Statewide one-day survey of central lines Impact of different denominators on public reporting of infection rates

Purpose-This paper aims to determine whether the rank order of hospitals changes when their central line-associated bloodstream infection (CLABSI) rate is computed using a traditional proxy measure for the denominator (number of patients with one or more catheter in place) versus using the actual number of catheters or catheter-lumens. Design/methodology/approach-The authors conducted a statewide voluntary one-day prevalence survey among all hospitals participating in Washington State's mandatory public reporting program. Hospitals counted the number of catheters and catheter-lumens as well as patients with catheters. Counts of patients with one or more catheter in place, of catheters, and of catheter-lumens were extracted from each hospital's completed survey form and transformed into a ratio. Three CLABSI incidence density rates were computed for each hospital by scaling their annual CLABSI rate in the previous calendar year by the ratio of patients to catheters to catheter-lumens. Influence of these three different denominators on rank order of the hospitals was assessed by scaling the corresponding Centers for Disease Control and Prevention's National Healthcare Safety Network incidence density rates for each participating hospital and examining position shifts with the Wilcoxon signed rank test. Findings-Statistically significant but only modest shifts in position became evident, which did not correlate with service complexity characteristics of the hospitals affected. Originality/value-Others have shown that the CLABSI incidence density rate in a single hospital is significantly affected by switching from a traditional proxy measure denominator to a more meaningful denominator. This is the first report on whether all hospitals' rates would be affected in a uniform or a non-uniform manner if a different denominator were to be selected by mandatory public reporting programs.