Diagnostic Testing Stewardship in Urine Testing: Impact of Changes in Testing Practice on Culture Rates and Outcomes

摘要

Urinalysis and urine culture are frequently ordered diagnostic tests among hospitalized patients, often for non-specific symptoms such as fever and leukocytosis rather than specific urinary tract infection (UTI) symptoms. Testing for these broad indications can result in inappropriate antibiotic prescription for asymptomatic patients. To address this problem, hospitals have implemented a variety of interventions, one of which is the use of diagnostic testing stewardship which optimizes the process of ordering, performing and reporting diagnostic tests. An example is urinalysis with reflex to culture (UARC), a practice which has been shown to reduce institutional culture rates by selecting patients who are more likely to have true infection. Optimal reflex criteria are not well established however, and downstream effects such as the impact on antibiotic prescription are less well studied.The present study utilized an interrupted time series analysis with negative binomial regression to assess the impact of a two-part diagnostic stewardship intervention at an academic medical center, namely, changing the UARC reflex criteria and the addition of the need to specify an indication for isolated urine cultures (UC). The interventions took place in October 2019 and November 2019 respectively, and data from December 2018-October 2020 were included. Our primary study outcomes were urine culture rates, from both tests ordered as UARC and isolated UC, and downstream clinical use of antibiotic prescriptions for suspected UTI and catheter-associated urinary tract infections (CAUTI). Adverse events that may have resulted from this intervention, in particular adverse outcomes that may have resulted from missed UTIs, were assessed by chart review of post-intervention period for sepsis secondary to an undiagnosed urinary tract infection.After the intervention, there was a significant decrease in the rate of urine cultures performed from UARC (32.5 cultures/1,000 patient days pre- vs. 8.67 cultures/1,000 patient days post-intervention, p<0.001) but not in isolated UC culture rates (11.2 cultures/1,000 patient days pre- vs. 7.75 cultures/1,000 patient days post-intervention, p = 0.17). Culture positivity from tests ordered as an UARC significantly increased (34.7% pre- vs. 62.1% post-intervention, p=0.01), but not from tests ordered as isolated UC (26.9% pre- vs. 26.7% post-intervention, p=0.99). There was a significant decrease in antibiotic prescription rates (20.5 antibiotic prescriptions/1,000 patient days pre- vs. 14.2 antibiotic prescriptions/1,000 patient days post-intervention, p=0.047), with a significant pre-intervention monthly decrease. There was a non-significant change in CAUTI rates (2.7 CAUTIs/1,000 urinary catheter days pre- vs. 0.6 CAUTIs/1,000 urinary catheter days post-intervention, p=0.29). The intervention was associated with low rate of adverse events (1 case of potential sepsis in 130 reviewed charts).In conclusion, the changing of UARC reflex criteria to a stricter cut-off resulted in significant decreases in urine culture rates from tests ordered as an UARC without significant adverse events seen, supporting the use of this criteria in institutional urine testing algorithms. The addition of indication selection for isolated UC did not impact utilization of this test, with many providers still inappropriately utilizing this test. While antibiotic prescriptions for suspected UTIs decreased, this was primarily pre-intervention, suggesting that this intervention, while decreasing rates of culture performance, may not have impacted antibiotic prescription.