188

GG Kaplan, D Bates, D Macdonald, J Romagnuolo

Department of Medicine (Division of Gastroenterology), University of Calgary, Calgary, Alberta

BACKGROUND & AIMS: Indications for intravenous proton pump inhibitors (IV PPI) include 1) reduction of rebleeding in upper gastrointestinal bleeding (UGIB) from peptic ulcer disease with high risk stigmata (HRS) and 2) PPI delivery to “NPO” patients. Evidence that IV PPI are being prescribed inappropriately with costly implications has been previously provided by our group. The aim of this study was to assess whether interventions could be instituted to improve inappropriate utilization and costs of IV PPI (pantoprazole).

METHODS: Prospective study of all patients prescribed either bolus or infusion of IV PPI in a 60-day period before and after implementation of interventions. Patients were divided into 2 groups based on the setting in which the IV PPI was used: UGIB or non-UGIB. The interventions included educating physicians (newsletter), instituting a dose template for ordering IV PPI in our computerized inpatient order system, allowing pharmacists to review and alter IV PPI orders in non-UGIB patients (autosubstitutions if receiving oral food/meds), and recommending a GI consult (to confirm HRS) for infusion orders. Proportion of prescribing errors and total IV PPI costs were compared between the two groups, as were balance measures (e.g. number of UGIBs, total admissions to hospital, and GI consults). The Chi-square test was used.

RESULTS: The post-intervention (n=61) group (vs. the pre- group; n=67) showed a significant absolute reduction in the rates of inappropriate indication in both UGIB (26%; 95%CI [10–42%]; p<0.0001) and in non-UGIB (17% [0–39%]; p<0.0001). There was a 15% [2–27%] absolute increase in endoscopy prior to IV PPI (to confirm HRS) for UGIB, and a 29% [16–42%] increase in infusions rather than boluses for UGIB. However, a greater improvement in underspending than overspending errors, meant that no overall IV PPI cost difference was seen between the 2 groups. The post-intervention period and the control period were similar in the total number of UGIBs (117 vs 106) and of admissions to hospital (5884 vs 5739). Total number of GI consultations is being tallied.

CONCLUSION: In a tertiary center, relatively simple maneuvers resulted in significant improvements in IV PPI prescribing behavior. Overspending behaviors appear more difficult to change than underspending ones, resulting in a lower ability to achieve anticipated cost savings.

NEXT ABSTRACT >