BACKGROUND: CDC’s National Healthcare Safety Network (NHSN), used by CDC and its partners for surveillance of healthcare-associated infections (HAIs), provides standardized infection ratios (SIRs) to participating hospitals to help promote healthcare quality. In 2011, the NHSN began reporting SIRs measuring central line-associated bloodstream infection (CLABSI) incidence to the Centers for Medicaid and Medicare Services (CMS) to evaluate hospital reporting and performance. A CLABSI SIR is the ratio of the observed to expected CLABSI incidence.
METHODS: The expected CLABSI incidence was estimated using a log linear regression model based on hospital and unit-level factors. Bayesian hierarchical random effects model was used to adjust SIRs resulting in Adjusted Ranking Metric (ARM) values.
RESULTS: In 2011, there were 62,637 months of data summarizing CLABSI incidence reported to NHSN from 3,326 hospitals. Hospital-level CLABSI SIRs are prone to excessive variation depending on exposure volume and unmeasured factors. Bayesian analysis was used to produce ARM values that account for these sources of variability and adjust for differences reliability.
CONCLUSIONS: The ARM obtained by shrinking the observed CLABSI incidence toward the overall predicted incidence given exposure volume and relevant factors is more comparable and stable to better measure quality performance.