CDS EHR Integration Reduces Inappropriate Urine Culture Orders

A clinical decision support (CDS) EHR integration helped significantly reduce inappropriate urine culture (UC) orders among hospitalized patients in the country’s large safety-net hospital system, according to a study published in the American Journal of Infection Control (AJIC).

Asymptomatic bacteriuria (ASB), the isolation of bacteria without symptoms or signs of a urinary tract infection (UTI), is overdiagnosed nationwide. Overtreatment of ASB leads to harm, including adverse effects from antibiotics, antibiotic resistance, and increased length of stay.

The researchers implemented a quality improvement project to decrease unnecessary urine culture orders across 11 acute care hospitals.

The initiative was two-fold. First, the researchers changed the EHR ordering screen for urine cultures to require the entry of an indication for the culture.

Next, they created a best practice advisory (BPA) that appeared when clinicians ordered urine cultures for patients with urinary catheters hospitalized for more than 48 hours. The CDS alert reminded providers of the high rate of ASB in patients with foley catheters.

The study found that inpatient UCs decreased by 20.9 percent post-intervention, and inpatient UCs on patients with urinary catheters fell by 21.6 percent. Outpatient UCs were unchanged.

The authors noted that while other studies have shown larger reductions in UCs, the interventions often were multifaceted and required reflex testing, education, cognitive aids, and stewardship involving specialists.

“In our case, we successfully reduced UC with a low-effort, EHR-focused intervention, adding to prior literature demonstrating effectiveness of mandating indications prior to ordering,” they wrote. “More significantly, this was scaled across 11 hospitals in the largest safety net system in the United States, modeling scale and generalizability in a low-resourced setting.”

The research revealed that reductions in urine culture ordering rates were highly variable among the 11 hospitals. The study authors observed low BPA acceptance rates in internal medicine and family medicine clinicians. Additionally, attendings showed high acceptance of the BPA compared to other clinicians.

“Variation in behavior change among individual hospitals and clinician types and specialties warrant further study,” the authors wrote.

The researchers also noted that the CDS design could generalize to other hospital-acquired infections, such as Clostridioides difficile, or low-value testing, such as stool cultures for hospital-acquired diarrhea.

There were several limitations to the study. First, the researchers noted that they did not include randomization or a control arm.

Additionally, they did not assess appropriateness or indications of urine culture pre- and postintervention.

The researchers also noted that the intervention occurred during the COVID-19 pandemic, which may limit generalizability to non-pandemic settings.