OBJECTIVES To understand benefits and risks of backing out of an alert for warfarin administration when high International Normalized Ratio (INR). Adverse drug events cause clinically significant adverse outcomes and complications 1-3. Analyses of a national surveillance study of showed that there were more than 175,000 visits for adverse drug events yearly, among them warfarin topped the list as the most commonly implicated medications. Such events are a huge economic burden to the nation impacting patients in a myriad different ways. Therefore, it is widely thought that mitigating warfarin related complications could help decrease medical expenses. METHODS We developed an alert for our computerized pharmacy order system for all patients being treated with warfarin. The system generated an alert during the ordering of warfarin if the INR was ≥ 3.5. Upon encountering an alert providers had options such as back out from an order, or override, or override with some action. From June 1, 2007, to November 30, 2008, our system had 12,041 inpatient encounters and received 21,234 orders for warfarin. We had 820 alerts for a total number of 635 encounters. The mean warfarin dose was 4.76 mg. The mean INR for the 635 encounters with alert was 4.24 with a standard deviation of 1.69. RESULTS Average length of stay, average cost per case and complications were significantly lower among patients whose providers cancelled warfarin orders than patients who received by ignoring the alert. In addition, providers who backed out from giving warfarin were more likely to prescribe vitamin k which may have helped mitigate complications. There was no statistical significance in the hematoma related complications (1.7% versus 2.0%) between backed out and overridden groups. CONCLUSIONS Further investigation with additional factors is required before attributing the reduction in length of stay to backing out of an alert.