Extracting social determinants of health from inpatient electronic medical records using natural language processing
Bibliographic record
Abstract
BACKGROUND: Social determinants of health (SDOH) have been shown to be important predictors of health outcomes. Here we developed methods to extract them from inpatient electronic medical record (EMR) data using techniques compatible with current EMR systems. METHODS: Four social determinants were targeted: patient language barriers, employment status, education, and whether the patient lives alone. Inpatients aged 18 and older with records in the Calgary-wide EMR system were studied. Algorithms were developed on the January 2019 hospital admissions (n=8,999) and validated on the January 2018 hospital admissions (n=8,839). SDOH documented as structured data were compared against those extracted from unstructured free-text notes. RESULTS: More than twice as many patients had a note documenting a language barrier in EMR data than in structured data; 12 % of patients indicated by EMR notes to be living alone had a partner noted in their structured marital status. The Positive Predictive Value (PPV) of the elements extracted from notes was high, at 99 % (95 % CI 94.0 %-100.0 %) for language barriers, 98 % (95 % CI 92.6 %-99.9 %) for living alone, 96 % (95 % CI 89.8 %-98.8 %) for unemployment, and 88 % (95 % CI 80.0 %-93.1 %) for retirement. CONCLUSIONS: All SDOH elements were extracted with high PPV. SDOH documentation was largely missing in structured data and sometimes misleading.
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 | 0.007 | 0.034 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.010 | 0.004 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.001 |
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".