By Deborah Borfitz
November 13, 2019 | Point-of-care (POC) testing devices are getting smarter, but better planning on the front end can mean fewer disappointments for technology developers trying to win market acceptance of their products and care providers hoping to use them in real-world practice settings. Ping Wang, associate professor of pathology and laboratory medicine at the Hospital of the University of Pennsylvania, and director of the clinical chemistry section and core laboratory at Penn’s Perelman School of Medicine, has come up with checklists for both occasions.
Wang shared her insights at the 2019 Next Generation Dx Summit, as well as her definition of a “clinically ideal” POC test: convenient test location and sample type, simple test components, results in 10 to 20 minutes, adaptable to resource-limited settings, robust and mistake-proof, and clinically valid and actionable.
Clinically valid means “accurate and precise,” Wang says, generating results comparable to a central lab. “That’s not true of a lot of point-of-care tests today,” she adds. A POC test is clinically actionable if the results enable quick intervention in a clinical unit—immediate specialty referral from a primary care clinic, for example, or better chronic disease management.
The value proposition for a POC test is to “improve chronic disease monitoring, empower patients to improve outcomes, and decrease acute admissions and readmissions,” says Wang. It depends not only on merits of the technology itself, but its clinical adoption.
Approval by the U.S. Food and Drug Administration (FDA) doesn’t guarantee adoption, she notes. Among the common implementation barriers for POC devices are a mismatch between the technology and clinical need, lack of understanding of the test’s clinical utility and user-ability, and suboptimal user experience, engagement and support.
In an article published in Clinical Chemistry last year, Wang included a diagram of the development pathway for novel POC technology as well as a checklist of questions for different stakeholder groups to be asking themselves on the journey to adoption. Among the questions for clinicians and laboratorians, for example, are “Has the assay been FDA approved or cleared?” “Is it subjected to individualized quality-control practice?” “Are there clinical guidelines supporting its use?”
To operationalize technology, the question set includes “What is the intended patient population and clinical unit?” “Will it be placed in a core laboratory or clinical unit?” “What are the specific needs of the hospital, clinician, or laboratory?” “Can the platform be interfaced to electronic medical records [EMR]?”
Service providers and payers will want to find out if there’s an existing or newly developed procedural terminology code for reimbursement, Wang says, and if the device is cost-effective when compared to downstream clinical costs. The questions need to be revisited multiple times as technology evolves, she adds.
Among the long list of checklist questions for technology developers: “Is there an unmet clinical needs niche?” “Is there a clinically well-established biomarker?” “Is it currently being used clinically?” “Are there competing existing assays on the market?” “Is the performance of these assays meeting clinical needs?” “What is the testing sample (fingerstick whole blood, venous whole blood, urine, saliva)?”
Some of the most significant advances have been in the diabetes area, notably continuous glucose monitoring (CGM)—the “poster child” of POC testing, says Wang. FDA-cleared CGM devices include Abbott’s Freestyle Libre, Dexcom’s Dexcom G6, Medtronic’s iPro2 and MINIMED 670G, and Senseonics’ Eversense.
Innovation at Penn
Innovative POC device development at the University of Pennsylvania involves partnerships with engineers and individual scientists and clinicians, says Wang, and implementation is integrated into patient care. Among the resulting innovations is a microbubbling digital assay capable of ultrasensitive quantitation of protein biomarkers. Bubbling is an “ideal signaling strategy,” not unlike the technique for finding a hole in a tire.
The POC test comprises a microchip and integrated microwell array that traps protein complexes, Wang says, and a smartphone serves as the readout device. Detection antibodies generate oxygen, generating bubbles that can be seen under a microscope.
The microbubbles are too numerous for a human to count, so Penn researchers are instead taking a machine learning approach to the task, says Wang. The neural network algorithm used correlates well with lab results and performs well under different imaging conditions. In a recently published paper (DOI: 10.1002/ange.201906856), the microbubble technology compared favorably with Roche Diagnostics’ Elecsys PSA test for prostate cancer.
In the wearables arena, Penn researchers are working on arm bands that absorb sweat and test the glucose concentration of people with diabetes, Wang says. The group also has a smartphone optosensing platform under development for the detection of viral and bacterial infections that correlates “close to 100%” with gold standard tests. A low-cost smartphone spectrometer is also performing well in comparative studies.
Researchers recently submitted a paper on a rapid fentanyl screen for use in emergency rooms, says Wang. In the clinical study involving 218 patient samples the test was validated against liquid chromatography tandem mass spectrometry.
Rapid CYP2C19 genotyping has already been integrated into patient care to guide use of the blood thinner clopidogrel following surgery, Wang says. Results interpretation and a dosing algorithm are programmed into the EMR (Epic) and the test gets repeated in the lab. Pharmacist prescribing of Clopidogrel and a procedure for lab notification were critical to implementation success.