Unadjusted and adjusted odds ratios (ORs) from logistic regression models for coding of hospitalizations for people experiencing homelessness, 2015/16 to 2020/21*
| Period | Unadjusted† | Quarter-adjusted‡ | PT-adjusted§ | CSD group–adjusted¶ |
|---|---|---|---|---|
| OR (95% CI) | Adjusted OR (95% CI) | Adjusted OR (95% CI) | Adjusted OR (95% CI) | |
| Pre-mandate | Ref. | – | – | – |
| Post-mandate | 2.31 (2.27–2.34) | 2.31 (2.27–2.34) | 2.29 (2.25–2.32) | 2.30 (2.26–2.34) |
| Late post-mandate | 3.11 (3.06–3.16) | 2.99 (2.94–3.04) | 2.89 (2.85–2.94) | 2.98 (2.93–3.03) |
Note: CA = census agglomeration, CI = confidence interval, CMA = census metropolitan area, CSD = census subdivision, PT = province or territory, Ref. = reference.
↵* From the Hospital Morbidity Database, Canadian Institute for Health Information.
↵† The unadjusted logistic regression model is testing the impact of period on identification of people experiencing homelessness.
↵‡ Controls for fiscal quarter and the Q1 2020 dummy variable.
↵§ Controls for fiscal quarter, Q1 2020 and PT.
↵¶ Controls for fiscal quarter, Q1 2020 and CSD group, defined as major CMAs (Vancouver, Toronto, Montréal), other CMAs (all other CMAs not classified as major), CAs and rural areas (all areas not defined as a CMA or CA).
Source: Hospital Morbidity Database (HMDB), 2015/16 to 2020/21, Canadian Institute for Health Information.