ePCRs suck in a very specific, grinding way that only makes sense if you’ve actually done EMS work. On paper they’re supposed to make things cleaner and safer and more accountable, but in real life they mostly feel like something that was designed far away from the street and then dropped on top of it. They assume the world is cooperative and temporally coherent.
EMS is none of those things. A call doesn’t unfold in neat steps. You don’t get information when the software wants it. You get it when you can, from whoever happens to be conscious, coherent, or nearby, and sometimes it’s wrong or incomplete or changes halfway through the call.
The software doesn’t tolerate that very well, so the burden shifts to the medic to translate this madness into something that fits the form.
A lot of EMS medicine lives in narrative and judgment, but ePCRs tend to flatten that into checkboxes and dropdowns. The story ends up shoved into a text box at the end, like an optional add-on, even though that’s the part another clinician would actually want to read. You wind up with charts that look thorough but don’t communicate reasoning, change over time, or why you did what you did. It’s documentation that satisfies no-body - going from a completed run to a valid CM1500 is a chasm that spawns many a millionaires off-shore because the slog is a nightmare to do on your own or locally at a reasonable price. TO make all of this worse, the fields are scattered across tabs. Tiny buttons on a tablet in a moving ambulance. Slow load times on bad cell service. Forms that forget what you just entered or kick you out. None of these things sound catastrophic on their own, but they add up. Every little bit of friction steals attention and energy, and that cost gets paid over and over again, on every shift, for years.
Now consider this, is mostly for insurance. Hand-offs are long, and annoying, because documentation and handoffs are separate process. At Ambra911.com what we aim to solve is that we linearize the process.
Medics dictate unstructured data
We structure it
Allow medic to press inform hospital that uses a voice agent to inform the hospital and allow them to ask any questions from the chart
Automatically integrate this with their existing PCR systems so they’re not inconvenienced
Push the run into insurance flow directly
Model the kickback probability, collect more information automatically from either patient or provider to resolve
Push it into our QA
Get you paid
What was a 1-2 hour ordeal of documentation, and waiting 7 days to even submit a bill, letting debt accrue, is now less than 5 minutes of documentation, handoff and claims submission. I struggle to see how there’s a better system.
Mobile support is another lie everyone pretends not to see. EMS documents in low light, in noise, in motion, with gloves on, sometimes with one hand free and sometimes with none. Often time on Windows - which aren’t actually phones. Most ePCRs still feel like desktop software awkwardly shrunk to fit a tablet. They technically work, but they clearly weren’t designed for the physical reality of the job. When documentation is physically annoying, people delay it. When they delay it, accuracy drops. Then QA gets involved, and the cycle gets worse. Voice is the answer.
Capture first, structure later.
The system is especially cruel on the hardest calls. Cardiac arrests, pediatrics, psych, messy refusals, multi-unit chaos. These are the calls where information is incomplete and timelines are fuzzy, and they’re also the calls where the ePCR throws the most errors and demands the most precision. Right when a crew is drained and shaken, the software decides this is the moment to be pedantic. That sticks with people.
Imagine you had someone die on you, and you then had to click through AngelTrack. Shudder.
The chart has become a legal and financial artifact first and a clinical record second, and that shift leaks into how people document and what they prioritize.
QA doesn’t help when it’s built around field completeness instead of clinical sense. When feedback shows up late, ignores context, and focuses on missing boxes instead of clarity, it feels adversarial. People stop writing for humans and start writing for the system. The chart becomes something you have to get past, not something that helps anyone.
What makes this especially frustrating is that the fix isn’t mysterious. The core idea is simple and obvious to anyone who’s done the job: capture-first workflows. Get the story down fast while it’s fresh, in whatever form works best in that moment - voice, quick tags, rough timelines - and worry about structuring it later. Treat narrative as the backbone, not an afterthought. Respect the fact that EMS work is human, nonlinear, and messy.
Until ePCRs are built around that reality instead of pretending it doesn’t exist, they’re going to keep sucking, no matter how many features they add or dashboards they sell.