. Additional evaluation will determine if organizational factors predict which ED sites would benefit from the academic detailing vs. dashboard-based audit and feedback approach to guide resource allocation for dissemination of EQUIPPED across VHA.
EQUIPPED implementation including audit and feedback with academic detailing is associated with consistent improvement in medication safety toward older Veterans discharged from VA EDâ€™s. A centralized approach to provide electronic dashboard-based audit and feedback may be feasible in some ED settings and additional research will determine if baseline organizational factors predict the best implementation approach.