Modeling the Cost Effectiveness of Facility-wide Active Surveillance for Methicillin-resistant Staphylococcus aureus (MRSA) in Acute Care Hospitals

Objectives: Our goal was to determine the thresholds of prevalence levels of MRSA in acute care hospitals that make universal surveillance (i.e. admission culture of all patients) cost effective.

Methods: We developed a computer simulation model representing the decision to perform universal surveillance versus no surveillance using TreeAge Pro 2008. A model outlining the choices and assumptions made by hospital administration when designing an MRSA surveillance program was created using data from prior studies. Inputs included sensitivity and specificity of surveillance, MRSA prevalence, R₀, probabilities of medical outcomes of MRSA and all associated costs. Sensitivity analyses were performed by systematically varying the prevalence of MRSA and R₀ since these variables are unknown. 100,000 Monte Carlo simulations were performed for each pairing of variables.

Results: Performing universal MRSA surveillance was cost-effective [incremental cost-effectiveness ratio (ICER) <$50,000 per quality-adjusted life year (QALY)] at the following MRSA R₀ and prevalence combinations: R₀ ³0.25 and prevalence ³ 0.01. In fact, surveillance was the dominant strategy when R₀ =1.5 and prevalence ≥ 0.15; R₀ =2.0 and prevalence ³ 0.10; R₀ ³2.5 and prevalence ≥ 0.05.

Conclusions: MRSA surveillance of medical hospital admissions appears to be cost-effective at a wide-range of prevalence and R₀ values. Hospital administrators need realistic estimates of these variables in their facility to design cost-effective surveillance.

Implications for Public Health: All hospitals should consider performing universal MRSA surveillance based on our model. Further research to determine realistic estimates of R₀ and prevalence can assist in creating better public health surveillance policy.