Document Type

Master's Culminating Experience

Publication Date



Background: Methicillin-resistant Staphylococcus aureus (MRSA)-specific hospital discharge codes were introduced in October 2008. The purpose of this project is to study the correlation of these codes with laboratory-proven pediatric MRSA bloodstream infections.

Methods: Laboratory and discharge databases were used to identify patientsChildren’s Hospital Medical Center from October 2008 through December 2010 with MRSA bloodstream infections. The laboratory database identified patients with a positive blood culture for MRSA, and the discharge database identified patients with an MRSA bloodstream discharge code (038.12 for MRSA sepsis or 041.12 for MRSA infection with either 790.7 for bacteremia or 771.81 for newborn sepsis or 771.83 for newborn bacteremia). The sensitivity and positive predictive value (PPV) of the codes were determined, using laboratory-confirmed infection as the gold standard.

Results: During the 27 month study period, 65 patients with MRSA bloodstream infection were identified from laboratory data and 58 patients were identified from administrative data; 36 were concordant. The overall sensitivity was 55.4%, and the overall PPV was 62.1%, and neither showed a significant trend over time. The sensitivity varied with the clinical manifestation of the bloodstream infection. Of the 22 patients coded for MRSA bloodstream infection without a positive blood culture, almost half (45.4%) had only a localized MRSA infection. Of the 29 patients with positive MRSA blood culture but insufficient discharge coding, the majority (55.2%) were coded only for MRSA infection without bloodstream involvement.

Conclusions: Discharge coding for MRSA bloodstream infection was found to be insensitive and to have a low PPV. Using these codes alone will lead to an underestimation of pediatric MRSA bloodstream infections.