Discussion

CAUTI prevention is a universal need for both acute and long-term care facilities. The diagnostic stewardship of urine specimens has become an increasingly recognized element of CAUTI prevention, ^{Reference Coffey, Claeys and Morgan3} yet there are few published examples of clinical decision support tools. Trautner and colleagues ^{Reference Trautner, Grigoryan and Petersen2} described a diagnostic algorithm that reduced urine specimen use but did not address catheter exchange before obtaining a specimen. ^{Reference Raz, Schiller and Nicolle4} Mullin and colleagues ^{Reference Mullin, Kovacs and Fatica5} showed that limiting catheterized urine specimens to specific populations reduced urine culture use and CAUTI, but similarly did not address catheter exchange and was limited to fever workup of critically ill patients only.

In this study, we report a significant and durable improvement in CAUTI and a corresponding improvement in the key process measures targeted by the algorithm - urine specimen use and exchange of indwelling urethral catheters prior to obtaining urine specimens. The improvement in these process measures suggests that the improvement in CAUTI did not occur by chance but more likely as a result of true changes in practice brought on by the algorithm.

Indwelling urethral catheter utilization decreased during the study period despite no changes being made to policies around catheter utilization. This could be a result of increased scrutiny of catheters in general due to our intervention. While reducing catheter utilization is a well-recognized CAUTI prevention strategy, the effect on CAUTI rates can be unpredictable and conversely may increase rates by decreasing the denominator.

The ideal time point after which an indwelling urethral catheter ought to be exchanged prior to obtaining a urine culture is unclear. We used a threshold of 48 hours to place a high priority on avoiding contaminated specimens at the cost of performing more exchanges. Others have studied 7 days ^{Reference Yarrington, Reynolds and Dunkerson6} or 14 days ^{Reference Hooton, Bradley and Cardenas7} as thresholds for catheter exchange but a timeframe as short as 24 hours has shown benefit. ^{Reference Frontera, Wang and Phillips8} We achieved 39% compliance with catheter exchange; the remaining 61% may have been missed due to provider judgment, patient preference, or other clinical contraindications. Our approach to catheter exchange did not result in any reported harms and was sustainable for nursing staff.

The limitations of this study include its quasi-experimental nature and the potential influence of other unmeasured factors affecting CAUTI. Our preimplementation period coincided with the onset of the COVID-19 pandemic which has been associated with increased healthcare-associated infection rates, however internal data shows that CAUTI rates at our institution did not meaningfully change after onset of the COVID-19 pandemic (1.35 vs 1.39 CAUTIs per 1,000 catheter-days in calendar years 2019 and 2020, respectively). We did not collect data on patient-level demographics, comorbidities, or trends in antibiotic prescribing. We did not use a “wash-out” period after implementation, but this omission would have only biased our results toward the null. Finally our study was performed at an academic medical center and may not be generalizable to other settings.