Testing regular expression searches and machine learning models to determine housing instability and low income status from primary care electronic medical record data in Toronto, Ontario
Bibliographic record
Abstract
BACKGROUND: Housing and income are important social determinants of health (SDoH). Primary care providers often do not have information about these determinants, which could be used to support equitable health system planning and care delivery. The aim of this study was to use primary care electronic medical record (EMR) data to test two approaches (machine learning and regular expression searches) to obtain information about patients' housing instability and low income status. METHODS: We used de-identified EMR data from the St. Michael's Hospital Academic Family Health Team (Toronto, Ontario, Canada). A Health Equity Questionnaire is also routinely distributed to patients and includes questions about income and housing status; this formed the reference standard. First, a regular expression (REGEX) classifier was created using key text terms and codes; the second approach used supervised machine learning models (XGBoost). Discrimination and calibration metrics were calculated as compared to the patient-reported responses. RESULTS: 11,794 eligible patients were included in the housing cohort and 10,454 were in the income cohort. Overall, both approaches had poor sensitivity for determining both housing instability (XGBoost: 3.1%, REGEX: 29.0%) and low income status (XGBoost: 41.7%, REGEX: 17.6%). Positive predictive value (PPV) was satisfactory for the machine learning approach (83.3% for housing, 72.9% for income). CONCLUSION: While the machine learning approach demonstrated reasonable PPV, the overall metrics were poor and unlikely to be useful in a clinical setting for identifying patients with housing or economic needs. More robust analysis could be explored, but continued patient-captured SDoH information is necessary.
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 | 0.004 | 0.019 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.000 |
| 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".