Innovation in the clinical laboratory has enabled a seismic shift in how—and how quickly—patients are diagnosed and treated. But laboratory medicine professionals know well that an analytically perfect test is not automatically the best one to use: It’s the impact on patient care that matters.
This fact has been a driving force in the movement to focus on laboratory stewardship, which combines the expertise lab experts bring about using the right test, at the right time, for the right patient—and combines that with awareness of financial and other responsibilities that define the full context of care for institutions, patients, and society.
We asked Joseph D. Yao, MD, a consultant in the division of clinical microbiology and an associate professor of laboratory medicine at the Mayo Clinic Alix School of Medicine in Rochester, Minnesota, about what role the laboratory plays in determining how tests are used in clinical practice, and how defining clinical utility fits in. Yao’s research focuses on the development of new and improved clinical diagnostic tests for viral hepatitis, HIV infection, and transplant-associated viruses.
As someone in lab medicine, how would you describe the difference between analytic validity, clinical validity, and clinical utility?
Analytic validity verifies that the performance characteristics of a diagnostic test are what the assay manufacturer intended them to be. The performance characteristics of a diagnostic test have been defined by CLIA and various laboratory accreditation agencies.
Clinical validity is whether the assay can accurately determine if a positive result correlates with disease in the patient. For example, if I’m designing an assay or have an assay to detect a virus in a transplant recipient, if I have a result that’s 10,000 IU/mL, does that indicate the patient has the disease? And what severity of disease does that patient have? Answering those questions requires clinical correlation with the laboratory generated assay results.
You can do it quantitively or qualitatively. For example, if it’s an influenza A detection assay, does a positive result mean that the patient has the disease? The assay developer has to validate that a positive result correlates with disease in the individual from whom the specimen was collected. That requires clinical data.
Clinical utility is defined as, what does the assay do in helping to manage the patient? In the influenza A virus example, the results allow clinicians to implement treatment.
In some cases, like with genetic tests, I may determine a specific mutation in a patient’s gene, but I don’t know what disease it will cause or what disease this patient will have in the future. That means that test does not have clinical validity.
On the other hand, if I have a test that can detect a mutation in a patient’s gene that can predict high probability of breast cancer occurrence in the next 3 to 5 years, then that has clinical utility if it enables the clinician to decide on treatment that improves that patient’s outcomes.
How has the approach to clinical utility for molecular infectious disease tests evolved along with the science?
Things have really changed a lot in the 20 years that I have been in the field. Molecular tests revolutionized diagnostic microbiology and infectious disease diagnosis, and overall disease management. They allow us to detect microbial pathogens that are not easily detectible by conventional methods. Cultures and antigen tests are just not very sensitive. Molecular tests allow us to detect infections that are caused by relatively low microbial burden.
We also are able to detect multiple pathogens in the same specimen at the same time. With rapid PCR tests, in one hour we can do up to 36 targets or microbial pathogens at a time. These are significant improvements.
These advances enable rapid, accurate diagnoses in patients, which allows them to get proper treatment. It also reduces how long it takes for patients to be treated, and avoids giving them other unnecessary tests.
What is the role of lab medicine in establishing clinical utility and working with payers, patients, and physicians given that a laboratory result is never used in isolation?
The clinical laboratory has an important role in partnering with clinicians, health system management, and payers. We give them expert data and also participate in studies to show clinical utility of tests.
One hot topic right now is whether payers should cover very expensive, multipanel molecular tests, like those for pneumonia, gastroenteritis, soft tissue infection, or sepsis. A panel with 32 combinations of viruses, bacteria, fungi, and parasites can cost $1,200. Should they pay for a test that detects 32 pathogens at the same time, or pay individually for one or three viruses, bacteria, or fungi at a time until you finally find the right one?
Each of these tests, whether it’s one or three targets, can be $100 to $150 each. You’ll quickly reach $1,200 by ordering tests sequentially. That could also take a week, and the patient may have lost the opportunity to get well in those 7 days.
In certain circumstances, especially with patients cared for in the ICU or transplant recipients who are immunocompromised, they should probably be tested with syndromic panels upfront to expedite a diagnosis and enable implementation of correct treatment as soon as possible.
But it has to be targeted. A healthy individual with a urinary tract infection who develops diarrhea 3 days after starting antibiotics probably does not require a $1,200, 28-pathogen stool test. They can probably just get one test for Clostridioides difficile because that’s most likely the cause.
I think payers are getting smarter, and working with experts in the field to come up with criteria for what they would and would not pay. That is a good trend. It’s not perfect yet, and a lot of work still needs to be done on clinical utility to define the patient groups or subsets of patients for whom these complex and expensive tests are indicated.
Some payers look for large or randomized studies to show utility, but can you explain why certain types of studies can work for a diagnostic test but others do not? How do you reconcile that?
It’s important to have a well-designed study and look at clinically appropriate and useful outcomes. About 10 years ago, our institution conducted blood culture panel tests. Instead of just doing blood cultures and conventional detection of pathogens and conventional phenotypic drug resistance tests, we compared that to using a panel of microbial pathogens that cause bloodstream infection and resistance determinants.
The goal was to see whether doing this test on a positive blood culture bottle would improve patient outcomes. Do they get resolution of their infection sooner than through conventional diagnostic methods? And does this lower the total cost of hospitalization and reduce length of stay? Ultimately payers are willing to pay for these tests if they can save on other aspects of patient care, like length of stay and disease severity, as less severity means less costs to the healthcare system.
What do you think lab medicine professionals need to be doing right now to establish clinical utility and value for testing services, and how should they work with infectious disease physician colleagues?
For greater impact, the laboratory should partner with clinicians to conduct clinical utility studies for new laboratory tests that could potentially have an impact on patient outcomes and healthcare costs. They have to partner together and design and conduct these trials.
For ordinary, routine care, the laboratory should have a good rapport with the infectious disease or clinical group in order to educate them on current tests and explain the pros, cons, and pitfalls so that they can understand the nuances of them.
We can also partner with them on diagnostic and antibiotic stewardship programs to help guide appropriate and judicious use of antibiotics, as well as appropriate use of expensive but powerful laboratory tests.
Jen A. Miller is a freelance journalist who lives in Audubon, New Jersey. +Twitter: @byJenAMiller