As an oncologist, I’ve sat through the 24- to 48-hour EMR trainings that claim to prepare clinicians for new systems. They rarely do. The average physician still spends one to two days in technical onboarding and another month fumbling through screens before feeling comfortable. Versions vary wildly even within the same vendor, and every move to a new hospital resets the learning curve.

One night in a rural hospital, I watched a patient in pain wait 40 minutes for medication because a locum nurse couldn’t find the correct screen for the dose. She clicked through six pages before calling pharmacy. That delay wasn’t a training issue. It was a design failure, and a patient paid for it.

We talk endlessly about AI, interoperability, predictive analytics, and clinical decision support, yet the essential tool clinicians touch every day still behaves like a 1990s operating system. The CrowdStrike update failure in July 2024 wasn’t a cyberattack, but it still crippled access to medical records, imaging, and fetal monitoring across 759 U.S. hospitals. More than 200 hospitals lost access entirely. The incident wasn’t a freak event; it exposed the brittleness of a system that was never built with resilience or usability in mind.

Clinicians adapt to new treatments and scientific advances constantly. EMRs don’t adapt with them. They’re built in corporate laboratories, not clinical environments. Hospitals purchase budget-limited versions with stripped-down workflows. Interfaces differ. Labels change. Buttons move. Even identical EMR systems behave like different species from one health system to another.

The cost of this chaos is enormous:

• Productivity loss: Hours vanish into tutorials, workarounds, and late charting.
• Safety hazards: Unfamiliar layouts lead to misclicks, wrong doses, missed alerts, and delayed care.
• Emotional burnout: Clinicians feel incompetent not because they are, but because the software demands it.
• Widening inequity: Under-resourced hospitals get bare-bones configurations, making care slower and riskier in the communities that can least afford it.

And the industry’s response? “That’s just how it is.” That complacency must end.

Health care doesn’t need another feature-heavy platform. It needs clarity. Apple succeeded not because it had more features, but because it understood human hands and human attention. Health care deserves the same standard.

But real reform can’t come from vendors alone. EMRs are too entangled with compliance, billing, and federal regulations to be fixed piecemeal. A redesign requires a coordinated effort: EMR companies, health systems, and, critically, the Office of the National Coordinator for Health Information Technology (ONC).

A redesigned EMR ecosystem should follow three core principles:

• Universal logic: One menu structure. One navigation language. One consistent set of labels. A nurse or physician should not feel like they’re learning a new dialect every time they change hospitals.
• Simplicity over features: Most EMR features exist to satisfy billing, audits, and compliance. A clinical tool should prioritize safety, speed, and clarity.
• Usability as a regulated metric: If usability impacts safety (and it does) then EMR vendors should be publicly scored on it. Those scores should determine whether their products can be deployed in clinical settings. Usability is not a luxury. It is a patient-safety issue.

Right now, EMRs track codes more effectively than they track care. They were optimized for insurers and auditors, not for clinicians or patients. If we want to improve outcomes, we need tools designed around the bedside, not the billing office.

We tolerate this inefficiency only because we’ve normalized it: “That’s how health care works.” But we are long past the point where another tutorial, another onboarding, or another 40-minute delay for a patient in pain is acceptable.

Every other industry has undergone usability revolutions. Health care is the only one where the primary work tool gets more complicated the more “advanced” it becomes.

We need our Apple moment.

Because until we design technology around clinicians’ hands, attention, and workflow, rather than insurers’ dashboards, we will keep wasting six weeks of every doctor’s life on tasks that should take six minutes.

Sriman Swarup is a board-certified hematologist-oncologist and the cofounder of OncoNexus, an AI-driven oncology workflow platform focused on improving efficiency and care delivery. He also leads Swarup Medical PLLC, where he consults on clinical systems design, health equity, and digital health transformation. Practicing in rural Arizona, Dr. Swarup manages more than 3,000 patient encounters each year while advising startups and health systems on innovation and health care strategy. He writes about medicine, technology, and health care leadership at his website.