Funguria is a common finding in hospitalized patients with indwelling urinary catheters.Reference Richards, Edwards, Culver and Gaynes 1 However, funguria infrequently represents true infection and rarely leads to fungemia,Reference Kauffman, Vazquez and Sobel 2 and treatment of patients with urinary catheters who have funguria usually provides no clinical benefit.Reference Sobel, Kauffman and McKinsey 3 Despite the fact that true urinary tract infections due to yeast are rare, many patients with funguria meet current National Healthcare Safety Network (NHSN) surveillance criteria for catheter-associated urinary tract infection (CAUTI). 4 These episodes of clinically insignificant funguria are inappropriately labeled as CAUTIs, and this miscategorization leads to overestimation of CAUTI rates.
Several experts have proposed the exclusion of funguria from the current NHSN surveillance definition for CAUTIs. However, the impact of such a proposed change on rates of reported CAUTIs has not been extensively studied. We examined surveillance data from a network of community hospitals as well as a tertiary-care medical center to describe the potential impact of excluding yeast as a urinary pathogen from the CAUTI definition on CAUTI rates.
Methods
We performed a retrospective analysis of prospectively collected CAUTI surveillance data from 30 community hospitals participating in the Duke Infection Control Outreach Network (DICON) and Duke University Hospital (DUH), a 900-bed academic, tertiary-care medical center. DICON is a network of 41 community hospitals in 5 states throughout the southeastern United States that has been described previously.Reference Anderson, Miller and Chen 5 Trained infection preventionists at each hospital enter surveillance data into a standardized database using NHSN definitions. 4 DICON and DUH infection preventionists validate a subset of surveillance data each month. Of the DICON hospitals, 11 had incomplete CAUTI surveillance data during the study period and were excluded from the analysis. CAUTI rates were evaluated separately in DICON hospitals and DUH because of differences in patient case mix and catheter utilization.
We compared CAUTI rates with and without yeast isolates from surveillance data between January 1, 2013, and December 31, 2013, in several ways. Overall CAUTI rates including and excluding yeast species were calculated for DICON hospitals as a whole and then separately for each DICON hospital. The preceding analyses were repeated for the 2012 calendar year for comparison. Finally, we compared CAUTI rates in patients at DUH by first including then excluding yeast isolates for the 2012 and 2013 calendar years.
Results
Complete surveillance data on CAUTI were available from patients in 30 acute-care, community hospitals in DICON between January 1, 2013, and December 31, 2013. A total of 313 CAUTIs were observed during 271,118 urinary catheter-days (1.15 CAUTIs per 1,000 urinary catheter-days). Yeast species were the cause of 70 (22%) of these infections and were the second most common CAUTI pathogen after E. coli, which caused 83 of these infections (27%). Removal of the yeast CAUTIs reduced the CAUTI rate to 0.90 CAUTIs per 1,000 urinary catheter-days and yielded an absolute incidence rate reduction of 25%. The mean CAUTI rate and the change in CAUTI rate excluding yeast varied among DICON hospitals (Table 1). The overall mean CAUTI rate per DICON hospital when yeast isolates were included was 1.26 CAUTIs per 1,000 urinary catheter-days (95% confidence interval [CI], 0.84–2.09). When yeast isolates were excluded, the overall mean CAUTI rate per DICON hospital decreased by 20% to 1.01 CAUTIs per 1,000 urinary catheter-days (95% CI, 0.67–1.34). Similarly, the overall CAUTI rate in 2012 dropped 22% when yeast isolates were excluded.
Mean Catheter-Associated Urinary Tract Infection (CAUTI) Rates Including and Excluding Yeast as a Urinary Pathogen in 30 Community Hospitals in 2013
A total of 203 CAUTIs were observed throughout DUH over 56,806 urinary catheter-days (3.57 CAUTIs per 1,000 urinary catheter-days) during the 2013 calendar year. Removal of 56 yeast CAUTIs reduced the CAUTI rate by 28% to 2.59 CAUTIs per 1,000 urinary catheter-days. The CAUTI rate at DUH in 2012 dropped 26% when yeast isolates were excluded.
Discussion
We found that urinary isolates of yeasts that meet NHSN criteria for CAUTI are common. Exclusion of yeast from the CAUTI definition decreased both community and tertiary-care hospital CAUTI rates by nearly 25%. We believe that CAUTI rates would decrease by a similar magnitude if other acute care hospitals that have yeast CAUTIs used a modified CAUTI definition that excluded yeast.
There are several important and foreseeable implications of excluding funguria from NHSN definitions. First, CAUTI rates excluding funguria cannot be directly compared to historical CAUTI data that included funguria. Similarly, standardized incidence ratios determined with and without funguria cannot be directly compared. Thus, use of different definitions will invariably impact existing interhospital comparisons or rankings. Second, excluding funguria will reduce rates of CAUTIs from all reporting institutions that have yeast CAUTIs. Additionally, without inclusion of funguria, each unit change in CAUTI events (the numerator) or in urinary catheter-days (the denominator) will have more quantitative impact on the resultant CAUTI rate than before the definition change. Thus, CAUTI rates and interhospital performance rankings using CAUTI rates will be more susceptible to period-to-period fluctuations in CAUTI events and urinary catheter-days.
In summary, excluding funguria from the CAUTI definition will likely reduce the overall reported CAUTI rates by at least 20% in hospitals that have yeast CAUTIs. Yeast in the urine rarely represents true infection; thus, excluding yeast from the CAUTI definition will lead to a more clinically relevant metric. At the same time, limitations of reporting CAUTI rates as a performance measure without also reporting the absolute number of events will be amplified by excluding yeast from the CAUTI definition. Reporting the number of CAUTI events along with the CAUTI rate will best marry clinical relevance with infection surveillance.
Acknowledgments
Financial support: DJA was supported by NIAID/NIH (K23AI095357).
Potential conflicts of interest: All authors report no conflicts of interest relevant to this article.
