Effect of urine reflex culturing on rates of cultures and infections in acute and long-term care
JournalAntimicrobial resistance and infection control
MetadataShow full item record
AbstractBackground: Urine cultures are often positive in the absence of a urinary tract infection (UTI). Pyuria is generally considered necessary to diagnose a UTI. Problem: Urine cultures are often positive in the absence of UTI leading to unnecessary antibiotics. Methods: Quasi-experimental pre-post study of all patient urine cultures ordered in a VA acute care hospital, emergency department (ED), and two long-term care (LTC) facilities from August 2016 to August 2018. Urine cultures performed per 100 days were compared pre- (August 2016 to July 2017) versus post-intervention (August 2017 to August 2018) using interrupted time series negative binomial regression. Intervention: We examined whether reflexing to urine culture only if a urinalysis (UA) found greater than 10 WBC/hpf decreased urine culturing. Results: In acute-care, reflex culturing resulted in a 39% time series regression analysis adjusted decrease in the rate of cultures performed (pre-intervention, 3.6 cultures/100 days vs. Post-intervention, 1.8 cultures/100 days, p < 0.001). Pre-intervention, 29% (4/14) of Catheter-associated UTI (CAUTI) would not have been reported if reflex culturing was employed. In the ED, reflex culturing was associated with a 38% (p = 0.0015) regression analysis adjusted decrease in cultures, from 5.4/100 visits to 3.3/100 visits. In LTC, there was a small absolute, but regression analysis adjusted increase of 89% (p = 0.0018) in rates from (0.4/100 days to 0.5/100 days). Conclusion: In acute care and ED, urine reflex culturing decreased the number of urine cultures performed. A small absolute increase was seen between pre-post time periods in LTC. Reflex testing generally decreases cultures and may lead to more accurate diagnoses of CAUTI.
Identifier to cite or link to this itemhttps://www.scopus.com/inward/record.uri?eid=2-s2.0-85087398270&doi=10.1186%2fs13756-020-00762-1&partnerID=40&md5=ab4539029b59ac1b525be1ef475c3242; http://hdl.handle.net/10713/13299