The CISSS des Laurentides confirms the occupancy rate is divided by licensed stretchers
On 6 September 2026 we wrote to the CISSS des Laurentides. Its six emergency rooms publish an occupancy rate every hour, with no gaps, but the six rates do not move in the same steps: at one the number jumps 20 points at a time, at another 2. We asked what the percentage was being divided by. The communications directorate answered on 10 September. Here is the question, the answer, and the measurements behind it.
The six gauges, measured
Window: 163 consecutive hours ending 10 September 2026. All six facilities published a rate in every one of those hours — no gaps. Counting the distinct values each ER published gives the step of its gauge, and the step gives the denominator.
| Facility | Distinct values | Gauge step | Published range | Stretchers in the denominator |
|---|---|---|---|---|
| Rivière-Rouge | 15 | 20 pts | 0 % – 280 % | 5 |
| Mont-Laurier | 14 | 12 pts | 50 % – 213 % | 8 |
| Argenteuil (Lachute) | 13 | 12 pts | 13 % – 163 % | 8 |
| Sainte-Agathe | 24 | 5 pts | 67 % – 200 % | 20 |
| Saint-Jérôme | 28 | 2 pts | 85 % – 151 % | 50 |
| Saint-Eustache | 28 | 2 pts | 103 % – 186 % | 50 |
Extracts from the 10 September answer
Communications directorate, CISSS des Laurentides, email of 10 September 2026, answering the question above. The French is the original; the English follows it:
Votre observation est juste : les variations du taux d’occupation ne se traduisent pas de la même façon d’un établissement à l’autre. Cela s’explique par le fait que le pourcentage publié est calculé à partir du nombre de civières autorisées (au permis) dans chaque urgence.
Your observation is correct: variations in the occupancy rate do not translate the same way from one establishment to another. This is because the published percentage is calculated from the number of authorised (licensed) stretchers in each emergency room.
Nous convenons qu’un pourcentage, lorsqu’il est présenté seul, peut ne pas refléter toute la réalité et gagne à être interprété en tenant compte notamment, de la capacité de l’installation.
We agree that a percentage, when presented on its own, may not reflect the whole reality, and is better interpreted taking the facility’s capacity into account.
The licence as denominator
The licensed count and the count of stretchers actually open are two different numbers. An ER can close stretchers without its licence moving, and the published rate goes on dividing by the licence. We had written elsewhere on this site that the denominator was the functional stretcher count; that line is corrected as of today.
If the denominator is a licensed stretcher count, then a 20-point step corresponds to five stretchers and a 2-point step to fifty, which gives the right-hand column of the table without our having to assume it.
For a resident it reads like this: one more person on a stretcher at Rivière-Rouge moves the published number 20 points, and the same person at Saint-Jérôme moves it 2. A 280 % at Rivière-Rouge rests on five stretchers. A 143 % at Saint-Jérôme counts fifty.
Checked for all 120 ERs, not just the six
The answer above covers six emergency rooms. We tested the same rule against the full hourly archive, 62,274 rows covering 120 ERs. 55,493 rows carry a published rate and a non-zero licensed-stretcher count. In those 55,493 rows, the published rate matches occupied stretchers divided by licensed stretchers, times one hundred, to within one point. No mismatches. No ER among the 120 out of line. Over those same 519 hours, licensed-stretcher count does not change for any of the 120 ERs.
Comparing one facility with another
The establishment maintains that the rate stays a useful indicator for following pressure on a given ER over time. A gauge compared against itself hour to hour does keep the same denominator. Comparing two doors would need the licensed stretcher count beside the rate, a number that already appears in the operating licences.
None of the establishment’s published figures are in question here, and its six ERs publish every hour while several establishments elsewhere in Québec leave emergency rooms silent for weeks.
Method
The step reads without us: take /api/facility/<permit>/history?hours=250 on this site, field occ_pct, sort the distinct values and take the smallest non-zero difference. Permits: 51217628 Rivière-Rouge, 51217610 Mont-Laurier, 51245306 Argenteuil, 51246874 Sainte-Agathe, 51228880 Saint-Jérôme, 51228864 Saint-Eustache. The 163 hours for all six facilities are available as a CSV, free, on request.
Page published 10 September 2026. The quoted answer was received 10 September 2026 at 11:38 EDT.
Do you work with these numbers?
This page is one cut of an hourly archive I have been saving since 10 August 2026: every Quebec emergency room, every hour, every column of the ministry’s feed, including the ones that vanish from the public table an hour later. The ministry keeps no such history.
CA$149 a month per organisation for the full extract refreshed hourly plus an API key over the same history, or CA$400 for a one-time extract of the whole archive. The CSV on this page stays free.
Leave an address and I will write back with the extract and how it is built. No card, no charge.