Despite American College of Chest Physicians (ACCP) guidelines, venous thromboembolism (VTE) prophylaxis rates remain unacceptably low. We sought to evaluate several approaches to routine ordering of thromboembolism prophylaxis to identify optimal ordering strategies and evaluate outcomes.
Computerized provider order entry (CPOE) has been used at Duke Hospital since 2004. In 2006, an HTML-based graphical user interface module with embedded decision support to guide appropriate ordering of thromboembolism prophylaxis was implemented, replacing the standard text-based (unguided) command line approach. We utilized DEDUCE, a Duke designed tool to query complex clinical and administrative databases at Duke hospital, to determine rates of VTE prophylaxis before (Phase I), 1 month after (Phase II), and 2 years following (Phase III) implementation of the module. At-risk medical and non-cardiac surgical patients per ACCP criteria were identified from a list of ICD-9 diagnosis codes using the Agency for Healthcare Research and Quality clinical classification bundling schema.
Approximately 2,000 medical and 2,000 non-cardiac surgical patients were identified for analysis. During Phase I, VTE prophylaxis rates were 65% for medical patients and 92% for surgical patients. This improved to 72% in medical patients and remained the same for surgical patients during Phase II (P < 0.0001 for medical patients). Two years later, prophylaxis rates were 75% for medical patients and 95% for surgical patients. A significant improvement in use of pharmacologic prophylaxis in the surgical group over the three phases was observed (52%, 60.5%, and 74% respectively; P < 0.0001) with concomitant reduction in sole use of mechanical prophylaxis (40.6%, 30.8%, and 21.0% respectively; P < 0.0001).
Even with a CPOE system, a well-designed interface with an embedded decision support module can significantly improve both rates of and types of VTE prophylaxis in at-risk patient populations. The relationships of optimized VTE prophylaxis with outcomes will also be reported.
Hospital systems should consider a decision support module within CPOE to improve both VTE prophylaxis rates and utilization of chemical prophylaxis.
Gregg Stashenko, Grant monies (from industry related sources) This research was sponsored by a grant from Sanofi Aventis to Dr. Tapson. Dr. Stashenko's salary support comes from an institutional NIH T32 grant.; No Product/Research Disclosure Information