Despite guidelines recommending against the use of urine cultures to screen for asymptomatic bacteriuria and treatment with antibiotics, these practices are still very common. Investigators developed a quality improvement intervention that focuses on reducing the number of unnecessary urine cultures and antibiotic use.

The quality improvement study was performed at four intervention and comparison sites in the Veterans Affairs (VA) health care system. A coordinating center provided external facilitation, and each site had a site champion who served as an internal facilitator. The intervention encompassed an evidence-based algorithm and case-based education to train clinicians to use the algorithm, which guided them through two questions to ask themselves before ordering a urine culture or starting antibiotics when a urinary tract infection was suspected.

A total of 11,299 patients in the acute and long-term care units of the participating sites were included in the study. Of 12,260 urine cultures ordered, 5,867 were positive; 4,981 patients were treated, more than half with two or more antibiotics.

The reduction in the number of urine cultures ordered before and after the intervention wasn't significant at the intervention sites, but when these sites were compared with the comparison sites, there was a significant reduction in the number of urine cultures by 3.24 urine cultures per 1,000 bed-days in the intervention sites. The relative percentage decrease in days of antibiotic therapy at the intervention sites in the postintervention period was 21.7%, whereas days of antibiotic therapy at the comparison sites increased by 35.1%. Similarly, the relative percentage decrease in length of antibiotic therapy at the intervention sites was 21%, whereas length of antibiotic therapy increased in the comparison sites.

The investigators point out that data collection was limited because of the COVID-19 pandemic and that the implementation strategy may not be generalizable to non-VA facilities. In addition, the outcome measures didn't take into account whether a urine culture was appropriate.