1084. Estimated Burden of Methicillin-Resistant Staphylococcus aureus (MRSA) in California Hospitals following Changes to Administrative Codes, 2005-2010
Session: Poster Abstract Session: Surveillance of HAIs: Implementation and National Perspectives
Friday, October 4, 2013
Room: The Moscone Center: Poster Hall C
  • IDWeek Poster_Trends in MRSA Burden in CA Hospitals.pdf (907.8 kB)
  • Background: MRSA rates have previously been estimated from ICD-9 codes that have shown poor accuracy for identifying MRSA. In 2008, revised ICD-9 codes were instituted to improve capture of MRSA colonization and infection.

    Methods: 2005-2010 mandatory California (CA) hospital discharge data were reviewed for ICD-9 codes indicating MRSA infection, colonization, and personal history. Trends in carriage and infection rates were compared before (pre-2008) and after (post-2008) ICD-9 revisions using segmental regression analysis. Infections were stratified by community- and hospital-onset. Consistency of MRSA colonization and personal history codes were assessed across serial admissions.

    Results: Among 340 CA hospitals and 17,354,517 adult admissions, 15.8 per 1,000 admissions had any of the MRSA codes pre-2008 and 23.7 per 1,000 admissions had a code post-2008, with an increase post-coding change (0.3 cases per 1,000 admissions, p<0.01).  Total MRSA infections declined from 15.8 to 14.6 cases per 1,000 admissions (-0.3 cases per 1,000 admissions, p<0.001), including declines in both pneumonia and sepsis (Figure). MRSA infection codes were rapidly adopted with a one-to-one correlation, while newly instituted MRSA colonization codes rose steadily. Hospital-onset MRSA infections of pneumonia (-0.03 cases per 1,000 admissions, p<0.0001), septicemia (-0.01 cases per 1,000 admissions, p<0.010, and unspecified infection (-0.02 cases per 1,000 admissions, p<0.05) all decreased post-coding change. Among MRSA infected patients, only 8.5% had a MRSA personal history code upon subsequent readmission. Among patients with either a MRSA colonization or personal history code, only 22.0% maintained either of the codes on next admission.

    Conclusion: Revised ICD-9 codes were rapidly adopted, identifying an increasing number of patients with MRSA carriage and a decreasing number with MRSA infection. However, due to poor tracking of colonization codes between hospitalizations, coding remains an exceedingly poor method for identifying MRSA carriage. It is unknown whether accounting for any presence of MRSA codes across serial hospitalizations will remedy this.


    David M. Tehrani, MS, Chenghau Cao, MPH, Homin Kwark and Susan S. Huang, MD, MPH, FIDSA, Division of Infectious Diseases and Health Policy Research Institute, University of California Irvine School of Medicine, Irvine, CA


    D. M. Tehrani, None

    C. Cao, None

    H. Kwark, None

    S. S. Huang, None

    Findings in the abstracts are embargoed until 12:01 a.m. PST, Oct. 2nd with the exception of research findings presented at the IDWeek press conferences.