ARPA-H Bets on Biosensors and Wearables to Move More Care Out of the Hospital
At INVEST Digital Health in Dallas, the most clinically interesting technologies were not framed as futuristic concepts. They were presented as tools aimed at moving information gathering closer to the patient, and in some cases out of the hospital entirely. The ARPA-H portfolio companies featured in the Innovation Showcase focused on two practical problems clinicians confront every day: how to collect better physiologic data and how to recognize deterioration before a patient becomes unstable.
One company, NIRSense, is building biosensors around optical sensing technology with the stated goal of collecting data that would previously have required a blood test. Chief Commercial Officer Dan Rodenhaver described the platform as a way to support care for more advanced disease states at home. In clinical terms, that matters because access to biomarker-level information outside the hospital can change the threshold for intervention. If a device can reliably capture information that once depended on venipuncture and a lab turnaround, clinicians gain the possibility of earlier triage, tighter longitudinal monitoring, and fewer blind spots between visits.
The signal here is not simply convenience.
The company’s development path also reflects a familiar translational problem in medtech: sophisticated sensing is only part of the equation. Rodenhaver described the need for new data inputs rather than better applications of existing data, which gets at a major bottleneck in digital health. Many tools have been designed to repackage the same information in a new interface. A biosensor that produces genuinely different physiological data has a stronger chance of changing care pathways, particularly if it can help characterize worsening illness before overt symptoms drive a patient to the emergency department.
Wearable Monitoring Moves Toward Earlier Deterioration Detection
Sibel Health brought a different but complementary approach to the same clinical challenge. The company has developed flexible, wireless wearable monitors intended to capture vital sign information and identify early signs of patient deterioration. Its portfolio includes a pulse oximeter and a chest monitor designed to capture blood oxygen, heart data, and other patient information. Those outputs may look familiar, but the value lies in the continuous, low-friction way they can be gathered.
Traditional spot checks are often too sparse to catch gradual decline. In many settings, a patient can move from apparently stable to clearly unstable between intermittent measurements, especially when respiratory status, heart rate trends, or subtle changes in oxygenation are drifting in the wrong direction. Continuous wearable monitoring is meant to narrow that gap. Beth Schreiber, director of customer success and clinical implementation, described how the wearables work and how they can support clinicians. The bedside implication is straightforward: earlier detection can trigger earlier review, which can change disposition, escalation, and resource use.
For hospitals, this kind of technology has particular appeal in settings where nurse workload, patient volume, and acuity all pressure routine observation schedules. For home-based care, the promise is even broader. A device that can reliably track physiologic deterioration outside a controlled environment may let health systems extend oversight beyond the ward and into post-discharge recovery, where complications often emerge quietly.
Why Federal Funding Is Steering Toward Better Data Inputs
ARPA-H’s presence in the showcase also signals a broader funding philosophy. The agency backs research across disease-agnostic focus areas, with an emphasis on broad, cross-cutting innovation rather than single-disease silos. In practice, that means technologies are judged not only by the condition they target, but by whether they can produce a new class of clinically actionable information.
That model fits biosensors and wearables well. The healthcare system already knows how to measure heart rate, oxygen saturation, and blood chemistry in conventional settings. The harder problem is collecting those data at the right time, from the right patient, with enough fidelity to alter management. A home-monitoring platform that can measure more advanced physiologic states could support chronic disease management, post-procedure surveillance, and earlier detection of decompensation. A wearable monitor that flags deterioration before a crisis can reduce the lag between physiologic change and clinical response.
These technologies still face the usual hurdles of adoption. Data accuracy, patient adherence, clinical workflow integration, and reimbursement all determine whether promising devices become routine tools or remain demonstration pieces. A sensor that looks impressive in a showcase must still prove it can operate reliably outside controlled conditions, when skin contact, movement, environmental noise, and inconsistent use can degrade performance. Even when the hardware works, the downstream question is whether the data arrive in a form clinicians can trust and act on without adding alert fatigue or administrative burden.
The Clinical Value Will Depend on Workflow, Not Just Hardware
The strongest case for both companies is not novelty. It is clinical timing. If biosensors can move lab-adjacent information into the home, and if wearables can identify deterioration before symptoms sharpen, then care teams may have more time to intervene with medication changes, additional testing, or escalation of care. In respiratory illness, heart disease, and other conditions marked by abrupt physiologic shifts, that extra time can be the difference between outpatient management and hospitalization.
For now, the most concrete takeaway is the direction of travel. ARPA-H is backing tools that attempt to replace episodic snapshots with continuous or near-continuous physiological insight. NIRSense is aiming to extend blood-test-type data beyond the laboratory. Sibel Health is aiming to make deterioration visible sooner through flexible, wireless monitoring. Each approach attacks a familiar clinical failure point: too little information, too late.
The remaining question is operational rather than conceptual. These systems will need to prove that the data they generate are accurate enough to change management, simple enough to fit into real-world care, and durable enough to function outside the showcase environment. Without that proof, the technology stays interesting. With it, the hospital wall starts to look a little less fixed.