Response to “Potential Misclassification of Urinary Tract Related Bacteremia Upon Applying the 2015 Catheter-Associated Urinary Tract Infection Surveillance Definition From the National Healthcare Safety Network”

To The Editor—In their concise communication, “Potential Misclassification of Urinary Tract Related Bacteremia Upon Applying the 2015 Catheter-Associated Urinary Tract Infection Surveillance Definition From the National Healthcare Safety Network,” ¹ Greene et al present findings from their retrospective review of cases they define as urinary tract–related bloodstream infection at 3 VA hospitals. The authors report that among 145 cases with documented indwelling urinary catheters, 93 cases (64.1%) would be deemed catheter-associated urinary tract infections (CAUTIs) according to the updated 2015 National Healthcare Safety Network (NHSN) criteria. The authors conclude that applying those criteria, specifically the criterion introduced in 2015 that requires a urine culture bacterial count of at least 1×10⁵ colony-forming units (CFU/mL), would lead to under-ascertainment of clinically meaningful CAUTIs.

The authors’ concern that “the new CDC surveillance definition has the potential to underestimate the burden of CAUTI-related illness” and that “this has the potential to undermine the faith that clinicians have in the reliability of the national surveillance system for CAUTI,” must be balanced by what was previously an even larger “disconnect” and widely shared concerns about a lack of clinical credibility. A study comparing clinical CAUTI determinations to earlier NHSN definitions indicated that the NHSN definitions had a positive predictive value of only 35% compared to Infectious Diseases consultant evaluation. ² While the CAUTI surveillance criteria that the CDC introduced in 2015 may omit some infections that are deemed clinically significant, the magnitude of missed cases associated with bloodstream infections likely is small, and this possible shortcoming should be placed in the larger context of criteria changes that improve the specificity of case findings. An analysis of NHSN data, completed as part of the review of predecessor CAUTI criteria, showed that the CFU/mL threshold change would result in 10% fewer reported CAUTIs; only 0.5% of all CAUTIs with a reported secondary bloodstream infection would not be reported as a result of the change, much lower than that identified by Green et al.³ Furthermore, many of the “secondary bloodstream infections” previously attributed to the urinary tract may still (and perhaps more appropriately) be captured in NHSN surveillance as central-line–associated bloodstream infections.

The authors suggest “moving toward a clinician-based approach for diagnosis” as an alternative to the current NHSN criteria. Clinical and surveillance definitions serve different purposes. Clinical definitions of UTI are used by clinicians for individual patient diagnosis, involve clinical judgment and consideration of all relevant clinical data, and guide treatment decisions. In contrast, surveillance definitions are aimed at populations, with the purposes of measuring disease incidence and outcomes and evaluating the impact of prevention activities. To be applicable and informative in multiple settings, surveillance definitions must be based on a circumscribed set of data. Because these definitions are used for performance measurement, the CDC is obligated to select laboratory tests and test result criteria that can be applied as widely as possible. A recent survey by the Association for Professionals in Infection Control and Epidemiology provides evidence of the variability of laboratory standards for working up and reporting urine specimens. ³ Until a standard clinical definition is available that can be operationalized for use in surveillance and widely adopted, the NHSN must continue to provide CAUTI criteria that enable intra- and inter-facility comparisons, and serve as a meaningful catalyst for infection prevention in US healthcare facilities.