physIQ provides cardiopulmonary insight from daily-wear sensor

Originally published in BioWorld MedTech

By Annette Boyle

The move away from in-clinic testing continues, with another company offering a convenient, at-home sensor that gathers critical health information without requiring any change on the part of the patient, which may provide even more accurate, real-world data on patient health than the tests they replace. 

physIQ Inc. has found a way to estimate maximum oxygen utilization (eVO2MAX) on par with cardiopulmonary exercise tests by using from a non-invasive biosensor patients wear as they go about their daily lives. The device enables more frequent, less burdensome and as accurate assessments as in-clinic tests such as the cardiopulmonary exercise test (CPET) or the six-minute walk test. Cardiopulmonary functional evaluation provides insight into patient status and can provide an early alert to physicians of deterioration. In clinical trials, these functional assessments indicate the effectiveness of therapies.

“Digital biomarkers from wearable sensors show great promise for serving up huge amounts of rich data to better understand how patients are doing in their real world lives,” said physIQ co-founder and Chief Digital Health Officer Matt Pipke. “eVO2Max is an example of how rich data can update our health measurement toolset from data-sparse practices like the six-minute walk test to continuous, intelligent and easy-to-use evaluations with higher accuracy.”

More data, better predictive ability

The sensor enables round-the-clock streaming of vital sign and activity data. While “the eVO2Max biomarker makes an estimate using 24 hours of continuous data, the more 24-hour periods used, the more estimates can be provided, allowing for a distribution mean that can provide high accuracy to the patient’s intrinsic functional capacity limit,” Pipke told BioWorld.

“One of the challenges of in-clinic testing is that it assesses capacity only during that one time, in that moment, in an unnatural environment. eVO2Max is designed to estimate a person’s maximal oxygen consumption capacity in their natural environment without formal ‘performance’ testing,” Pipke added. The natural setting also eliminates expectation bias and anomalies that can skew results.

That could provide critical information for identifying patients at risk. The company’s LINK-HF trial demonstrated that the biomarker flagged deterioration at an early stage, predicting hospitalization in patients with chronic heart failure up to 13 days before it occurred. The platform detected precursors of hospitalization with 76% to 88% sensitivity and 85% specificity, which the study investigators noted was comparable to the predictive accuracy of implanted devices.

A presentation by the Chicago-based company at the 2022 American HeartAssociation Scientific Sessions showed that eVO2MAX had a very high correlation (r=0.91) to CPET-mea-sured VO2Max, the current gold standard. The 262 participants in that study were drawn from both healthy as well as chronic illness patient populations.

The improvement comes down to use of artificial intelligence (AI) to improve prediction. “The AI is doing what AI does best: Trained with lots of data and with labeled targets, it learns on its own to look for telltale dynamics of vital signs and activity data, that reveal the intrinsic functional capacity of their cardiopulmonary system,” Pipke explained. “We’ve trained the AI to output a value that is known to clinicians and pharma scientists, so that they can easily interpret it in view of the conventional gold standard CPET.”

More useful trials

The sensor could also produce pharmaceutical or device trial results that more closely match what users will see in the real world by extending the assessment period. “For providing comparison between a baseline and post-drug treatment in a clinical trial, it may be typical for a study participant to wear a patch for a week at baseline, and for a week at several intervals plus end of trial, to provide a clean, highly detailed estimation of their VO2Max as it changes potentially with trial treatment over time,” Pipke noted.

physIQ is preparing a 510(k) submission to the FDA and plans to make eVO2Max available as an evaluation tool for real-time care shortly after clearance, which is expected in 2023.