Advice worth repeating: Avoid duplicate labs

An electronic best practice alert reduced unnecessary repetition of labs.


Background

The idea to reduce repetitive lab testing at the University of Florida came not from doctors but from patients, including one health professional who pointed out that his HbA1c was ordered on the wards even though the test had already been done in the ED, said nephrology fellow Harini P. Bejjanki, MD, ACP Member. “Errors like these gave birth to the Reducing Lab Costs project,” she said, adding that before the intervention, there was no way for physicians to know the status of a lab test that's still in process.

To coordinate the project, Dr. Bejjanki formed a team including the laboratory director, three other hospitalists, the chief of hospital medicine, and information technology (IT) staff, who came up with the idea for a best practice alert in the electronic health record. “When one physician orders a test, the other one needs to know that this test was already ordered,” she said. “We should avoid poking the patient multiple times for the same test.”

How it works

Members of the team identified 17 tests that they believed to be frequently duplicated and applied the alert to each. To determine when a test could appropriately be repeated (from one day for complete blood count [CBC] with differential to three months for HbA1c), the team used best practice guidelines and sought expert opinion from the hospital's chiefs of various divisions, Dr. Bejjanki said.

In September 2015, the alert went into practice for all inpatient units. The way it works is simple: If a recent order for the same lab test already exists, a pop-up message informs the ordering physician and provides the order time, collected time, result time, and the result itself (if applicable). Physicians are unable to sign the orders until they have reviewed the popup, but they also have the choice to override recommendations, Dr. Bejjanki said. “You can't take away that clinical decision making from a physician,” she said. “There has to be a balance where we respect people but also try to deliver outcomes that are good for everyone, for the patient and the hospital.”

Results

The team measured the percentage of test duplicates for all 17 tests before and after the intervention. After the intervention, from October 2015 to February 2017, duplicates for all lab tests decreased by an average of 15.4% compared to the preintervention period, according to results published in October 2018 by Clinico Economics and Outcomes Research.

The greatest change was seen in duplicate orders for vitamin D (−58.7%), vitamin B12 (−49.6%), and respiratory viral panel nasopharynx swab (−33.4%). Duplicate orders decreased for all tests but two: CBC with differential, which increased by 3.3%, and serum C-reactive protein (CRP), which increased by 7.8%. Cost-savings calculations showed that the intervention saved about $72,543 by reducing the number of duplicate tests.

Challenges

The biggest challenge was driving physicians to change practice, as evidenced by the increase in CRP test duplicates in particular. “Some people believe in testing it more frequently, especially the infectious disease specialists, and some people think once a week is fine. Our window period was once a week,” Dr. Bejjanki said. “Despite our best efforts and getting expert opinions and guidelines, there was still some difference in opinion as to when that particular test should be repeated to monitor for improvement.”

After the study, the team conducted some educational sessions for hospitalists and residents to review various evidence-based guidelines on repeat testing from organizations such as Choosing Wisely and ACP, she said. “I think changing people's behavior patterns is the biggest challenge. At the same time, some people may argue [to] just place a hard stop . . . but that's a very big ethical dilemma,” said Dr. Bejjanki.

Lessons learned

The project's success is due in large part to involving the right people, from the lab director to the chief quality officer to IT staff, and having easily measured outcomes, she said. “Overall, it was a low-cost, high-impact intervention.”

Next steps

The project has expanded to include 40 to 50 commonly ordered tests, including blood cultures. Next, Dr. Bejjanki plans to analyze data on physician overrides of the alert, breaking them down by hospitalists versus residents and by unit. “That's our next step, to actually see where can we do even better,” she said. “But for the first step, I think we've achieved our goals and we're very happy with the project.”

Words of wisdom

“My biggest hope is that more and more hospitalists do these things. I think it's time to realize that we're leaders of quality improvement and that we bring about the right change,” said Dr. Bejjanki.