Identifying homelessness using health administrative data in Ontario, Canada: how coding policies impact case validity
Bibliographic record
Abstract
Conducting longitudinal research about the health of people experiencing homelessness poses unique challenges. In Canada, we previously demonstrated that identification through health administrative databases permits population-level studies, despite relatively low sensitivity. Since then, coding of homelessness became mandatory in hospitals nationally. As a result, case validity since 2018 is unknown. We re-validated case definitions for identifying homelessness in health administrative databases between 2018 and 2022 against the longitudinally collected housing history of a representative sample of people experiencing homelessness (n=640) and randomly selected, housed people (n=128,000) in Toronto, Canada. We calculated sensitivity, specificity, positive and negative predictive values, and positive likelihood ratios for 42 unique case definitions. We compared the resulting true positives against false positives and false negatives to identify potential causes of misclassification. The optimal case (best sensitivity/scalability beyond Ontario) definition included any indicator during a hospital-based encounter within 180 days of a period of homelessness (sensitivity=52.9%; specificity=99.5%). Among periods of homelessness with ≥1 hospital-based encounter, the optimal case definition had good sensitivity (75.1%) with minimal reduction in specificity (98.5%). Review of false positives suggests homeless status is sometimes improperly carried forward in healthcare encounters occurring after transitioning out of homelessness. Case definitions to identify homelessness using health administrative data exhibit moderate sensitivity and excellent specificity, with two-fold increases in sensitivity since the implementation of mandatory coding in Canada. Mandatory collection of social determinants of health information within administrative data present invaluable opportunities for advancing research on the health and healthcare needs of people experiencing homelessness.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.106 | 0.318 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.005 | 0.010 |
| Science and technology studies | 0.006 | 0.005 |
| Scholarly communication | 0.005 | 0.002 |
| Open science | 0.005 | 0.005 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".