There’s a line we use constantly where I work: access is not scheduling — access is architecture. I’ve come to believe it’s one of the most useful ideas in health informatics right now, and I want to make the positive case for it, because I think the field is standing on top of an opportunity that’s genuinely exciting to build for.
Start with the stakes, because they’re bigger than most people outside the space realize. Musculoskeletal conditions are the leading cause of disability in the United States and account for more healthcare spending than any other category — an estimated $420 billion a year, affecting more than half of adults.[1][2] This is not a niche. It is one of the largest, most data-rich problems in American medicine.
Here’s the opportunity inside it. In musculoskeletal medicine, the richest data of the entire journey is created in the first thirty seconds — the moment someone describes what just happened to them. “I stepped off a ladder and my knee gave out.” That sentence is dense with clinical signal: mechanism, location, acuity, context. It is the highest-value data event in the whole pathway, and it happens before an EHR, an imaging system, or an analytics dashboard is anywhere in the picture. For years, informatics has poured its best thinking into everything after the encounter. The front door has been sitting there the whole time, quietly generating the best raw material in the system.
What’s changed is that we can finally build for it. Digital musculoskeletal care is on a steep growth curve — one analysis projects the market will reach roughly $10.9 billion by 2032, growing at about 16.4% a year.[3] Remote or digital-first guidance is becoming normal, and orthopedic organizations are actively rethinking how demand enters. That combination makes this the right moment to treat the access layer as an engineering problem worth solving, not a routing inconvenience to be tolerated. Structure the injury well at the point of capture, and everything downstream inherits clean, connected data instead of reconstructing it. That’s not a fix for something broken. That’s a new place to create value.
The technical part is the fun part, and it’s worth saying plainly to an informatics audience: the interesting constraint is designing a schema that can absorb human narrative without flattening it — one that stays legible to a clinician, not just to a model. That constraint makes the work better. The moment you decide a person, not an algorithm, is accountable for what happens next, every structured field has to carry meaning a human can act on and stand behind. Accountability and data quality turn out to be the same design decision seen from two sides.
That’s the choice I feel most strongly about, and it’s an optimistic one. A lot of the current conversation frames the model as the decision-maker and the clinician as a check bolted on afterward. I’d flip it: keep the clinician accountable, and let the technology make their judgment faster, more consistent, and better documented. When you build that way, the data you generate is cleaner, more auditable, and more useful to every system it flows into next — scheduling, coordination with a practice, an employer’s reporting, a workers’ comp file. Human-led isn’t the cautious option. In my experience, it’s the higher-performance one.
This is the conversation I’m most looking forward to having with peers at Rosettafest this year. Not “here’s a product” — the pattern. Any organization moving musculoskeletal demand is sitting on the same extraordinary front-end data and the same chance to structure it early and well. The tools required aren’t exotic. It’s careful schema design, consistent capture, and a deliberate decision about where human judgment sits in the flow.
The organizations that handle MSK demand best over the next decade won’t be the ones with the most data or the flashiest models, instead they’ll be the ones that connected the moment of injury to the moment of decision — and kept a person accountable across that gap. That connection is the work, and it’s some of the most rewarding engineering I’ve done. The data was always there. We’re finally building the architecture worthy of it.
Sources
[1] Cleveland Clinic ConsultQD, “Musculoskeletal Disease Causes Most Disability, Healthcare Spending in U.S.” — https://consultqd.clevelandclinic.org/musculoskeletal-disease-causes-most-disability-healthcare-spending-in-u-s-but-gets-less-than-2-of-nih-research-funds
[2] Cigna Newsroom, “10 stats on musculoskeletal conditions in the United States” (annual MSK healthcare costs ~$420B; affects more than half of U.S. adults) — https://newsroom.cigna.com/10-stats-on-musculoskeletal-conditions-in-the-united-states
[3] GlobeNewswire / ResearchAndMarkets, “Digital Health for the Musculoskeletal Care Market: 2026–2032 Outlook” (projected ~$10.9B by 2032, 16.4% CAGR) — https://www.globenewswire.com/news-release/2026/06/24/3316973/28124/en/Digital-Health-for-the-Musculoskeletal-Care-Market-2026-2032-Outlook-Increasing-Investment-in-Health-and-Fitness-Apps-Drives-Adoption-of-Digital-Health-Solutions-for-Musculoskeleta.html