Implementation of an evidence-based stress ulcer prophylaxis protocol in the ICU

Abstract

Stress ulcer prophylaxis is one of the most common medications used in the intensive care setting. Despite well-established risk factors and guidelines, most critical care physicians use prophylaxis therapy inappropriately. In the absence of a standard of care, we evaluated the efficacy, safety, and potential cost savings of an evidence-based stress ulcer prophylaxis protocol in the surgical ICU (SICU). We conducted a prospective, clinical study at a tertiary care teaching hospital in Chicago. One hundred and nine patients admitted to the SICU were screened daily for 12 risk factors of stress ulceration and followed until discharge from the unit. Patients with at least one identified risk factor were started on famotidine. Prophylaxis was discontinued when risk factors were no longer present and/or the patient tolerated at least 50% of his/her daily caloric need. Data collected included patient demographics, daily risk factor assessment, number of doses received, nutritional intake, estimated creatinine clearance, and adverse outcomes. Comparison data was collected for the 109 patients admitted to the SICU prior to the adoption of our protocol. Of the 109 patients meeting inclusion criteria, 99 required stress ulcer prophylaxis per our protocol. The most common risk factors were major cardiovascular or abdominal operation within 48 hours, mechanical ventilation greater than 48 hours, and intolerance to enteral feeding. A comparison of pre-protocol versus post-protocol implementation revealed an average SICU length of stay of 5.2 vs 8 days (P = 0.003), an average duration of prophylaxis therapy of 5 vs 4 days (P = 0.02), and an average number of doses administered per patient of 8.6 vs 5.7 (P = 0.001), respectively. This translated into a cost difference per patient of $4.53 vs $3.84 (P = 0.02). Efficacy and safety analysis revealed six adverse drug reactions, with one treatment failure documented by a clinically significant bleed requiring transfusion (0.92% incidence), and five incidents of thrombocytopenia (0.01% incidence) for which famotidine could not be ruled out as a causative agent. These findings are consistent with previously established adverse reactions associated with histamine-2 receptor blockers. Pantoprazole was subsequently initiated in all six cases. Our midpoint data analysis suggests that our evidence-based protocol is a safe and efficacious tool for the standardization of stress ulcer prophylaxis in an ICU. Although potential cost savings may be minimal, the reduction of unnecessary medication administration, prevention of potential drug-drug interactions, and recognition of patient risk factors through the use of a stress ulcer prophylaxis protocol are achieved.