Creating the foundation data for building a population grouping methodology – lessons learned at the Canadian Institute for Health Information (CIHI)
Bibliographic record
Abstract
Background In April 2013, CIHI initiated a project to develop a population grouping methodology that stratifies a population based on past clinical information and produces risk measures (i.e., costs for the prospective year). The methodology includes all individuals in the population at a given moment, including those who are not using the health system. The foundation data for this project include historical clinical administrative and utilization data that are linkable at the individual level. Ideally, such a methodology includes multiple years of data that cover multiple health sectors and the full population. For example, if two years of data are used to develop clinical profiles, and if predictive indicators aim to estimate need one year in the future, then three consecutive years of data are needed. It is also important to assess the stability of the predictive indicators over time and to determine the optimal historical review period for the clinical classification; as a result, there is a need for additional years of data for such a project. One goal for this project was for the methodology to be useful to the majority of Canadian provinces. A province will not benefit from any methodology where personlevel linkable clinical information are not available to apply (vs. develop) the methodology. Clinical data only are needed to apply the final methodology, so they need to be of a consistently high quality across the provinces. This rationale influenced the choice of health sectors to incorporate into the methodology. CIHI does not have full pan-Canadian coverage of patient-level linkable clinical data for all health sectors. Some sectors have partial or no coverage within some provinces (e.g., emergency department, home care, long-term care, drug prescriptions). Some sectors have full coverage within the provinces, but CIHI does not have access to patientlevel linkable data (e.g., physician billing). Additionally, while data coverage may be sufficient in the present, historical data can be limited. These coverage issues influenced the sectors and provinces included in the foundation data.
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 | 0.204 | 0.324 |
| Meta-epidemiology (narrow) | 0.002 | 0.002 |
| Meta-epidemiology (broad) | 0.002 | 0.002 |
| Bibliometrics | 0.016 | 0.025 |
| Science and technology studies | 0.007 | 0.006 |
| Scholarly communication | 0.013 | 0.008 |
| Open science | 0.012 | 0.009 |
| Research integrity | 0.003 | 0.007 |
| Insufficient payload (model declined to judge) | 0.014 | 0.005 |
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; the direct Gemma label and the distilled Codex classifier agree on what is shown here.
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".