From the 2018 HVPAA National Conference
Joseph Meserve (University of California San Diego), Janet Ma (University of California San Diego), Andrew Nguyen (University of California San Diego), Amie Nguyen (University of California San Diego), Arkady Komsoukaniants (University of California San Diego), Gregory Seymann (University of California San Diego)
Anticoagulants and antiplatelet medications confer increased risk for gastrointestinal bleeding. Patients on triple antithrombotic therapy (dual antiplatelet therapy plus oral anticoagulants) are at particularly high risk, and current Cardiology and Gastroenterology guidelines recommend prophylactic proton pump inhibitor (PPI) use in this patient population. Studies suggest that this practice is underutilized.
To assess the incidence of PPI initiation for GI bleed prophylaxis among hospitalized patients started on triple antithrombotic therapy at a tertiary-care medical center
We used the electronic medical record to identify a cohort of patients who were discharged alive from our hospital over a 6 month period between 06/2017 – 12/2017 . Patients were included if they had a combination of aspirin and clopidogrel on their medication list at time of discharge, in addition to any of the following anticoagulants: coumadin, rivaroxaban, apixaban, or dabigatran. The primary endpoint was rate of PPI use. Secondary endpoint was rate of GI bleeding after triple antithrombotic initiation.
We identified 50 patients discharged on triple therapy; 11 of these were on a PPI prior to admission. Of the 39 patients not previously on a PPI, only 4 (10.3%) had PPI initiated with triple therapy. Overall, 70% (35/50) of this cohort did not receive guideline directed initiation of PPI and only 8% (4/50) started PPI for the specific indication of GI prophylaxis due to triple therapy. Notably, 10% (5/50) of these patients were readmitted with a GI bleed within 12 months from discharge with no deaths attributable to GI bleeds.
Using local data to compute mean costs per DRG for GI hemorrhage and mean PPI costs, we extrapolated savings with appropriate PPI prophylaxis as high as $116,000 annually at our institution, assuming 100% of the admissions for GI bleeding were preventable. Using a more conservative estimate of 45% reduction in bleeding, potential cost savings would approach $52,200 annually. These values reflect an underestimate of the impact, as outpatient costs, morbidity and mortality are not considered.
Rates of prophylactic PPI use for patients on anticoagulants were low at our institution, and unfavorable outcomes of GI bleeding were high following initiation of triple therapy. We hypothesize that reasons for underutilization may include unfounded concern for P2Y12 and PPI interactions, low awareness of guideline recommendations, and lack of clinical decision support.
Implications for the Patient
Increasing adherence to best practices that minimize the risk of anticoagulants and antiplatelet medications should be an area of focus for high-value care initiatives. We are developing an intervention with our cardiology and pharmacy colleagues for introduction of decision support and pharmacy oversight to improve the reliable use of prophylaxis.