Urine Tests Before Surgery Blasted as Mostly Unnecessary

"The vast majority" of urine tests conducted prior to scheduled surgeries to check for infections "were not plausibly indicated," said researchers who examined claims data from private insurers and Medicare.

And while the individual tests didn't cost much -- about $17 -- they added up to almost $50 million over the study's 11-year duration, plus another $5 million for antibiotics prescribed to patients with no clinical signs of infection.

"Patients and society bear the risk of inappropriate antibiotic use, which can result in adverse drug reactions, increased risk of infections such as Clostridioides difficile, and emergence of antibiotic resistance," wrote authors Erica Shenoy, MD, PhD, of Massachusetts General Hospital in Boston, and two colleagues in a JAMA Internal Medicine research letter, published as part of the journal's long-running "Less Is More" series spotlighting overused tests and treatments.

Preprocedural urinalyses were once routine in order to check for infections that could increase the likelihood of complications. But studies have demonstrated that, in reality, such testing rarely improves outcomes or even alters clinical management. Organizations including the Infectious Diseases Society of America and the U.S. Preventive Services Task Force have recommended against testing and prescribing for asymptomatic infections except in certain narrow indications.

Shenoy and colleagues wanted to see just how common the practice has been. They obtained data on some 13 million procedures performed from 2007 to 2017 from Medicare and the IBM Watson Marketscan database of commercial insurance claims. Mean patient ages in the two datasets were about 74 for Medicare and 45 in Marketscan. Procedure types spanned 14 specialties. The researchers did not count kidney and urological surgeries because guidelines do recommend urinalysis for most such procedures.

For the others, urinalysis was deemed appropriate when the claims data listed diagnoses of urinary tract symptoms, fever, or altered mental state. In the absence of any of those codes, the researchers classified the testing as "not plausibly indicated."

It was true that 75% of surgeries covered by the data did not involve preprocedural urinalysis, suggesting that adherence to guidelines on such testing was pretty good. But in the 25% that did, fully 89% across all types of surgery had no apparent indication; in no specialty was the rate of non-indicated testing lower than 84%.

The latter figures, according to Shenoy's group, indicate that traditional practice patterns "remain entrenched." The team called on insurers to take more steps to be more aggressive in denying claims for unneeded testing.

An accompanying editorial echoed that recommendation, noting, for example, that one reimbursement policy change implemented in Canada cut "low-value" vitamin D testing by 93% in subsequent years.

"Eliminating insurance coverage of low-value services based on national guidelines is a promising and underused approach for curtailing the use of preoperative tests," wrote Niloofar Latifi, MD, of Beth Israel Deaconess Medical Center in Boston, and Deborah Grady, MD, a JAMA Internal Medicine deputy editor.

But the editorialists suggested other approaches as well, including policies set by hospitals and provider groups and alterations in electronic records systems' decision-support modules. "Notably, educating clinicians on high-value practices was only found to be effective [in one study] if paired with a supplemental strategy," the pair cautioned.

Limitations to the analysis by Shenoy's group included the possibility of incomplete patient data -- patients may have had legitimate indications for testing and antibiotic prescriptions that were not recorded with the relevant diagnostic codes. Also, about half of the 11-year study period preceded the movement to limit "low-value" testing, which first gained broad support and attention in 2012 with the American Board of Internal Medicine Foundation's "Choosing Wisely" campaign.

Comment