Integrating the Electronic Laboratory Reporting of Methicillin-resistant Staphylococcus aureus (MRSA) Data

The development of algorithms for integrating ELR of MRSA can improve public health surveillance; reduce the burden on clinical, laboratory, and public health partners; and enhance public health action. The ELR from BioSense, as well as from national commercial ELR systems, can help develop and evaluate MRSA ELR algorithms.

Our objective was to examine variability among MRSA results in ELR systems from laboratories supplying data to BioSense and define algorithms for the integration of MRSA ELR data from hospitals and national commercial laboratory for case reporting to public health.

We analyzed ELR messages from more than 263,000 patients from 30 BioSense hospital data suppliers between 03/2004 and 12/2007 and more than 3,300,000 patient records from the national laboratory during 08/2007.

MRSA ELR results were reported in free-text format by eight hospital laboratories. The remaining 22 hospital sources and the national ELR sources reported results in coded format. We found 78.1% (1,684) MRSA ELR results were reported in coded elements. Accurate test names from MRSA patients were reported from 28 hospital laboratories (93.3%) and from the national laboratory. Local codes (coding format) for ELR, as well as targeted MRSA descriptors for free text format, were the most reliable elements of MRSA ELR.

We classified types of tests and responses associated with positive MRSA ELR findings, and developed an algorithm for the integration of incoming MRSA ELR results. Further work on the development of algorithms for integrating various ELR data sources will improve the public health reporting and follow-up of MRSA.