A tale of 2 hospitals: a staggered cohort study of targeted interventions to improve compliance with antibiotic prophylaxis guidelines.

Background: The purpose of this prospective study was to determine the effectiveness of targeted interventions to improve compliance with antibiotic prophylaxis guidelines (timing, spectrum, and discontinuation) at 2 university-affiliated hospitals.
Methods: Based on barriers identified previously, hospital-specific interventions were developed such as educational conferences, standardized forms, an extended time-out, and feedback. Guideline compliance and surgical site infection (SSI) data were recorded on all patients who underwent elective laparotomies for colorectal procedures, vascular operations, and hysterectomies during four 6-month study periods. Prestudy data from July to December 2006 served as a baseline. One year later, a prospective cohort study was performed. The interventions were introduced to the 2 hospitals in a staggered fashion with 2-month implementation periods before reassessing compliance during the 6-month study periods. General linear modeling was performed (P < .05 significant).
Results: Compliance with all 3 guidelines combined improved during the year preceding the study, after attention only, at both hospitals. Hospital-specific differences were found in the effectiveness of the intervention package on individual guidelines. Hospital 2 but not 1 improved in timing after the interventions; both hospitals improved in spectrum, and neither hospital improved in discontinuation. Overall compliance with all 3 antibiotic prophylactic measures was greater at hospital 1, but hospital 2 had lower SSI rates.
Conclusion: Simply increasing attention to a quality problem can result in a significant and sustained improvement. Quality improvement interventions should be evaluated rigorously for effectiveness given hospital-specific differences in effectiveness and for correlation of guideline compliance with outcome.
(Copyright 2010 Mosby, Inc. All rights reserved.)