PURPOSE: Physicians whose high risk patients for venous thromboembolism (VTE) who were not receiving prophylaxis were randomized either to receive an electronic alert or to be controls with no alert. The alerted physicians’ patients had a 41% reduction in symptomatic deep vein thrombosis and pulmonary embolism [NEJM 2005;352:969]. To determine whether this trial had a long-lasting impact after publication, we performed a cohort study of 411 patients in which all physicians were alerted, without a control group, if their high-risk patients were not prophylaxed.
METHODS: The computer calculated VTE risk scores based upon 8 common VTE risk factors, each weighted according to a point scale. “High risk” was defined as > 4 score points. Major risk factors (cancer, prior VTE, and hypercoagulability) each scored 3 points, major surgery 2 points, and age, obesity, bed rest, and hormone therapy scored 1 point each. We compared alerts following study publication to alerts from our published randomized trial.
RESULTS: The percentage of high risk patients not receiving VTE prophylaxis decreased from 18% in the randomized trial to 7% in the cohort study (p<0.001). Most high-risk patients not receiving prophylaxis were Medical Service patients (80%). Prescribing of prophylaxis in response to the alerts increased to 40% in the cohort compared with 34% in the randomized (p=0.01). Pharmacologic prophylaxis with unfractionated heparin decreased (17% to 8%) and use of low molecular weight heparin increased (4.4% to 14%) in the cohort. Overall VTE rates remained unchanged, although proximal DVT decreased (0.8% to 0.2%) and distal DVT increased (0.4% to 1.0%) in the cohort.
CONCLUSION: Continued use of an electronic alert system, tailored to detect high VTE risk patients not receiving prophylaxis, can increase utilization of VTE prophylaxis.
CLINICAL IMPLICATIONS: The implementation of an electronic alert system to encourage VTE prophylaxis among high-risk hospitalized patients maintains efficacy in reducing rates of DVT and PE through improved prescribing patterns.
DISCLOSURE: Steven Baroletti, None.