The following is a guest article by Matthew Blosl, CEO at DexCare, a Leading Patient Navigation Platform for Health Systems
It’s 10 p.m. A patient in rural America logs onto her hospital’s web portal and sees test results outside the normal range, with no explanation of what they mean. Her doctor’s office won’t open for another 12 hours, and the nearest urgent care center is 45 minutes away. She’s left with a long, anxious night and no way to get answers.
This gap—no access and no next step—is what AI is beginning to fill. Companies like OpenAI and Anthropic have introduced tools to help patients better understand their health and support clinicians in delivering care more efficiently.
Some experts, rightly so, remain cautious and point to AI’s tendency to hallucinate as a risk to patient safety. But for many patients, the more immediate problem is simple access to healthcare. The ability to see a doctor that day. And to understand test results when the email pings their inbox. And when over 37% of Americans live in healthcare deserts—with no access to essential medical services—AI is connecting patients to care in ways the current system can’t.
Today, ChatGPT answers healthcare questions for 230 million people worldwide each week. Patients get answers in seconds, without needing a copay, waiting for a callback, or scheduling an appointment.
Not surprisingly, most ChatGPT healthcare queries happen after traditional office hours. And with wait times for primary care appointments now measured in weeks or months— sometimes longer for certain specialties—patients are increasingly willing to seek help elsewhere, even if that help is imperfect.
Benefits are showing up in the exam room, too.
The technology helps patients arrive at their appointments more prepared, leading to deeper in-office conversations. As David Liebovitz of Northwestern University explains, AI chatbots can synthesize a patient’s history, surface potential concerns, and provide more context-aware insights that can improve decision-making. That’s progress from the era of ‘Doctor Google,’ where you’d type symptoms into search and brace for the worst. Could this headache really be a tumor?
However, enthusiasm and caution often travel together.
“The stakes are exceptionally high in healthcare,” notes Dr. Peter Bonis, Chief Medical Officer at Wolters Kluwer Health. “Whether these applications prove safe and effective over time is still uncertain.” He’s right. And it’s too early to know the technology’s long-term impact or limitations. But we need something to shake up our healthcare system. To force change.
As Medicare expands to cover nearly one-fifth of Americans, while the supply of physicians dwindles, seeing a doctor is only going to get harder. For AI to move from a chatbot in your pocket to the transformative technology that dominates news headlines, health systems must tackle a problem in plain sight.
The culprit is bad data.
Healthcare produces nearly one-third of the world’s data, yet much of it remains siloed and disconnected within health systems. The result? Information that patients need is often invisible, partly right, or flat-out wrong. One in three patients who book a doctor’s visit online encounters inaccurate information. And one in five can’t find the right provider at all. The consequences are felt everywhere. Doctors with open appointments go unfilled. Patients who could have been seen end up in the ER. AI, handed the same broken data, reaches the same dead end.
This data chasm stands between AI’s promise and what it can deliver.
AI can now tap directly into live healthcare data, but connectivity was never really the problem. If the data on the other end is fragmented, buried across too many systems, and inaccurate, then AI inherits those flaws, creating a faster, more confident path to the wrong answer. What AI ultimately needs is data worth trusting. Data that tells a system which doctor is available, who is the right fit, and how to get a patient to that doctor.
When the data is right, AI can do more than just answer questions after hours. It can match patients to the right doctor, surface the right history, and handle the work that clogs the system. As a result, routine appointments become earlier diagnoses and more precise treatment. And a country running short on doctors gets a little more out of the ones it has.
For the patient staring at test results at 10 p.m., AI is already in the room. What’s missing is the data foundation beneath it.
No comments:
Post a Comment