What EMR regulation can learn from the trans-Canada railway

In a previous post, I asked a simple question: if digital health has the technology, the funding and widespread adoption, why does it still feel broken?The answer pointed to something deeper than any single stakeholder. Different parts of the system failed in different ways, which suggests the problem isn’t the players. It’s the structure.It turns out we’ve seen this pattern before. Just not in healthcare.Over 90% of Ontario’s family physicians use an electronic medical record. And almost none of those records can talk to each other.Thousands of separate, disconnected systems. A patient’s health information scattered across clinics, hospitals, labs and pharmacies with no way to bring it together. Emergency physicians flying blind because they can’t see what the family doctor documented. A system that was supposed to improve care is actually fragmenting it.How did we get here? By regulating the wrong thing.The provincial gauge disasterMost people know the Canadian Pacific Railway story. British Columbia joined Confederation in 1871 on the promise of a transcontinental rail link. The CPR got built by 1885. Last spike. National unity. Done.What most people don’t know is what happened 30 years earlier.Canada’s first railways were built to the international standard gauge of 4 feet, 8½ inches—compatible with American lines. Very sensible.Then in 1851, a Royal Commission convinced the province of Canada to adopt a different gauge: 5 feet, 6 inches: the “provincial Ggauge.” The government tied it to subsidies. If you wanted public funding, you had to lay broad-gauge track. Broader track was more stable, carried wider cars and had the added benefit of being incompatible with American railways, which would prevent military incursion.It sounded clever. It was a disaster.Canadian railways couldn’t interchange freight with American lines. Every shipment crossing the border had to be unloaded and reloaded. Trade was hobbled. The economic cost was enormous. By the 1870s, Parliament repealed the Provincial Gauge Act and the railways spent years and enormous sums ripping up track and converting to standard gauge. The Grand Trunk started converting in 1872. The Intercolonial converted in 1875 while still under construction. By the time the CPR was built, nobody would have considered using anything but the international standard.The lesson: standardize the thing that connects systems to each other. Let everything else compete.The gauge is invisible. Passengers don’t care about it. But get it wrong and nothing works.Ontario got it right onceOntario’s OHIP billing software market tells a different story.Starting in the mid-1980s, the government needed physicians to submit billing claims electronically. The Ministry of Health didn’t try to build billing software or certify what it should look like. It did something much simpler: it published a data transmission specification. First for 5¼-inch diskettes, then 3½-inch diskettes, then electronic data transfer. The spec defined the content and the format—the gauge and the stations, if you will. What the software looked like, how it handled workflow, what features it offered—that was up to the vendors.The result was a healthy, competitive market. Many vendors entered. Over time the market consolidated naturally—not because switching was painful, but because better products won. Switching costs stayed low because the data content and format were standardized. Prices stayed reasonable. Functionality kept improving. The market segmented along practice size, which is exactly what you’d want.This wasn’t central planning. It was a clear infrastructure standard with open competition on everything above it.Then we moved to EMRs and forgot everything we’d learned.We did the provincial gauge thing with EMRsWhen Ontario started investing in primary care EMRs, the government regulated the product instead of the infrastructure. OntarioMD created certification specifications: what the EMR had to do, what functions it had to support, what data it had to capture. Vendors built systems that met the specs. They got certified. Mission accomplished.Except nobody standardized the layer underneath. Each EMR stores data in its own proprietary format. There is no common clinical terminology, no standardized data model, no shared infrastructure for exchanging information between systems.We regulated the train. We forgot the gauge.Once data is locked into proprietary formats, the economics shift. Switching becomes expensive, so physicians stay even when systems underperform. Vendors, knowing their customers are unlikely to leave, have less pressure to invest in usability or innovation. Over time, competition weakens not because better products win, but because the cost of moving becomes too high.We built beautiful, government-certified trains. They sit on five different gauges of track.Vendor lock-in does not occur because vendors want to hold data hostage. It occurs because the industry structure designed it.What the CPR got rightThe CPR wasn’t just an engineering achievement. It was an infrastructure decision.Standard gauge didn’t tell railway companies how to design their engines, how many cars to run, or how to price their tickets. It standardized one thing: the distance between the rails. Everything above the track was open to competition. Everything at the track level was shared.That’s the model digital health needs.What procurement needs to get rightOntario is now procuring a new province-wide primary care medical record system. The language is encouraging—integration, interoperability, a complete patient record accessible across the province. A multi-vendor approach that preserves physician choice and keeps competition alive.But multi-vendor only works if every vendor lays standard-gauge track. Two tests tell you whether the procurement has succeeded. Can a patient’s key health information—medications, allergies, diagnoses, test results—be shared in a way that’s useful to patients, other clinicians and other doctors, regardless of which EMR produced it? Can that information be automatically incorporated into a receiving EMR without someone re-keying it?If the answer to either is no, we haven’t standardized the gauge. We’ve just certified more trains.If instead the procurement focuses on certifying products—what the EMR looks like, what features it offers, what workflows it supports—we will have repeated the Provincial Gauge Act of 1851.The gauge is the hard partProcurement expediency has a way of overriding architectural discipline. It’s easier to certify a product than to build invisible infrastructure. You can hold a press conference about a new EMR system. It’s harder to hold one about a data exchange specification.But nobody remembers who regulated the colour of the first Canadian locomotives. The decision that mattered, the one that made the national railway possible and the nation it connected, was the decision to standardize the gauge.Ontario is about to lay track for the next generation of primary care. We can continue to certify trains and hope they connect. Or we can standardize the gauge and allow a real system to emerge.One path leads to more fragmentation, rising costs and slower progress. The other creates the conditions for competition, innovation, and information that actually flows.We’ve made this mistake before. The question is whether we’re about to make it again?Dr. Karim Keshavjee practiced family medicine for 20 years. He is now program director of the master of health informatics program at the University of Toronto. This item originally appeared on Medium.

What EMR regulation can learn from the trans-Canada railway

Register for free access or login.

Related Topics