Prospective evaluation of a clinical guideline recommending hospital length of stay in upper gastrointestinal tract hemorrhage

Context.-Upper gastrointestinal tract hemorrhage (UGIH) is a common and potentially life-threatening disorder. Resource utilization can vary without adverse effect on patient outcome. Clinical practice guidelines are a potential solution to reduce variation in practice while improving patient outcomes. Objective.-To validate prospectively the safety, acceptability, and impact of a clinical practice guideline defining the medically appropriate length of stay (LOS) for patients hospitalized with UGIH. Design.-Prospective, controlled time-series study with an alternate-month design. Outcome surveyors and patients were blinded to study group allocation. Guideline.-A retrospectively validated scoring system using 4 independent variables: hemodynamics, time from bleeding, comorbidity, and esophagogastroduodenoscopy (EGD) findings to predict risk of adverse events. The quantitative risk for the low-risk subset was 0.6% (95% confidence interval [CI], 0.0%-2.0%) for subsequent complications and 0% (95% CI, 0.0%-0.9%) for life-threatening complications from this retrospective evaluation. Setting.-A 1000-bed, not-for-profit, university-affiliate teaching hospital. Patients.-Consecutive adult patients hospitalized for acute UGIH. Intervention.-Concurrent feedback of guideline recommendation (same-day hospital discharge) to physicians caring for patients at low risk for complication. No risk information was provided during control months. Results.-Seventy percent (209/299) of UGIH patients achieved low-risk status according to the guideline and were therefore potentially suitable for early discharge from the hospital. Providing real-time quantitative risk information (intervention group only) was associated with an increase in guideline compliance from 30% to 70% (P<.001) and a decrease in mean (SD) LOS from 4.6 (3.5) days to 2.9 (1.3) days (mean reduction of 1.7 days per patient; P<.001). No differences in complications, patient health status, or patient satisfaction were found when measured 1 month after discharge. An independent variable predicting decreased hospital LOS for low-risk UGIH patients was early EGD. Conclusions.-Implementation of the clinical practice guideline safely reduced hospital LOS for selected low-risk patients with acute UGIH. Further prospective validation in other settings is warranted.