Health Technology’s $50 Billion Reality Check

Last week at CES in Las Vegas, the healthcare industry gathered to showcase its vision for the future—AI-powered diagnostics, advanced wearables, and autonomous systems promising to revolutionize care delivery. The week brought notable policy developments as well: the FDA released updated guidance clarifying its approach to wellness wearables, and Utah launched a pilot program allowing autonomous AI prescribing for routine medication refills. These announcements reflect the broader momentum behind technology integration in healthcare, raising important questions about how we evaluate what works and what doesn’t.

2026 brings real opportunity for digital health technology. At PHTI, we are eager to continue generating evidence-based assessments of the leading digital health solutions on the market, while advancing a critical dialogue on AI integration and emerging technologies like GLP-1s. Next week, we are introducing a first-of-its-kind Performance-Based Contracting (PBC) Playbook—a comprehensive report and interactive toolkits designed to help employers and health plans tie payment for digital health solutions to measurable clinical outcomes.

But it would be a mistake to look at the road ahead with nothing but unbridled optimism. The health technology sector is at an inflection point. AI is poised to transform care delivery and clinical workflows, and healthcare companies are racing to deploy new AI-powered solutions. At the same time, federal officials are exploring ways to integrate emerging technologies and modernize our healthcare infrastructure. Before the U.S. health system pours billions more into the next big thing, we should take stock of how health technology has already changed—or not changed—the cost and quality of healthcare.

From apps that improve mental health symptoms to home-based physical therapy, the market is flooded with tools that promise better health and large ROIs. According to Pitchbook Data, companies and investors have directed more than $50 billion into digital health tools for four conditions. These four conditions—diabetes, hypertension, musculoskeletal disorders, and anxiety and depression—affect roughly half of all Americans and are primary drivers of health spending, lost productivity, and deepening health disparities.

Understanding whether these new technologies work—and for whom—is essential. Healthcare costs consume nearly 20% of GDP, and the sector faces persistent quality gaps, access barriers, and workforce shortages. As the Executive Director of the Peterson Health Technology Institute, I have overseen independent evaluations of 50+ leading digital health solutions. We now have rigorous, evidence-based assessments of both the clinical effectiveness and economic impact of these technologies.

Investors, entrepreneurs, clinicians, and healthcare purchasers need a clear view of where technology is driving meaningful improvements in healthcare—and where it may just be hype. The next wave of investment requires a more deliberate approach.

Here are four lessons to guide successful development of future health technologies.

The most effective solutions are ones that identify proven care models and use technology to scale them efficiently. Too often, health tech is treated like pixie dust—sprinkle it on and hope something magical happens. Tools that work best are grounded in clinical care models that have been proven in traditional settings, like medication management for hypertension and cognitive behavioral therapy for anxiety and depression. The magic isn’t in inventing new forms of care but in finding new ways to deliver that care while improving access and reducing burden on the clinical workforce.

When it comes to economic impact, replacement beats augmentation. Digital health delivers the most financial value when it replaces aspects of traditional care models rather than layering on top of them. For instance, clinically-effective, virtual physical therapy can replace in-person PT or reduce the frequency of visits, which creates cost savings and health benefits (e.g., avoided imaging and pain medications). By contrast, tools that promote healthy behaviors for diabetes often layer costs on top of standard care without delivering enough incremental clinical benefit to justify the additional cost.

Early in the digital health boom, companies focused on engagement and satisfaction scores as a metric of success. More recently, attention has (appropriately) shifted to clinical outcomes. Success requires identifying the right patients at the right time in their care journey. Diabetes management tools, for example, deliver limited benefits when deployed across broad swaths of users with well-established habits for managing their diabetes. However, if these tools are supplied to patients who are newly starting insulin, they can help establish better habits that drive meaningful clinical benefits.

Generations of healthcare leaders have understood that population-level improvements in diet and exercise would have dramatic benefits for our health. Hundreds of health tech tools promise to motivate healthier lifestyles and self-monitoring of health data, including weight, blood pressure, and blood sugar. Yet, changing individual habits that have accrued over a lifetime is incredibly difficult. To date, apps that seek to remind, encourage, and “nudge” users toward healthier behaviors haven’t proven effective. While some solutions show an initial spike in engagement, most patients quickly regress to their original habits. We shouldn’t give up on strategies to motivate healthier decisions, but we cannot continue to pay for behavior change apps that fail to deliver durable clinical benefits.

The next chapter of digital health isn’t about novelty but measurable impact that stands up to scrutiny and evaluation. Health plans, health systems, and employers are becoming more sophisticated, demanding solutions with proof of clinical effectiveness and value rather than just engagement metrics. We have every reason to believe that a new generation of heath tech solutions, especially with the introduction of AI, will surpass historic performance. However, this is only possible if rigorous assessment, evaluation, and real-world monitoring are baked in at the outset.

The $50 billion invested has taught us valuable lessons. If we follow the evidence, we have no excuse not to get this right.