How EMR Integration Makes OR Forecasting Seamless

By Bradley McLain, Chief Technology Officer, ORlogic

Surgical forecasts are only as good as the data beneath them. You can have the most sophisticated forecasting model in the world, but the OR moves fast — add-on cases, shifting blocks, cancellations, reclassifications — and a forecast built on yesterday’s snapshot is stale before anyone reads it, much less acts on it.

This is why direct EMR integration sits at the foundation of everything ORlogic does. It’s what turns OR forecasting from a quarterly planning exercise into something operational.

The Mechanics
Our EMR integration polls the surgical schedule on a five-minute cadence. Every five minutes, we pull the current state of all booked cases across the perioperative suite: expected start and end times, actual start and end, anesthesia type, procedure category, primary surgeon, room assignment, and the rest of the fields hospitals use to encode the operational signals embedded in the schedule.

Most of what comes back, most of the time, is unchanged. A schedule is fairly stable hour to hour. But the changes that do arrive — the ones that matter — show up in our system within minutes of being entered in the EMR. A booking added at 6:42 AM is in our forecast by 6:45. A case that gets bumped to a different room flows through within a single polling cycle. An actual start time that diverges from the expected start — the kind of small signal that tells you a day is about to run long — is captured before the next case has even been wheeled in.

When substantive changes occur — typically in booked case volume, case mix, or expected duration — the daily forecast updates accordingly. Not every change triggers a meaningful shift; we have no interest in noise. But when the underlying picture has actually moved, the forecast follows. That’s the loop. It’s not complicated, and simplicity is the point.

Cadence Matters More Than Cleverness
There’s a temptation, when talking about forecasting, to focus on the modeling. Models are interesting, but in operational settings, cadence almost always wins.
Consider what an OR manager actually needs at 6:30 AM. They don’t need a model that’s three percent more accurate than last year’s. They need to know whether the staffing they have on the floor matches the work that’s actually going to walk in — including the case that got booked overnight, the one that got cancelled an hour ago, and the one that is running thirty minutes long.

They need the forecast to reflect current reality, not the version that existed when someone last clicked refresh.

A five-minute polling cadence isn’t there to impress anyone. It’s there because the gap between when a change happens and when a manager can act on it is where staffing problems live. A forecast that updates at the speed of the schedule is a forecast people will actually use. One that lags by half a day becomes another report nobody opens.

This is the part of OR forecasting that doesn’t get talked about enough. The accuracy of the underlying model matters, of course — but only after the data layer is solved. If your forecast doesn’t know about the case that was booked twenty minutes ago, model accuracy is beside the point.

Beyond the Intraop Suite
Once continuous, structured data flows in from the EMR, the boundaries of what’s worth forecasting expand on their own. The OR is the obvious place to start — it’s where the highest costs live, and where the schedule is most visible — but it’s not the whole suite.

That’s why our forecasting now extends into preop and PACU as well. A booked case isn’t just an intraop event; it’s a patient who needs a preop bay, a prep nurse, and a recovery slot when they come out. The schedule already implies all of this — durations, case types, and acuity together suggest where each patient will be at each point in the day — but most operations teams don’t see it as a single integrated picture. The preop charge nurse, the OR coordinator, and the PACU lead all plan an approach to the same day, often without any shared view of how those three plans actually fit together.

When you compute expected length of stay, track the actual start and end times through each phase, and apply appropriate staffing ratios, the same schedule that drives the intraop forecast becomes a full-suite forecast: how many preop nurses you need at 6 AM, how many staff in intraop at 10, how many PACU nurses at 2. It’s the same EMR feed. It’s just being asked to answer a wider question.

Where We’re Going: Hour-Level Forecasting
Another direction worth pursuing is finer, not wider. A daily forecast is a useful management object — it tells you, at the scale of a shift, what kind of day you’re looking at. But days aren’t uniform internally. A surgical suite at 7 AM and the same suite at 1 PM may be very different. Staffing decisions, break rotations, room turnover, and add-on capacity all operate at hour-scale, not day-scale.

Hour-level forecasting is where we’re heading next. It uses the same EMR feed, the same case-level data, the same five-minute refresh — but instead of collapsing a day into a single number, it preserves the structure that’s already there. The result is a view that matches how OR managers actually think about their day: not as twelve hours of “today,” but as a sequence of distinct operational moments, each with its own pressure.

The Integrated View
None of the individual pieces here are revolutionary. EMR integration exists. Polling a schedule isn’t novel. Staffing ratios have been applied for decades. Computing length of stay isn’t a research problem.

What’s different is having all of it move together, in real time, against the actual schedule the hospital is running today. The intraop forecast, the preop forecast, the PACU forecast, and eventually the hour-level breakdowns — they’re not separate products bolted together. They’re different views of one continuously updated picture.

The goal is straightforward: a perioperative suite where the OR coordinator, the preop charge nurse, and the PACU lead are looking at the same forecast, can see how their pieces interact, and can make decisions that account for each other rather than around each other. Better visibility across the suite. Fewer silos. Staff who feel like they’re working from a shared map rather than three different ones.

EMR integration for OR forecasting isn’t the interesting part of the story — but it’s the part that has to work before any of the rest of the story gets told.