What if your front door isn’t completely random?
Watch any episode of the HBO series The Pitt and you’ll get a realistic idea of what it’s like to work in a fast-paced emergency department: the endless stream of arrivals, interplay between clinical factors and social determinants of health, and the constant juggling of patient needs against available resources.
Although the series is set in Pittsburgh, it accurately reflects day-to-day hospital operations here in Canada, too. Much like Dr. Robby and his colleagues, our emergency departments serve as the “front door” to the hospital. Clinicians are continually triaging and treating patients — often with too few resources.
Historically, there’s been a general acceptance that the front door is random and, therefore, the chaos is wholly unpredictable and uncontrollable. As a hospital IT leader with 40+ years of experience, I’d like to challenge that notion.
Finding patterns to help predict healthcare demand
To be fair, no hospital can control who walks into the ED or when they arrive. Whether experiencing chest pain, signs of a stroke, or other urgent symptoms or injuries, people show up when they need care. Clinicians can’t suggest that a patient come back next week when things are less hectic.
And yet, using data and advanced analytics can help hospitals get a better sense of when people are more likely to show up with certain clinical needs. Think of it like rolling a die: While you can’t predict any single roll, after a thousand rolls, you can start to anticipate the shape of how it will land.
Hospital demand behaves more like that than we may assume. Look at any single morning in an ED, and it looks like noise. But stretch the view to five or 10 years, and the trends become unmistakable — from the flu-season window that prompts an influx of cases to heat waves that send the elderly into distress.
The signals that are already available
Many of the variables that affect demand distribution are visible before any patients arrive.
Consider the wildfire smoke that drifted toward Toronto over the summer. Respiratory presentations were about to surge. I didn’t need a hospital system to warn me; I needed weather data. Reports of flu activity in Australia are a reasonable leading indicator for what Canada will see in the Fall. And a snowstorm in the forecast is like an early warning for staffing and transport problems.
Modern data and analytics capabilities make it possible to monitor external signals and turn them into actionable intelligence. Here are some signals to consider incorporating into operational planning:
- Use the Air Quality Health Index as a standing feed to help predict when respiratory cases are likely to climb.
- Monitor local event calendars for festivals, playoff runs, or concerts that will have a predictable footprint in trauma and intoxication. (Plenty of The Pitt episodes would back this up!)
- Track school calendars, which move paediatric respiratory volume every September.
Individually, each of these data streams is interesting. When you match them against your own historical data, they start to become a forecast. Most hospitals already do a version of this, but it relies on individuals, such as the nurse with 20 years of experience who instinctively knows when to call in extra staff. That instinct is real, but it’s also one retirement away from walking out the door. We shouldn’t ask one veteran’s memory to do the work of a system.
What it looks like when you connect the dots
Tapping into external data signals isn’t an all-or-nothing exercise. As I’ve emphasized in my work with OnX, hospitals can make real progress by focusing on one or two signals at a time. Smaller wins can help build credibility and momentum to fuel additional initiatives.
I know the work pays off from my own experience helping a regional health authority build an integrated decision support system. We used the data that hospitals were already collecting and shipping off to government agencies. Rather than waiting six months for that data to come back analyzed (and, candidly, stale), we linked it ourselves in close to real time.
I still remember one finding in particular: the same patients turning up at emergency departments across the region as many as 150 times in a single year. No single hospital could see this trend because each saw only their own set of visits. These were often people without a family physician who were cycling through with no one managing their care. Once we could see the pattern, we could act on it.
As with all things, you can’t manage what you can’t see — and you can’t see these patterns when data is stuck in silos.
Where are your opportunities to turn noise into signal?
While I would never assert that a dashboard can fix Canadian healthcare, I would argue that the most under-used asset in most hospitals is the data they already own, matched against signals sitting in plain sight. These are the opportunities I’ve been exploring lately with the team at OnX and that they want to discuss with you at the Canada East Healthcare Innovation Summit 2026 in November. In the meantime, get in touch to share the data or other challenges your hospital is prioritizing.