The FDA cleared 295 AI-enabled medical devices in 2025 alone, bringing the cumulative total to 1,451 authorized devices. Yet a JAMA Network study found that fewer than 2% of these devices cited a randomized clinical trial, and just three reported actual patient health outcomes.
Health technology companies are building products faster than the evidence infrastructure can keep up. The regulatory environment is increasingly favorable, with new reimbursement codes for digital therapeutics, expanded FDA pathways for software as a medical device, and growing payer interest in real-world outcomes. But the companies capturing this opportunity are those generating evidence that connects clinical performance to the outcomes payers and providers actually care about.
The Evidence Gap in Health Technology
Health technology occupies a unique position in the evidence landscape. Unlike pharmaceuticals, where clinical trials have decades of established methodology, and unlike traditional medical devices with well-defined post-market surveillance pathways, health tech products often generate continuous streams of data from the moment a patient begins using them.
This should be an advantage. Wearable devices, digital therapeutics, and remote monitoring platforms capture real-world patient data at a scale and granularity that traditional clinical research never could. Activity trackers reveal whether a new therapy truly improves daily mobility. Cardiac monitors show how effectively a medication controls heart rhythm throughout the day, not just during a clinic visit. Continuous glucose monitors capture metabolic patterns that periodic blood draws simply miss.
Yet most health tech companies struggle to translate this data abundance into the structured evidence that regulators and payers require. The gap is not in data collection. It is in evidence generation strategy.
A Regulatory Landscape Built for This Moment
The regulatory and reimbursement environment for health technology has shifted meaningfully in the past two years.
FDA clearance is accelerating. In 2025, 62% of all AI/ML device clearances were classified as Software as a Medical Device, and 10% included Predetermined Change Control Plans, signaling a shift toward lifecycle management of adaptive AI systems. The FDA has joined Health Canada and the UK’s MHRA to establish international guiding principles for these evolving frameworks. Industry groups like the American Hospital Association are advocating for streamlined 510(k) processes that allow expanded indications through robust post-market evidence rather than multiple separate submissions.
Reimbursement is catching up. In January 2025, CMS began covering digital mental health treatment devices under Medicare, creating the first Medicare reimbursement pathway specifically designed for digital therapeutics. By late 2025, commercial payers like Cigna Healthcare announced coverage of FDA-approved digital therapeutics. The CPT 2026 code set includes 288 new codes, with key additions for digital health services, remote patient monitoring, and AI-enabled diagnostics, signaling that these technologies are now viewed as established clinical tools rather than experimental.
Wearable data is becoming clinical-grade. Over 1,000 interventional drug trials have incorporated wearable-derived data since 2001, with adoption accelerating sharply in recent years. Late-2025 FDA guidance recognized real-world evidence from wearables, provided that standards for data integrity, bias mitigation, and explainability are met.
Where Real-World Data Delivers Unique Value
Health technology companies that invest in structured evidence generation can address several critical business needs simultaneously.
Algorithm validation in real-world settings. Pre-market testing often uses curated datasets under controlled conditions. Real-world performance across diverse patient populations, clinical workflows, and care settings provides the evidence needed to demonstrate that a product works reliably outside the lab. This is increasingly important as the FDA explores lifecycle management approaches for AI-enabled devices.
Outcomes data for reimbursement. The shift from volume-based to value-based care means payers want evidence that digital health products reduce hospitalizations, improve medication adherence, or lower total cost of care. Published research has shown that digital disease management programs can meaningfully reduce adverse events and readmission rates in chronic conditions. Generating this evidence proactively positions companies ahead of payer requirements.
Clinical utility demonstration. Adoption depends on demonstrating value to the clinicians using these tools in practice. Real-world evidence showing how a product integrates into clinical workflows, improves diagnostic accuracy, or reduces time to treatment informs both clinical adoption and purchasing decisions by health systems.
Post-market surveillance at scale. For software-based products that evolve through updates and learning algorithms, continuous real-world monitoring addresses concerns about model drift, bias, and performance degradation. This kind of evidence generation is becoming essential as the FDA develops post-deployment evaluation frameworks specifically for AI-enabled devices.
Patient engagement and adherence evidence. Digital health products succeed only if patients actually use them. Real-world data on engagement patterns, completion rates, and adherence provides both the clinical evidence of sustained benefit and the commercial evidence of product stickiness that investors and partners evaluate.
What Makes Health Technology Evidence Different
Health tech companies face evidence challenges that are fundamentally distinct from those in traditional pharma or medical device development.
The data sources are new and continuously evolving. Wearable sensors, mobile health applications, patient-reported outcomes collected through apps, and remote monitoring platforms generate data types that did not exist a decade ago. Determining which endpoints are clinically meaningful, how to standardize measurements across devices, and how to handle the sheer volume of continuous data requires expertise in both the technology and the regulatory science.
The regulatory pathways are still being defined. Over 96% of AI-enabled medical devices are currently cleared through the 510(k) pathway, but the FDA has acknowledged that this process may not be ideally suited for adaptive AI systems. Companies that engage proactively with evolving regulatory frameworks, rather than waiting for final guidance, position themselves for faster market access.
Interoperability remains a practical challenge. Health tech products generate data that needs to integrate with electronic health records, claims databases, and clinical workflows to produce the kind of evidence payers and regulators value. FHIR standards are improving this, but designing an evidence strategy that accounts for integration from the start saves significant time and cost downstream.
Building an Evidence Strategy
For health technology companies evaluating their real-world data strategy, a few practical considerations can guide the approach.
Start with the evidence requirements, not the data. Understanding what regulators, payers, and health systems need to see before designing data collection ensures that the evidence generated actually supports business objectives.
Design for continuous evidence generation. Unlike traditional one-time studies, health tech products can generate evidence continuously. Building this capability into the product architecture from launch creates a compounding advantage over time.
Choose data sources strategically. The right combination of product-generated data, claims data, EHR data, and patient-reported outcomes depends on the specific evidence question. Not every product needs every data source, but knowing which data vendors and partnerships align with your evidence needs is critical for efficient execution.
Plan for evolving standards. Regulatory frameworks for digital health are actively being developed. An evidence strategy that builds flexibility for emerging requirements protects against costly pivots later.
Exploring how real-world data can support your health technology product development, regulatory strategy, or reimbursement positioning? Schedule a consultation to discuss your specific evidence needs.
This concludes our series on real-world data across healthcare and life sciences. Previous posts covered medical devices, biotech and rare disease, and pharma. Each vertical presents distinct evidence challenges and opportunities, and the common thread is clear: organizations that invest in structured, strategic real-world evidence generation gain a meaningful competitive advantage.
Sources
- Innolitics. (2025). 2025 Year in Review: AI/ML Medical Device 510(k) Clearances. https://innolitics.com/articles/year-in-review-ai-ml-medical-device-k-clearances/
- The Imaging Wire. (2026). FDA Updates AI List with New Clearances. https://theimagingwire.com/2026/03/11/numbers-from-the-fda-show-radiology-is-maintaining-its-lead/
- Sivakumar, R., et al. (2025). FDA Approval of Artificial Intelligence and Machine Learning Devices in Radiology: A Systematic Review. JAMA Network Open, 8(11), e2542338.
- Lin, J.C., et al. (2025). Benefit-Risk Reporting for FDA-Cleared Artificial Intelligence-Enabled Medical Devices. JAMA Health Forum, 6(9), e253351.
- Bergman, A., et al. (2025). Characterizing industry payments for FDA-approved AI medical devices. Health Affairs Scholar, 3(12), qxaf211.
- American Hospital Association. (2025). AHA Letter to FDA on AI-enabled Medical Devices. https://www.aha.org/lettercomment/2025-12-01-aha-letter-fda-ai-enabled-medical-devices
- IntuitionLabs. (2026). FDA’s AI Medical Device List: Stats, Trends & Regulation. https://intuitionlabs.ai/articles/fda-ai-medical-device-tracker
- Centers for Medicare & Medicaid Services. (2025). 2025 Medicare Physician Fee Schedule: Digital Mental Health Treatment Device Codes.
- American Psychological Association. (2025). New reimbursement pathways have opened doors for using digital therapeutics. https://www.apaservices.org/practice/business/technology/tech-talk/reimbursement-pathways-digital-therapeutics
- Reilly, L. & Molnar, A. (2025). Policy and reimbursement strategies to improve patient access to prescription digital therapeutics. npj Digital Medicine, 8, 740.
- IQVIA. (2025). Digital Health Trends 2025. https://www.iqvia.com/insights/the-iqvia-institute/reports-and-publications/reports/digital-health-trends-2025
- Nature Reviews Drug Discovery. (2026). Wearable technologies in clinical trials for drug development: trends and emerging opportunities. https://www.nature.com/articles/s41573-026-01403-9
- Galen Growth. (2026). Digital Health 2026: Ten Predictions as the Industry Shifts From Hype to Hardwiring. https://www.galengrowth.com/digital-health-2026-predictions-hype-to-hardwiring/
- ICON plc. (2025). Understanding FDA regulations for AI in SaMD. https://www.iconplc.com/insights/blog/2025/06/24/fda-regulations-ai-medical-devices
- Grand View Research. (2025). Digital Health Market Size & Growth Forecast to 2030. https://www.grandviewresearch.com/industry-analysis/digital-health-market
- American Medical Association. (2025). 288 new CPT codes cover digital health, AI and more. https://www.ama-assn.org/practice-management/cpt/288-new-cpt-codes-cover-digital-health-ai-and-more
