Opportunities for Diagnostic Stewardship in Clinical Microbiology

April 18, 2021

In graphs of life expectancy vs. annual per capita healthcare spending for the 37 Organization for Economic Co-operation and Development (OECD) countries, the United States really stands out. The U.S. spends more than double the average of the other 36 countries (adjusted to U.S. dollars), but American life expectancy is not any longer relative to countries who spend less. In fact, life expectancy in the U.S. has stalled in recent years and has actually decreased during the COVID-19 pandemic. A lot of things contribute to total healthcare spending and to life expectancy, but pause to consider—how is the clinical microbiology lab impacting this relationship?

Life expectancy vs. healthcare expenditure per capita for 37 OECD countries from 1970-2015.
Life expectancy vs. healthcare expenditure per capita for 37 OECD countries from 1970-2015.

Several recent publications point to a need for diagnostic stewardship. Diagnostic stewardship is similar to, and may even promote, antimicrobial stewardship: making efforts to use diagnostic tests efficiently and eliminate unnecessary, ineffectual or even harmful misuse of testing. Diagnostic stewardship seeks to improve diagnostic test utilization for better patient care, which may mean reducing the use of some tests that are not helpful, leading to subsequent cost savings. Our first priority is to do all that we can to increase life expectancy and, relatedly, quality of life. But some tests only push the graphic to the right—more spending—without improving patient outcomes.

High-Cost Tests With Dubious Clinical Value

Some very expensive tests related to infectious diseases and clinical microbiology have come into use in recent years, but what is the evidence that they improve patient outcomes? There is a temptation to cast a wide net, but more data isn’t always helpful. There should be evidence that a particular test will be beneficial for patients. If there isn’t established literature on a particular testing method, clinical microbiology lab leadership should review test performance for their own patients periodically to see if the test is providing a benefit and should discuss the findings with physicians who inquire about the test.
For example, the Karius test, which is the only commercial test of its kind on the market, sequences cell-free DNA in blood specimens to attempt to detect microorganisms causing serious infection, and potentially avoiding invasive specimen collection—more than 1000 different organisms can be detected. Technologically, this test is exceptionally interesting, and it could be useful when current methods cannot identify an etiologic source of infection. But at a cost of $2000 per test, we should expect considerable clinical impact. Several studies have evaluated the impact of this test, including one that performed detailed chart review for 82 patients on whom the test was performed. Hogan et al. concluded, “The real-world impact of the Karius test as currently used in routine clinical practice is limited.” In only 6 of the 82 patients were the test results considered clinically helpful, whereas 3 had a negative impact and the vast majority were not helpful.
Another study looked at results for 79 pediatric patients at a single institution. A few of the results were helpful, but there were also a considerable number of microorganisms identified that are not likely to have played any role in the patient’s clinical course and may have led only to confusion. It is hard to know when to use such an expensive test, when the published results available now suggest it is seldom helpful. There may be a clear clinical benefit in selected patient subsets, or for select conditions. Hopefully, further studies will help clarify when it is more likely to be impactful, pushing us up on life expectancy and not just on healthcare spending. The same test has been suggested to have utility for monitoring cytomegalovirus (CMV) viral loads, but costs 10-20 times more than a quantitative CMV PCR (which directly measures CMV viral load only) costs in house or at a reference lab.
There are also reference labs that specialize in using next generation sequencing (NGS) assays on orthopedic revision specimens or urine specimens, for example. These tests are generally supported by small studies that “find” more organisms than were detected in culture. These types of applications need to be substantiated by large, multi-institution trials that demonstrate a clear clinical benefit—not just a bill—for patients and those who care for them. 

Low-Cost Tests Contribute to Unnecessary Spending Too

A test does not have to cost a lot to be a significant source of unnecessary lab (and health system) expense. The ASM Clinical Microbiology Virtual Journal Club meeting in December 2020 discussed 2 studies on urine and blood cultures that found that a significant percentage of orders were clinically unnecessary and that orders for unncessary cultures could be reduced by educational interventions aimed at clinicians. This type of diagnostic stewardship intervention ties in well with antimicrobial stewardship—unnecessary cultures or other testing can lead to unnecessary, and potentially harmful, antibiotic use.

Another lower cost per-test assay that can add up to a significant annual spend with over-utilization is Streptococcus pneumoniae urinary antigen testing (UAT). The most recent IDSA/ATS CAP (community acquired pneumonia) guidelines recommend not using this test routinely and reserving it for severe infection. This type of testing was also recently evaluated in a study published in Clinical Infectious Diseases entitled “Pneumococcal Urinary Antigen Testing in United States Hospitals: A Missed Opportunity for Antimicrobial Stewardship.” The study’s results actually make a rather compelling case that this particular test is an opportunity for diagnostic stewardship.

The researchers performed a retrospective cohort study of 159,894 adults admitted with community- or hospital-acquired pneumonia (CAP or HAP) at 170 different hospitals from 2010-2015 to evaluate variability in UAT use, association of UAT use with de-escalation of antibiotics and association of de-escalation with patient outcomes. They found that there was considerable variability between institutions in the percentage of CAP/HAP patients for which UAT was used: ~40% of institutions did not use UAT at all, others used UAT for CAP/HAP patients varying from <10% to >60% of the time. The researchers found that patients with positive UAT were less likely to have other resistant organisms recovered from lower respiratory tract cultures than patients with negative UAT (7.85% vs. 11.7%) and that in institutions that used UAT more often, there was a trend towards more antibiotic de-escalation in patients with positive UAT. Patient outcomes (deaths and transfers to ICU) were not different between patients with positive and negative UAT results, suggesting that the increased antibiotic de-escalation in UAT positive patients was thus not associated with worse outcomes, and therefore an opportunity for antimicrobial stewardship.

Let’s look at how the test actually performed in this study. 25,932 patients had UAT testing, of which 7.2% were positive. This comes out to ~ 1,867 patients with positive results and 24,065 with negative results. Patients with positive UAT were quite a bit more likely to have S. pneumoniae recovered from blood or sputum cultures (25.4% of UAT positive) than patients with negative UAT (1.9%). Applying these percentages to the above numbers means that approximately 474 of 1,867 (25.4%) with positive UAT also had positive cultures with S. pneumoniae, whereas approximately 457 of 24,065 (1.9%) with negative UAT had positive cultures for S. pneumoniae. So only about half (~51%) of CAP/HAP patients with blood or sputum culture positive S. pneumoniae had positive UAT, which is not great sensitivity since it is designed to detect S. pneumoniae infection. Studies have estimated the specificity for this type of test to be 94-97%. If 7.2% of patients test positive at 97% specificity, then more than 40% of all UAT positive results are likely falsely positive. In this large study set, UAT appears to be falsely negative in half of culture positive S. pneumoniae CAP/HAP cases and falsely positive for >40% of all positive results.

The authors note that mortality was not different among patients de-escalated in the setting of positive or negative UAT, or no UAT ordered. It is unclear what value UAT added, given the total picture. Whether or not antibiotics in CAP/HAP patients could safely be de-escalated doesn’t appear to correlate with UAT results (or testing at all), and the results do not appear to be reliable for diagnosing S. pneumoniae infection. Sounds like a good opportunity to save money and join the 40% of hospitals that don’t use it at all.

These are just some examples of potential opportunities for diagnostic stewardship. Clinical microbiologists should do everything possible to encourage better utilization of diagnostic tests to drive better patient outcomes and increase life expectancy, while also being good stewards of healthcare dollars and eliminating spending that doesn’t help patients.

Author: Matthew Pettengill, Ph.D., D(ABMM)

Matthew Pettengill, Ph.D., D(ABMM)
Matthew Pettengill, Ph.D., D(ABMM), is the Scientific Director of Clinical Microbiology at Thomas Jefferson University Hospital in Philadelphia, Pa., and is a Diplomate of the American Board of Medical Microbiology.