Latest News

Problems Aplenty Employing Point-Of-Care Devices In Clinical Settings

By Deborah Borfitz

September 11, 2019 | Clinical perspectives on point-of-care (POC) diagnostic testing was the subject of a lively panel session at the recent Next-Generation Dx Summit. Major themes were how implementations can go right—or terribly wrong—and the dubious practice of confirmatory testing.

Panelists included Jeanne Mumford, pathology manager for point-of-care testing at Johns Hopkins Medicine and president of Keystone Point-of-Care Coordinators; Joseph Campos, director, microbiology laboratory, in the infectious disease molecular diagnostics laboratory, laboratory informatics, at Children’s National Medical Center; and Robert H. Christenson, professor of pathology, and medical and research technology, at the University of Maryland School of Medicine.

“Near-individual” testing is more theory than practice at the three organizations represented. POC testing is hard to operationalize, Campos says. But labs may have a role to play in maintaining quality control, ensuring tests are performed correctly, and determining which ones need to be repeated in a hospital.

Mumford says the idea of POC testing replacing central lab testing makes her “cringe,” although Johns Hopkins documents results of at-home Coumadin tests in patients’ medical records and may use them to make medication adjustments. Christenson adds that the key challenge with POC testing is that it consumes nurses’ time, highlighting the value of emerging chip-based tests that automate much of the work.

Not all POC diagnostics are created equal when it comes to reliability of test results, in part due to user error, the panelists agree. A highly sensitive POC test for cardiac troponins is now available for detecting heart attacks, says Christenson, but clinician acceptance is required for a system to use it as part of a rapid rule-out strategy.

While home pregnancy tests are widely used, most obstetricians would want to repeat individual patient testing rather than trust the results, says Campos. Children’s National Medical Center has for some time allowed rapid strep testing on an outpatient basis, but all results get confirmed in the hospital’s central lab.

Continuous glucose monitors (CGMs) come to mind, adds Mumford, but there is no protocol indicating clinicians can or should rely on the readouts. Results interpretation is the issue, according to Christenson, questioning if the internet will provide 24/7 support in the future.

A representative from the U.S. Food and Drug Administration (FDA) attending the panel session indicates that the agency’s concerns include how primary care and emergency medicine physicians will be educated to respond to patients calling with their POC test results. If patients say they have strep throat, for example, would they be prescribed an antibiotic, or would a follow-up consult be required? “We’re really struggling with this,” she says. “If you have to confirm everything, [POC testing] becomes meaningless.”

Patients self-monitoring their glucose level or anticoagulant therapy know better than anyone about those POC tests because they do them repeatedly, notes Christenson. Not so with individuals who use a POC test to check for a sexually transmitted disease or learn if they are pregnant.

The “patient portal phenomenon” means individuals can now easily sign up with Quest Diagnostics or LabCorp to get their POC test results before their provider ever sees them, Mumford adds.

How information from POC tests gets reported back is a “regulatory gray zone right now,” responds the FDA representative. If the testing is taking place in a lab certified per the Clinical Laboratory Improvement Amendments, “the home collection point is what is new.”

Mixed Bag of Experiences

The input of pathologists, as well as workflow studies, is vital to the successful implementation of POC diagnostic testing, says Mumford, citing a failed initiative four years ago at Hopkins where pediatrics did all the negotiations with the device manufacturer. Nurses and respiratory therapists were the intended testing personnel and implementation in the neonatal intensive care unit dragged on for 13 months. Test kits still sit unused in her office.

At Children’s National Medical Center, efforts to introduce rapid flu and strep testing to the emergency room (ER) likewise failed because emergency physicians expected nurses to do the testing and, being short on time and resources, they refused, says Campos. “After we got a 24/7 lab a few years ago, we could do strep, flu and RSV [respiratory syncytial virus infection] testing and turn results around in less than 30 minutes, and that met the needs of ER physicians.”

POC lead testing has also “worked well,” continues Campos, noting that Washington, D.C. mandates checking blood lead levels in kids under the age of 5 due to municipal water contamination. The central lab had trouble meeting Department of Health expectations regarding results turnaround time.

But POC lead testing has not been successful everywhere, including two of Johns Hopkins’ busiest pediatric sites. “There have been a lot of falsely elevated samples and reporting is still an issue,” says Mumford.

Christenson says glucose testing is the “big dog” point-of-care test used by the University of Maryland Medical Center, but he named POC Coumadin testing as the biggest success with PharmDs offering immediate counseling to patients following a fingerstick. The same-day surgery unit requested POC pregnancy testing, but staff were “too busy to do it… so we had to take it away from them.”

The panelists agree that electronic interfaces are crucial. Christenson says the University of Maryland has for years used the Alere RALS POC data management system (now an Abbott product) that has a user lookup feature.

Johns Hopkins Medicine uses a vendor-neutral telecare system that complements a proprietary CGM system using artificial intelligence, says Mumford, but patient management is a “huge headache” because it can’t be done via the telecare system. Middleware is required to interface laboratory instrumentation with any electronic health record “proprietary or not,” she adds.

“When did lab testing become a commodity?” Christenson rhetorically asks. While POC testing can be valuable, they can also be more expensive than lab testing—and the value can be eroded by having nurses perform the test.

Christenson’s advice for new test developers is to make results clinically actionable, as with the Coumadin management devices and a newly FDA-cleared blood test to evaluate mild traumatic brain injury in lieu of a CT scan. Mumford adds that she is a fan of user focus groups where marketers, engineers and product developers “fire questions” at clinicians for hours at a stretch.

“Make the test as foolproof as possible,” Campos further recommends, “so if the test is done incorrectly it poses little harm to patients.”

POC devices are susceptible to interfering substances, as Christianson points out. One of the biggest problems in pathology is contamination of lab syringe blood, which can inadvertently throw off results of arterial blood gas tests, he says.

Christenson adds that he has yet to see in clinical practice best practices in sample collection at the pre-analytical phase. Results can differ depending on who is collecting the blood. Access devices used in the ER can literally “beat up red blood cells,” requiring specimens be recollected.