Edmonton Daily Brief · Sources

Sources

<a href="/sources.html">Sources</a>

EDB · Sections
  1. Front page
  2. Weather & Alerts
  3. City Hall
  4. Public Safety
  5. Development & Business
  6. Courts & Legal
  7. Sports
  8. Schools & Campus
  9. Community & Events
  10. Tech & Research
  11. The Province
  12. What’s On — every listing, weeks ahead
  13. Buses — the live tracker
  14. Search the paper

Alberta Health Services — Emergency Department wait times JSON API

Read
yes, on the basis of a decision the paper's publisher recorded in writing
What the paper uses
figures and listings, credited
How it is read
open-data API
Terms
its terms, read 2026-10-08: “…names and logos, may be reproduced in any form, or by any means, without…”
Robots
robots.txt allows our reader (checked 2026-10-08)
Offers its feed to readers
no
Permission
to be asked
Decided by
the paper's publisher, 2026-10-08 (DECISIONS #36: AHS ER waits published in the public interest; permission requested, pending refusal)
Notes

One of the strongest recurring daily-brief items available for Edmonton: a single concrete number readers feel. 'Longest ED wait in the Edmonton zone this morning: 10h24m at Fort Saskatchewan.' Twelve Edmonton-zone facilities including Royal Alexandra, University, Misericordia, Grey Nuns, Sturgeon.

WIRED 2026-08-23: read by pipeline/sources/health.py as the `ed_waits` fixture for the page (model-free; raises on failure so the rail records −1). Rendering is the front end's; contract in docs/INSTRUMENTS.md. Live 2026-08-22: 12 Edmonton-zone EDs reporting, longest Misericordia 7 hr 51 min. COMPLETELY UNDOCUMENTED — no link from the public waittimes.aspx page, which renders via a JS bundle. I found it by probing /Webapps/WaitTimes/api/waittimes; the shorter /waittimes/api/waittimes also returns 200. Because it is undocumented it carries no stability guarantee and could move without notice — monitor for schema drift. WaitTime is a preformatted STRING ('3 hr 15 min'), not minutes; parse it. Note field contains HTML entities (&amp;). AHS robots.txt disallows /org/ and /rls/ but not this path. Its mode has no generic collector and no code refers to this endpoint, so nothing fetched it and no health record ever existed. Disabled so the registry stops counting it as coverage; the entry is kept because the endpoint was verified and this is a wiring backlog, not a dead source. See ops/wiring.py. DELIBERATELY NOT WIRED, decided 2026-08-21 after verifying it works. The endpoint is live and good: keyed by city, and d['Edmonton']['Emergency'] returns 12 facilities with a current wait and a link each. The problem is not the source, it is us. This brief is a static page rebuilt about twice a day and ER waits move hourly, so a published number is wrong by the time it is read — and a reader choosing a hospital on a ten-hour-old figure is the harm case, not an inaccuracy case. Publishing it with 'check before you go' would mean the number carries all the risk and none of the value. The honest form is a plain link to the live page, which needs no pipeline. Same reasoning as the emergency alert notice: a copy on a schedule is not an alerting service.

If this is yours and you want it left out, write to [email protected]. We stop reading it and withdraw everything that cited it, past editions included. Every source · How this paper is made