The standardized infection ratio (SIR) is an indirectly standardized morbidity ratio that has been used to compare the infection rate in a hospital with an expected number of infections from a national standard and is being increasingly promoted as a metric for the public reporting of healthcare-associated infections (HAIs).
To identify potential discrepancies between SIR and other measures of risk.
Hypothetical and real data were compared using relative risk, a directly standardized morbidity ratio, and SIR values across a range of varying hospital population compositions.
In real and hypothetical data, other summary statistics were consistent with each other and with underlying HAI incidence density rates. However, use of the SIR frequently led to conclusions inconsistent with these other inherently unbiased estimators.
Because of a recognized type of distortion inherent in the calculation of indirectly standardized ratios, use of the SIR can lead to conclusions that differ from those reached when using other traditional measures of risk and to incorrect assessments of conclusions about the performance of hospitals or states. In addition, the tendency to inappropriately arrange SIR values in rank order for comparison makes SIR unsuitable as a statewide metric for monitoring HAIs.