Changing urine culture order test indications during a switch to a computerized physician order entry (CPOE) sharply reduced unnecessary cultures and saved considerable costs in lab expenses, researchers report. Moreover, it spared patients with asymptomatic bacteriuria from treatment more appropriate for a fully diagnosed urinary tract infection (UTI).
The researchers compared urine culture rates before the CPOE switch with the rates after the new electronic order system was in place. CPOE changes that included a clarification of test names and resetting reflex tests — which prompt orders for further testing — “resulted in a 45% reduction in the urine cultures ordered with an estimated cost savings of $104,000,” the authors reported.1
“During the study period, 18,954 inpatients … had 24,569 urine cultures ordered. Overall, 6,662 urine cultures (27%) were positive,” they found.
The urine culturing rate decreased from 38.1 per 1,000 patient days preintervention to 20.9 per 1,000 patient days after the intervention, explains author David K. Warren, MD, MPH, hospital epidemiologist at Barnes-Jewish Hospital in St. Louis.
“It was a pretty substantial reduction — it was what we were hoping,” he says. “We were concerned about the amount of unnecessary urine culturing that was being done. It was a robust drop.”
Warren and colleagues looked for any unintended consequences of the intervention to ensure those who needed culture and treatment were not being missed.
“Infection prevention tracks CAUTIs [catheter-associated urinary tract infections] housewide,” he explains.
“We looked at the infection rate for CAUTIs, and it actually didn’t change. So we were still capturing [CAUTIs], but at the same time, we were reducing the number of urine cultures that were being performed.”
The researchers looked at the proportion of positive urine cultures that were associated with an abnormal or positive urinalysis as well as cultures associated with a negative urinalysis.
“We saw that post-intervention, there were a larger number of positive urine cultures that were associated with an abnormal urinalysis compared to the pre-intervention period,” Warren says.
“That is more evidence that we are still detecting those people that have abnormal urinalysis associated with their urine culture.”
A prior analysis in the hospital showed that serial testing of patients admitted without a diagnosis of a UTI yielded 8.5% with asymptomatic bacteriuria.
“They had positive urine culture without any evidence of infection,” he says.
“Finding patients that have bacteriuria in their urine leads to all the downstream problems in terms of antibiotic treatment, increases the risk of resistant bacteria, and could lead to C. diff infection. That was part of the rationale for trying to cut down on urine culturing, especially in the area where the pre-test probability was low.”
In resetting testing algorithms, the “bar was set low” for neutropenic patients so they would not be missed if they had a UTI. Any suspect indications in these patients triggers reflex testing to urine microscopy and urine culture, he explains.
Although the effort streamlined orders to some extent and removed defaults that could lead to overtesting, physicians still make the call if they want a test.
“Urine culture was still an orderable test for clinicians who wanted to order it by themselves. They were free to do so,” he says. “We didn’t restrict that.”
One take-home message from the study is a transition to electronic ordering can be an opportunity to improve practice — in this case, in the area of diagnostic stewardship.
“Be mindful when you are designing those systems of how they can have a big impact on reducing unnecessary testing,” Warren says.
- Munigala S, Rojek R, Wood H, et al. Effect of changing urine testing orderables and clinician order sets on inpatient urine culture testing: Analysis from a large academic medical center. Infect Control Hosp Epidemiol 2019;40(3): 281-286.