What happened in health care technology this week – and why it’s important.
Groundbreaking Israeli Diagnostic Test Gets FDA Approval. The Next Challenge: Selling It to America
An article by Sagi Cohen in Haaretz.com showcases the recent FDA clearance of MeMed’s diagnostic test and platform, which can distinguish whether an infection is viral or bacterial within 15 minutes. Using a blood sample, doctors can use MeMed’s test to correctly diagnose acute infectious diseases, such as respiratory illnesses, urinary tract infections, and more. The technology, which has been cleared for both adults and children, uses the body’s immune response – rather than the presence of microbes or pathogens – to classify the illness. To do so, the test measures the levels of three immune system proteins.
Why it’s important – Getting an answer about whether an infection is viral or bacterial within fifteen minutes certainly does contribute to earlier diagnosis and treatment decisions. However, it’s important to understand that the MeMed test does have one significant limitation: It cannot differentiate between different kinds of bacterial infections. For patients in more serious condition, those taking medication or with a weakened immune system, knowing which type of bacteria can be crucial. In these cases, the test alone is insufficient.
SCIENTISTS DEVELOP 3D PRINTED MICRONEEDLE VACCINE PATCH THAT OUTPERFORMS JABS
Hayley Everett authored a post on 3D Printing Industry.com on research being conducted at Stanford University and the University of North Carolina at Chapel Hill (UNC), where scientists have produced a 3D printed vaccine patch that they claim provides greater protection than a typical vaccine shot.
Applied directly to the skin, the microneedle patch reportedly delivered an immune response ten times greater than a vaccine given into an arm muscle via a needle jab. The microneedle patches developed by Stanford and UNC scientists were 3D printed using a prototype 3D printer from Carbon, based on the firm’s Continuous Liquid Interface Production (CLIP) technology.
Why it’s important – Able to be self-administered, the patches provide a more attractive alternative to those with a phobia of needles while reducing the need for trained personnel to administer vaccines. Microneedle patches also deliver logistical advantages. Vaccines given via a needle typically require cold storage in fridges or freezers and usually require a trip to a clinic, hospital, or vaccination center. In contrast, vaccine-coated microneedle patches can be shipped anywhere in the world without special handling, and as they can be self-administered, it would mean fewer trips to hospitals or clinics for patients.
Epic’s sepsis algorithm is going off the rails in the real world. The use of these variables may explain why
Casey Ross published a follow-up report in STAT+ (subscription required) on the fact that Epic’s sepsis algorithm is using a curious piece of data to make its prediction: whether a doctor has already ordered antibiotics. The use of that information, which the company has not publicly disclosed, is contributing to a discrepancy between the algorithm’s accuracy in Epic’s internal testing and its performance in the outside world. Those problems came into view after multiple health systems attempted to validate the tool but found it performed significantly worse than Epic advertised in their hospitals.
The company tested the algorithm on data from three health systems but left the pivotal work of scrutinizing its impact in live clinical settings to its users, who have no obligation to test the model, register its use, or report findings to any independent body. Only a handful of the 180 customers using the model have published studies on its performance. While they have reached different conclusions about its overall accuracy, their findings point to the algorithm’s struggle to reliably flag the condition in advance.
“They’re just pushing these models out with no regard for any kind of evaluation, let alone rigorous evaluation. They just say, ‘Our area under the curve is .80,’ and they’re sending doctors alerts and changing workflows.”
Ziad Obermeyer, a physician and professor of health policy and management at the University of California, Berkeley, School of Public Health
Why it’s important – Experts said relying on antibiotics as a predictor means the model is often identifying sepsis after clinicians have already recognized it. Another glaring lapse, according to specialists in data science and machine learning, is that since launching the algorithm in 2017, Epic has not examined its performance across demographic groups, even though it relies on variables such as marital status and whether a patient’s ethnicity was documented in the health record.
Virtual Heart Modeling Can Help Personalize Cardiac Treatment
Researchers at Johns Hopkins have developed a method of creating personalized computational models of the heart that can be used to help cardiologists and cardiac surgeons decide on the best treatment options for people with heart problems. Their ongoing research was reported in ClinicalOmics online.
The models are created by feeding in various information such as echocardiography, MRI, and electrocardiography test results, plus blood pressure and other factors such as genetic test results, among other things. A virtual heart model is then created based on this information combined with known physiological and physical principles.
Why it’s important – While these models are still relatively new, recent studies have shown that they can help improve outcomes for patients. One fundamental way they can be used is to predict which patients are most likely to develop ventricular arrhythmias based on their physiology and other characteristics. Another use of these models is to predict how new and experimental treatments might affect the heart and why.
How Genetic Testing Will Create Personalized Therapeutics for Rare Diseases
Tech by Vice contributor Audrey Carleton reported on multiple research efforts underway to demonstrate that genetic testing is also becoming increasingly useful in the presence of rare symptoms. In July, researchers at the University of California San Francisco published a study on the use of genetic sequencing to locate mutations that cause Aromatic L-amino acid decarboxylase (AADC) deficiency. This genetic disorder starves the brain of the dopamine and serotonin required for essential motor and behavioral skills, like holding up one’s head or saying hello.
Despite evidence of its efficacy in difficult diagnostic settings, there remains a dearth of support from insurance companies, who are dubious that the approach is anything more than costly and experimental. The Personalized Medicine Coalition (PMC) is devoting its research dollars to disproving this, most recently via a paper on rare childhood disorders on a sample of data from the Nicklaus Children’s Hospital in Miami. Next on the docket for the PMC is pushing for the coverage of genetic testing by the Centers for Medicare & Medicaid Services. This federal agency oversees public insurance coverage.
Why it’s important – Reducing the steps involved in the “diagnostic odyssey” saves money for patients and insurance companies alike. It also leads to better patient outcomes. The coalition has been pushing for the passage of the CURES 2.0 Bill, which, among other items, creates a pilot program to test approaches to delivering personalized medicine under Medicaid and Medicare. If passed, it would begin turning the wheels required to make individualized treatments an option for low-income Americans, who otherwise could not afford them.