Electronic prediction rules for methicillin-resistant Staphylococcus aureus colonization.

Area of Science:

• Infectious Disease Epidemiology
• Health Informatics
• Clinical Prediction Modeling

Background:

• Hospital resources are significantly allocated to minimize methicillin-resistant Staphylococcus aureus (MRSA) infections.
• Current MRSA colonization surveillance is costly, with false positives leading to unnecessary patient isolation.
• Targeting high-risk patients can improve the efficiency of MRSA surveillance.

Purpose of the Study:

• To develop and validate electronic prediction rules for identifying patients at high risk for MRSA colonization upon hospital admission.
• To assess the performance of these rules in a real-world hospital setting.

Main Methods:

• Five prediction rules of varying complexity were derived using prospectively collected electronic health record data from 23,314 patients.
• Rules incorporated demographic, admission, pharmacologic, laboratory, physiologic, and historical variables.
• The rules were validated in a separate cohort of 26,650 patients across two additional hospitals.

Main Results:

• MRSA prevalence was 2.2% in the derivation cohort and 4.0% in the validation cohort.
• Multivariable modeling identified key predictors of MRSA colonization.
• The prediction rules successfully identified the top 30% of patients accounting for over 60% of MRSA colonization cases and ~70% of MRSA-associated patient-days.

Conclusions:

• Electronic prediction rules can automate patient triage for MRSA testing upon hospital admission.
• These rules offer significant improvements over existing methods, potentially saving costs for infection control programs.
• The developed rules provide an efficient and effective approach to MRSA surveillance.