Correlation of venous thromboembolism prophylaxis and electronic medical record alerts with incidence among surgical patients.

Abstract

Venous thromboembolism events are potentially preventable adverse events. We investigated the effect of interruptions and delays in pharmacologic prophylaxis on venous thromboembolism incidence. Additionally, we evaluated the utility of electronic medical record alerts for venous thromboembolism prophylaxis. Venous thromboembolisms were identified in surgical patients retrospectively through Core Measure Venous ThromboEmbolism-6-6 and Patient Safety Indicator 12 between November 2013 and March 2015. Venous thromboembolism pharmacologic prophylaxis and prescriber response to electronic medical record alerts were recorded prospectively. Prophylaxis was categorized as continuous, delayed, interrupted, other, and none. Among 10,318 surgical admissions, there were 131 venous thromboembolisms; 23.7% of the venous thromboembolisms occurred with optimal continuous prophylaxis. Prophylaxis, length of stay, age, and transfer from another hospital were associated with increased venous thromboembolism incidence. Compared with continuous prophylaxis, interruptions were associated with 3 times greater odds of venous thromboembolism. Delays were associated with 2 times greater odds of venous thromboembolism. Electronic medical record alerts occurred in 45.7% of the encounters and were associated with a 2-fold increased venous thromboembolism incidence. Focus groups revealed procedures as the main contributor to interruptions, and workflow disruption as the main limitation of the electronic medical record alerts. Multidisciplinary strategies to decrease delays and interruptions in venous thromboembolism prophylaxis and optimization of electronic medical record tools for prophylaxis may help decrease rates of preventable venous thromboembolism.