2173. Surgical Site Infection Determination in Epic® ICON: A Utilization Model
Session: Poster Abstract Session: Healthcare Epidemiology: HAI Surveillance
Saturday, October 6, 2018
Room: S Poster Hall
Posters
  • Sarah2.pdf (255.8 kB)
  • Background:

    Prior to 2016, our hospital used microbiology results alone to investigate surgical site infections (SSI). Previous studies show that this practice can miss as many as half of clinically significant infections. To improve accuracy for fiscal year 2016 SSI surveillance was done by manual chart review of 100% of the surgeries we report to NHSN. While more accurate, this process was time and labor intensive. In May 2016, we began using Epic® ICON as our data mining software. ICON can abstract (create denominator data), determine SSI status (create numerator data) and upload to NSHN. Data indicates that partially automated SSI surveillance reduce manual chart review but our team found that many charts were being reviewed unnecessarily. We developed a computerized algorithm within ICON that would that would capture SSIs but limit the number of charts to be reviewed.

    Methods:

    Algorithm variables within Epic® ICON were modified to limit data collection to the following parameters: readmission, chief complaint, surgical log, diagnosis, antibiotic administration post 48 hours, and specific microbiology results. We excluded 31 keywords that were part of the Epic® ICON foundation system from our algorithm. For example, we removed the keyword “infection” which flagged whenever “no infection” was charted. The chief complaints grouper was most important as it allowed only meaningful complaints to be considered. Microbiology results were also limited to only include Aerobic, Anaerobic, Fungus, AFB, and wound cultures. To validate the algorithm, it was run retrospectively for fiscal year 2016.

    Results:

    There was 100% concordance of results comparing SSIs identified using chart review to the use of our computerized algorithm and Table 1 shows the average number of charts requiring review pre and post implementation.

    Table 1: Comparison of Charts Requiring Review

    Monthly Average with Chart Review

    Monthly Average with Computerized Algorithm

    Improved Efficiency

    BRST

    63

    27

    57%

    COLO

    43

    21

    52%

    HYST

    38

    7

    82%

    HPRO

    72

    30

    59%

    KPRO

    59

    27

    55%

    CBGB

    30

    5

    83%

    Conclusion:

    Careful modification of the ICON foundations system resulted in a 55% decrease overall in the need for chart review without affecting accuracy of reporting.

    Sarah Elizabeth Totten, Dr.PH, M(ASCP), Infection Control, University of Colorado Hospital, Aurora, CO, Aurora, CO, Shane Hansen, MT(ASCP)SC, MSM(IT), Epic Inpatient Orders, Memorial Hospital, part of University of Colorado Health, Colorado Springs, CO, Michelle Barron, MD, Internal Medicine/Infectious Diseases, University of Colorado Denver, Aurora, CO and Larissa Pisney, MD, MSc, Division of Infectious Diseases, University of Colorado School of Medicine, Aurora, CO

    Disclosures:

    S. E. Totten, None

    S. Hansen, None

    M. Barron, None

    L. Pisney, None

    Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 3rd with the exception of research findings presented at the IDWeek press conferences.