1076 — An Agent-Based Modeling Approach to Surveillance of Catheter-Related Bloodstream Infections
Rubin MA (VA Salt Lake City Health Care System), Mayer J
(VA Salt Lake City HCS), Greene T
(VA Salt Lake City HCS), Sauer B
(VA Salt Lake City HCS), Trick W
(Rush University, Chicago), Hota B
(Rush University, Chicago), Samore MH
(VA Salt Lake City HCS)
Public disclosure of nosocomial infection (NI) rates is gaining momentum across the United States. Traditional surveillance for NI is hindered by the fact that National Healthcare Safety Network (NHSN) criteria are complex and subjective, raising concerns about the reliability of these criteria. Simplified, objective criteria based on microbiologic data may be a less valid, but more reliable system for comparing institutional infection rates.
We created an agent-based model to simulate the occurrence of catheter-related bloodstream infections (CRBSI) in a 12-bed hospital intensive care unit. Traditional (clinical-based) surveillance was performed by simulated surveyors applying NHSN clinical criteria; their reliability at interpreting subjective criteria was modeled using signal detection theory concepts. Algorithmic surveillance was performed by applying simplified criteria to microbiologic data based on previously published work. Surveillance methods were modeled under the assumption that the simple criteria would have inferior specificity to clinical criteria. Reliability of the simulated surveyors was assessed by altering their accuracy and their variance in specificity in a sensitivity analysis. Preservation of rank order was assessed using Kendall's Tau method.
As per model design, clinical criteria demonstrated a higher average specificity than simple criteria (92.2% vs. 81.5%, respectively). However, as rater accuracy and variance in specificity were varied, and across a plausible range of inter-rater reliabilities (mean 0.53, range 0.33-0.72), the accuracy of clinical criteria approached, and was often worse than, simple criteria. In addition, under the majority of conditions and scenarios, ecologic correlation (i.e., the accurate ranking of CRBSI rates across institutions) was higher for simple criteria than clinical criteria, even in situations where individual accuracy was higher for clinical criteria.
Simplified CRBSI surveillance criteria, which remove subjectivity, are more reliable than traditional clinical criteria at accurately identifying the true differences in CRBSI rates between institutions. The more reliable simple criteria are most likely better suited for publicly reporting institutional CRBSI rates for comparative purposes.
For the purpose of public reporting of NI rates, states and healthcare institutions should strongly consider using simplified, objective surveillance criteria to improve the reliability and comparability of estimated infection rates across institutions.