BACKGROUND: At one institution, a clinical decision support (CDS) alert for venous thromboembolism (VTE) prophylaxis burdened providers but was considered vital to patient safety. Electronic clinical quality measures (eCQMs) incentivized the translation of quality measures into data elements within the electronic health record (EHR) and facilitated hospitalwide performance monitoring during CDS improvement. The aim was to reduce VTE alerts by 50% without compromising eCQM performance. METHODS: This quality improvement initiative was performed at a tertiary care academic medical center using an integrated EHR. Alert firings were revised in three rounds over a four-week transition period while monitoring VTE eCQM performance weekly. Postimplementation data were recorded for 12 weeks. Primary outcomes were VTE alerts per 100 admissions and VTE eCQM performance. Secondary outcomes were alert effectiveness (desired responses/patients), alert efficiency (desired responses/alerts), and dwell time (time between alert firing and provider addressing the alert). RESULTS: Alerts decreased from 157 to 74 per 100 admissions, a 52.9% reduction (p < 0.001). There was no change in eCQM compliance or the percentage of inpatients excluded from the VTE eCQM. Provider dwell time across the hospital dropped between 2.9 and 7.2 hours per day. After the interventions, alert effectiveness increased (66.1% to 73.3%; p < 0.001), but alert efficiency decreased (17.5% to 16.2%; p = 0.007) due to an increase in providers delaying definitive responses. CONCLUSION: Altering VTE alert criteria did not affect compliance with providing VTE prophylaxis to patients while reducing alert burden by more than 50%. Using preexisting quality data like eCQMs can facilitate near-time patient safety monitoring during quality improvement projects.
BACKGROUND: At one institution, a clinical decision support (CDS) alert for venous thromboembolism (VTE) prophylaxis burdened providers but was considered vital to patient safety. Electronic clinical quality measures (eCQMs) incentivized the translation of quality measures into data elements within the electronic health record (EHR) and facilitated hospitalwide performance monitoring during CDS improvement. The aim was to reduce VTE alerts by 50% without compromising eCQM performance. METHODS: This quality improvement initiative was performed at a tertiary care academic medical center using an integrated EHR. Alert firings were revised in three rounds over a four-week transition period while monitoring VTE eCQM performance weekly. Postimplementation data were recorded for 12 weeks. Primary outcomes were VTE alerts per 100 admissions and VTE eCQM performance. Secondary outcomes were alert effectiveness (desired responses/patients), alert efficiency (desired responses/alerts), and dwell time (time between alert firing and provider addressing the alert). RESULTS: Alerts decreased from 157 to 74 per 100 admissions, a 52.9% reduction (p < 0.001). There was no change in eCQM compliance or the percentage of inpatients excluded from the VTE eCQM. Provider dwell time across the hospital dropped between 2.9 and 7.2 hours per day. After the interventions, alert effectiveness increased (66.1% to 73.3%; p < 0.001), but alert efficiency decreased (17.5% to 16.2%; p = 0.007) due to an increase in providers delaying definitive responses. CONCLUSION: Altering VTE alert criteria did not affect compliance with providing VTE prophylaxis to patients while reducing alert burden by more than 50%. Using preexisting quality data like eCQMs can facilitate near-time patient safety monitoring during quality improvement projects.