The network approach for prevention of healthcare-associated infections: long-term effect of participation in the Duke Infection Control Outreach Network

Abstract

Objective: To describe the rates of several key outcomes and healthcare-associated infections (HAIs) among hospitals that participated in the Duke Infection Control Outreach Network (DICON).

Design and setting: Prospective, observational cohort study of patients admitted to 24 community hospitals from 2003 through 2009.

Methods: The following data were collected and analyzed: incidence of central line-associated bloodstream infections (CLABSIs), ventilator-associated pneumonia (VAP), catheter-associated urinary tract infections (CAUTIs), and HAIs caused by methicillin-resistant Staphylococcus aureus (MRSA); employee exposures to bloodborne pathogens (EBBPs); physician EBBPs; patient-days; central line-days; ventilator-days; and urinary catheter-days. Poisson regression was used to determine whether incidence rates of these HAIs and exposures changed during the first 5 and 7 years of participation in DICON; nonrandom clustering of each outcome was controlled for. Cost saved and lives saved were calculated on the basis of published estimates.

Results: In total, we analyzed 6.5 million patient-days, 4,783 EBPPs, 2,948 HAIs due to MRSA, and 2,076 device-related infections. Rates of employee EBBPs, HAIs due to MRSA, and device-related infections decreased significantly during the first 5 years of participation in DICON (P< .05 for all models; average decrease was approximately 50%); in contrast, physician EBBPs remained unchanged. In aggregate, 210 CLABSIs, 312 cases of VAP, 332 CAUTIs, 1,042 HAIs due to MRSA, and 1,016 employee EBBPs were prevented. Each hospital saved approximately $100,000 per year of participation, and collectively the hospitals may have prevented 52-105 deaths from CLABSI or VAP. The 7-year analysis demonstrated that these trends continued with further participation.

Conclusions: Hospitals with long-term participation in an infection control network decreased rates of significant HAIs by approximately 50%, decreased costs, and saved lives.