Flow analytics
Measuring emergency department flow: door to doctor and results to reassessment
Door-to-doctor gets the attention, but the emergency visit has at least six segments, and the second-longest wait many patients feel is the one after their results come back. Here is how to measure flow so the numbers mean the same thing to clinicians, leaders and the province.
- By
- EDFlow team
- Published
- Reading time
- 4 min read
Start with the definition the province uses
Physician initial assessment (PIA) is the anchor of emergency department measurement in Ontario. It runs from triage or registration, whichever comes first, to the first assessment by a physician or nurse practitioner.[1] The average in 2022/23 was 118 minutes, up from 87 minutes in 2013/14, with a 90th percentile of 257 minutes and large regional variation.[2]
The 90th percentile of time to PIA was the most selected indicator in 2025/26 hospital Quality Improvement Plans, chosen by about half of hospitals, and the province's Pay-for-Results priorities target 3.4 hours or less.[3][4] If a department measures one thing, it measures this, and it should use exactly the provincial definition.
Measure every segment, not just door to doctor
A visit is a chain of waits. Measuring each link shows where time goes and who can change it.
| Segment | What it usually reflects |
|---|---|
| Arrival to triage | Front-end staffing, ambulance offload |
| Triage to physician initial assessment | Physician capacity, space, acuity mix |
| Assessment to first order | Workup habits, directives at triage |
| Orders to results | Lab and imaging turnaround by test, day and hour |
| Results to reassessment | The hidden wait: physician attention after results |
| Decision to departure | Discharge process, or access block for admitted patients |
The segment that surprises teams most is results to reassessment. Patients experience it as a second peak of anxiety: their tests are done, and nothing seems to happen. Measured by hour, it often reveals when physicians are pulled toward new arrivals.
Medians for people, 90th percentiles for managers
Emergency waits are skewed: most visits are short and a few are very long. Canada's emergency physicians warn that averages mislead in such data and recommend medians for public reporting, with the 90th percentile for internal management.[5] Their benchmarks give a sense of scale:
| Measure | Median | 90th percentile |
|---|---|---|
| Time to physician initial assessment | 1 h | 3 h |
| Time to inpatient bed | 2 h | 8 h |
| Length of stay, discharged CTAS IV and V | 2 h | 4 h |
| Length of stay, discharged CTAS I to III | 4 h | 8 h |
| Length of stay, admitted | 8 h | 12 h |
Access block, admitted patients waiting in the department for a bed, is the main driver of crowding, not so-called inappropriate visits.[5][6] A flow dashboard that ignores boarding will blame the emergency department for a hospital problem.
Balancing measures keep improvement honest
Every improvement effort needs measures that would reveal harm. Two matter most in emergency care:
- Left without being seen. It rose from 3.5% in 2018/19 to 5.3% in 2022/23 in Ontario.[2] Any change that makes waits more visible must be watched against it.
- Returns within 72 hours. A faster department that sends people home too soon hasn't improved.
Patient experience belongs here too, measured with the questions Ontario hospitals already ask, such as "If you had a long wait, were you told why?"[7]
Clinician metrics without harm
Per-physician timing is powerful and easily misused. Public league tables drive corner-cutting and distress; audit and feedback work best when clinicians help design them and see their own data privately, with case-mix context.[8]
- Show each clinician their own numbers; show leaders blinded peer ranges (interquartile range), never a rank.
- Filter by zone, shift and acuity mix so a night in the acute zone isn't compared with a day in fast track.
- Pair speed with returns within 72 hours, and aggregate monthly rather than daily.
- Suppress small cells (fewer than five) in any export.
EDFlow builds its analytics on these rules, with a definitions page that states each metric's formula and source. A number nobody can check is a number nobody should act on.
Frequently asked questions
How is time to physician initial assessment measured in Ontario?
From triage or registration, whichever comes first, to the first assessment by a physician or nurse practitioner, as defined in Ontario's Quality Improvement Plan technical specifications.[1]
Sources
- [1]QIP Indicator Technical Specifications (opens in a new tab) · Ontario Health
- [2]Value-for-Money Audit: Emergency Departments (opens in a new tab) · Office of the Auditor General of Ontario, 2023
- [3]Highlighting the 2025/26 QIP Indicator Matrix and Selection Trends (opens in a new tab) · Health Quality Ontario (Quorum), 2025
- [4]Operational Direction: Priorities for Spring/Summer 2025 (opens in a new tab) · Ontario Health, 2025
- [5]Emergency department overcrowding and access block (opens in a new tab) · Canadian Journal of Emergency Medicine (CAEP), 2013
- [6]Emergency Department Overcrowding in Canada (opens in a new tab) · Canada's Drug Agency
- [7]Ontario Emergency Department Patient Experience of Care Survey (opens in a new tab) · Ontario Hospital Association
- [8]Involving end-users in the design of an audit and feedback intervention in the emergency department (opens in a new tab) · BMC Health Services Research, 2019