1769. Assessing The Impact of The National Healthcare Safety Network’s (NHSN’s) New Baseline on Acute Care Hospital Standardized Infection Ratios (SIRs)
Session: Oral Abstract Session: National Trends in HAIs
Saturday, October 7, 2017: 8:45 AM
Room: 08

Background: To more accurately measure the progress of healthcare-associated infection (HAI) prevention efforts, the CDC’s National Healthcare Safety Network (NHSN) surveillance system updated risk-adjustment models for computation of updated Standardized Infection Ratios (SIRs), the primary HAI summary measure by NHSN. This study sought to examine how the updated SIRs varied from the previous SIRs calculated using older baselines for acute care hospital HAIs.

Methods:  We analyzed NHSN data for healthcare facility-onset laboratory-identified Clostridium difficile [CDI] and methicillin-resistant Staphylococcus aureus [MRSA] bacteremia reported in accordance with the CMS’ inpatient quality reporting program requirement. The unit of analysis was CMS certification number (CCN) facility reporting in 2015. We compared overall distributions of CCN-level SIRs (CCN-SIRs) between new risk-adjustment models using a 2015 baseline (SIR_NEW) and old models using a 2011 baseline (SIR_OLD) and tested location shift (median away from null) of pairwise differences. We also examined the magnitude of shift in SIR from old to new baseline.

Results:  For each HAI, the national pooled mean SIR of the new baseline was ~1.0.  For CDI, the overall distributions of CCN SIR_NEW and CCN-SIR_OLD were different, and the median of pairwise difference was away from null with CCN-SIR_NEW slightly higher.  For MRSA, the SIR differences were not significant. Most CCN-SIRs (83% for CDI, 93% for MRSA) remained in the same significance category across the old and new baselines (‘worse’, ‘better’, ‘not different from national benchmark’), and few CCN-SIRs were reclassified to a less favorable category.  For 75% of CCN-SIRs, their relative position in the quartile distributions of SIR_NEW and SIR_OLD remained the same. The discrepancies between SIR_NEW and SIR_OLD tended to be larger among CCNs with high SIRs.

Conclusion:  The updated national pooled mean SIRs were close to 1.0, validating the potential use of new risk adjustment models and baseline as updated benchmarks for tracking CDI and MRSA prevention progress. The shifts in CCN-level SIRs between old and new baselines were not large, indicating a modest impact of new baselines at the CCN level, except among hospitals with high SIRs.


Minn Soe, MBBS, MPH1, Allan Nkwata, MPH1, Jonathan R. Edwards, MStat2, Margaret Dudeck, MPH, CPH1 and Daniel Pollock, MD2, (1)Centers for Disease Control and Prevention, Atlanta, GA, (2)Division of Healthcare Quality Promotion, Centers for Disease Control and Prevention, Atlanta, GA


M. Soe, None

A. Nkwata, None

J. R. Edwards, None

M. Dudeck, None

D. Pollock, None

Findings in the abstracts are embargoed until 12:01 a.m. PDT, Wednesday Oct. 4th with the exception of research findings presented at the IDWeek press conferences.