Background: The Joint Commission (TJC) included medication reconciliation (MedRec) as a 2005 National Patient Safety Goal to reduce errors related to medication omissions, duplications and interactions. Medication errors and harms continue to be one of the most widely reported healthcare problems. TJC’s sentinel event database includes > 350 medication errors resulting in death or major injury; approximately half could have been avoided through effective MedRec. Despite years of attention, hospitals continue to struggle to implement successful MedRec. Similarly, we have identified high variability, risk of harm, and opportunities to systematically improve our institution’s MedRec.
Purpose: At our institution, we define admission MedRec (AdmMedRec) as complete only when an order action is taken on every single home medication. Our SMART aim is to decrease the rate of failure to complete hospital AdmMedRec within 24 hours across our system to < 5% in 12 months.
Description: Methods: Plan-Do-Study-Act (PDSA) Cycles were initiated in April 2017 with a pediatric hospitalist (PHM)/CMIO co-led interdisciplinary team of MDs, RNs, pharmacists, residents, Quality Management, and IT specialists across acute and subacute hospital sites in our large academic children’s hospital. (P): Failure Mode Effects Analysis, Process map, and Ishikawa diagrams were used to identify AdmMedRec failures. Baseline data from Nov 2016 – Jun 2017 were analyzed. (D) EMR workflow was redesigned along with the addition of AdmMedRec status column, MD admission checklist, and clarification of MD vs RN roles in MedRec. (S) Biweekly unit and service-specific data reviewed with team. (A) Cycle 2-4 interventions chosen by team: C2: PHM led resident and attending level trainings. C3: Simplified EMR build and workflow to enhance RN to MD communication on home medication status, made discontinued prior-to-admit medications no longer visible, obtained RN access to reconciliation of outside medications, and developed RN training material. C4: Best Practice Advisory (BPA) alert fires when ordering medication and AdmMedRec is not complete.
Results: AdmMedRec by hospital units (Fig 1) shows MEDICAL UNIT AdmMedRec failure rate ↓ from 0.31 to 0.14, PSYCH ↓ from 0.99 to 0.15, and HEME/ONC ↓ from 0.25 to 0.18. Fig 2 shows PHM AdmMedRec failure rate ↓ from 0.30 to 0.10 in the first 5 months of this project. Trends are noted on the graph.
Conclusions: Using an interdisciplinary team across all sites, this QI initiative was successful at ↑ AdmMedRec rates for all sites with greatest improvements for PHM and PSYCH. Both process (EMR) and culture (education) changes were integral to its success. Next cycle includes sharing provider – specific performance rates and engaging unit mavens to decrease AdmMedRec failure rate to the < 5% goal by June 2018. Additionally, we would like to reduce time to complete AdmMedRec to 12 hours from admission, expand to include MedRec at transfer and on discharge, and evaluate the quality of completed MedRec.