Your old EMR is costing you more than you think — and not in subscription fees

A physiotherapist in Brampton told me something last month that I haven't stopped thinking about. She'd been on the same EMR for six years. Not because she loved it. Because switching felt like a project she didn't have a weekend for. "I know it's bad," she said. "I just don't know how bad."

So I asked her to track it. Two weeks. Every time the software slowed her down, made her re-enter data, forced her to click through a screen that served no clinical purpose — she wrote it down on a sticky note. She ran out of sticky notes on day three.

Fourteen sticky notes. Three days. And she's someone who considers herself "fine" with her current system.

The question nobody asks about EMR software isn't what the monthly subscription costs. The question is what you're losing by staying.

The cognitive tax you stopped noticing

KLAS Research has been measuring clinician satisfaction with EHR systems through their Arch Collaborative surveys for years now, and the 2024–2025 data lands hard. Only 46% of clinicians agree that their EHR enables efficiency. Less than half. And lack of efficiency is the single metric most strongly correlated with burnout in their dataset.

The physiotherapist in Brampton — her name is Gurmeet — wasn't burned out in the dramatic, quit-your-job sense. She was burned out in the slow, erosive way that doesn't make headlines. Seven extra clicks to reach the charting screen. A scheduling view that defaults to the wrong week every Monday morning. An intake form editor so unintuitive that she'd given up customizing it and was using a paper form alongside the digital one. Each of these is minor. Together, across a full day of patients, they compound into something measurable.

I timed her doing three common tasks on her current platform, and then the same three tasks on a modern system we had her trial. New patient intake: 6 minutes versus 2. Charting a follow-up session: 11 minutes versus 4. Finding a patient's last three visit notes: 45 seconds versus 8 seconds (her system required scrolling through an unsorted timeline). On twenty patients a day, the difference was roughly 90 minutes. Not theoretical. Measured.

Ninety minutes she could have spent with patients, or with her family, or not at the clinic at all.

The thing about cognitive load is that it's invisible until someone measures it. Gurmeet wasn't aware she was losing 90 minutes a day. She'd adapted. Her brain had normalized the friction the same way you stop hearing a hum after living next to a highway for a year. When I showed her the timing comparison, she sat quiet for about ten seconds. Then: "I thought I was slow. I'm not slow. The software is slow."

Missed data is a clinical problem, not a tech problem

A November 2024 study in Health Affairs found that every additional hour a primary care physician spends on documentation reduces the chance they'll review a patient's outside records by 7.1%. That number should alarm anyone. Documentation burden doesn't just steal time. It crowds out the very clinical thinking documentation is supposed to support.

Gurmeet told me she missed a medication interaction two months ago. A patient was taking St. John's wort — mentioned it in passing during the session — and Gurmeet intended to flag it in the chart because the patient was also on sertraline. By the time she got to the chart at the end of the day, that detail had dissolved into the blur of fourteen other patient conversations. She remembered the clinical concern three days later, at home, loading her dishwasher.

I started writing this section thinking missed data was about software design. It's not, really. It's about what happens when software consumes so much attention and time that the clinician operating it has nothing left for the details that matter. The platform becomes a bottleneck for clinical judgment. A system that requires thirty minutes of administrative overhead per patient doesn't just cost thirty minutes. It costs whatever clinical insight would have occupied that time instead.

There's a parallel here that feels almost too obvious, but I want to name it anyway. We all watched it happen with Siri and ChatGPT. Apple had a voice assistant for over a decade. It worked. Sort of. People used it to set timers and check the weather. Then OpenAI released something that actually understood context, and within months, Siri went from "fine" to embarrassing. The technology was available earlier. What changed was that someone built a system that didn't make you work around its limitations. The old tool wasn't cheap — you were paying for it through a $1,000+ phone. You just didn't notice the cost because it was bundled into something else.

Legacy EMRs have the same problem. The cost isn't on the invoice. It's in every workaround, every re-entered data field, every sticky note that exists because the software can't hold the information where it belongs.

Your patients can tell

Nobody talks about this enough. Your patients notice when you're fighting your software. They notice the screen getting more of your eye contact than they do.

Research on EHR-related gaze time — the percentage of a visit the clinician spends looking at the patient versus the screen — consistently shows a positive association between patient-facing gaze time and patient satisfaction scores. One study published in the Journal of Family Practice found that patients were significantly more satisfied when physicians spent a higher proportion of the visit looking at them rather than their monitors. This isn't complicated psychology. People want to feel heard. A practitioner who's toggling between three tabs, waiting for a page to load, and scrolling through an unsorted medication list — that practitioner is physically present but attentionally absent.

Gurmeet mentioned something I hadn't considered. She said her patient satisfaction survey scores had dipped over the past year, and she couldn't figure out why. Her clinical skills hadn't changed. Her caseload was the same. What changed was a "platform update" her EMR vendor pushed in late 2024 that added two extra confirmation screens to the charting flow. It was supposed to improve compliance. What it actually did was add another thirty seconds per chart where she was staring at her monitor instead of transitioning between patients. Thirty seconds doesn't sound like much. Over twenty patients, that's ten minutes of eye contact her patients lost.

I thought of the Notion versus Microsoft Word comparison. People who grew up on Word don't think it's slow. It does what they need. Then they try Notion and realize the five minutes they spent formatting a header and fighting with auto-numbering was friction they'd internalized as normal. The document wasn't the problem. The tool was. But the tool was so familiar that frustration felt like a personal failing rather than a design failure.

Your EMR is doing the same thing to your patient relationships.

Staff who leave aren't leaving the profession — they're leaving your software

A 2025 industry report found that hospitals and clinics with higher nurse and staff turnover were 3.5 times more likely to be rated by their own employees as using "difficult or outdated" EHR systems. The KLAS data corroborates this from a different angle: clinicians planning to leave their jobs report average Net EHR Experience Scores of -14.7, versus 32.7 for clinicians planning to stay. That gap — nearly 50 points on a -100 to 100 scale — is enormous.

I talked to a clinic manager in Austin named Rachel who lost two massage therapists in eight months. Both cited documentation and scheduling friction as primary reasons. Not exclusively — compensation and scheduling flexibility mattered too — but both independently mentioned the software as something that made the job feel harder than it needed to be. Rachel's EMR was ClinicSense at the time. The therapists she lost went to a competitor clinic that had switched to a modern all-in-one platform three months earlier. The competitor wasn't paying more. They just had better tools.

Recruitment cost for replacing a single allied health practitioner — job posting, interviews, onboarding, credential verification, reduced productivity during ramp-up — runs between $3,000 and $8,000 depending on the market and the role. Rachel lost two. She was paying $69/month for ClinicSense. The EMR subscription was $828/year. The cost of not switching was somewhere between $6,000 and $16,000 in turnover-related expenses, plus whatever revenue she lost during the gaps.

Gurmeet's associate — the new grad who started six months ago — told me she evaluates prospective employers partly on their tech stack. "If a clinic is still on a system that looks like it was built in 2012, I wonder what else they haven't updated." She's twenty-six. Every new practitioner coming out of school thinks this way. They grew up on tools that work smoothly. I keep hearing that the switching cost is the obstacle. I think the bigger risk, and this surprises me to say, is becoming the clinic that new grads don't want to join because your software feels like an artifact.

The Notion moment is happening in healthcare — you can wait, or you can move

Every software category goes through a version of this. A dominant tool gets entrenched. Users build workflows around its quirks. Then something newer arrives that does the same job with 80% less friction, and the old tool's market position goes from "industry standard" to "legacy" in about eighteen months.

We watched it happen with project management. Microsoft Project dominated for decades. Then Notion, Linear, and Asana showed up and made it clear that the old tool's complexity wasn't a feature — it was a tax. The people who switched earliest got the most benefit. The people who waited ended up switching anyway, just later, with more accumulated data to migrate.

We watched it happen with AI assistants. Siri had a ten-year head start. ChatGPT made it irrelevant in six months. Not because Apple lacked resources, but because architecture matters more than brand. Bolt-on AI over a 2011 framework can't compete with AI built from the ground up.

We're watching it happen right now with practice management software. The platforms built between 2010 and 2018 — Jane App, Cliniko, SimplePractice — are solid tools that solved real problems when they launched. They also carry the architectural weight of every feature that was bolted on after the foundation was set. AI documentation is a $15 to $125/month add-on because it was built by a different team in a different era. Telehealth was a pandemic retrofit. Patient portals were an afterthought. Each layer costs money because each layer was built independently, and the pricing reflects the geology of the product roadmap, not the cost of delivering the service to you in 2026.

A platform built from scratch after 2023 doesn't have that geology. AI is in the foundation, not bolted on top. Telehealth and patient portals and online booking aren't add-ons — they're table stakes that ship with the base product. That structural difference is why Oli Health charges $19.95/month per clinician for a feature set that costs $114 to $350/month across legacy platforms (we did the full pricing math on 11 platforms, line by line, with receipts).

I keep thinking about Gurmeet's sticky notes. Each one represented a moment where her software made her job harder. Not a catastrophe. Just friction. Repeated hundreds of times a month. The monthly subscription for her old EMR was $79. The actual cost — in lost time, missed data, patient satisfaction, and the associate who's quietly evaluating whether to stay — was multiples of that.

The clinic owners who move early on this will look prescient in about two years. The ones who wait will eventually move too — they'll just do it after losing more staff, more time, and more patients who can tell when their practitioner is present and when they're fighting a screen.

Nobody is going to look back in 2028 and wish they'd stayed on their 2015 EMR a little longer.