2017 HSR&D/QUERI National Conference

4072 — Improving Medication Reconciliation in a VA Nursing Home as part of MARQUIS2

Lead/Presenter: Amy Baughman
All Authors: Baughman AW (VA Boston Healthcare System (VABHS)), Murphy N, Driscoll L, Norstrom J, Lange K, Hanson A, Shahood B, Ruopp M, Mixon A, Simon SR

Objectives:
Accurate, efficient medication reconciliation (med rec) is a known safety concern for healthcare professionals caring for patients with polypharmacy, multimorbidity, and cognitive impairment. Specific Aims: 1. Improve med rec by implementing best practices in a nursing home. 2. Evaluate a novel software tool. 3. Improve clinician experience.

Methods:
Setting: VA Boston's Community Living Center (Brockton, MA) is a 110-bed nursing home that provides transitional (skilled or rehabilitation), palliative/hospice and long-term care. Patients are medically complex and often have limited medication knowledge. The interdisciplinary quality improvement (QI) team includes pharmacists, nurses, physicians, nurse practitioners, physician assistants, information technologists and QI experts. Intervention: This study spans October 2016 - June 2018 as part of MARQUIS2 (Multi-Center Medication Reconciliation Quality Improvement Study 2). Baseline assessments included policy review, provider survey, and process mapping of all current med rec processes. In the next phase, we will implement interventions based on identified challenges and established best practices. These include pharmacist-driven intensive med rec for high-risk patients and the Avicenna MedRec Tool, proprietary software to overcome existing barriers in our current electronic medical record.

Results:
Primary outcome: unintentional medication discrepancies, tracked from a random sample of 16 patients per month by study pharmacists. We will also evaluate clinician satisfaction with med rec processes as we test various interventions. In our baseline survey, 11 providers (response rate 100 %) reported spending approximately 30 minutes on each admission and discharge reconciliation. Most (7/11) lacked confidence in the accuracy of the med rec process. Yet the majority identified med rec as important. Few received formal education (2/11), incentives (1/11) or resources (2/11) to address medication discrepancies. After 4 months of baseline data collection, 66 patients have been reviewed. On admission, there were 38 patients with unintentional discrepancies with a total of 104 discrepancies: 89 of these were history errors and 15 were med rec errors.

Implications:
1) Medication reconciliation is time consuming and error-prone in the current state. 2) Stakeholder surveys and process maps will inform targeted interventions by providing detailed information on work flow and practice variation.

Impacts:
Nursing homes represent a critical setting for medication reconciliation efforts due to the often challenging but clinically important determination of medication lists.