Derivation and Validation of Machine Learning Models for the Prevention of Unplanned Dialysis in Advanced CKD Patients
Bibliographic record
Abstract
Background: A short timeframe kidney failure risk prediction model may serve to prevent unplanned dialysis starts, a detrimental outcome associated with increased morbidity, mortality, and healthcare costs. To date, no such clinical tool exists. Methods: We developed and externally validated models for prediction of kidney failure over short timeframes of 6 and 12 months. The models were fit in 2,432 consecutive advanced CKD patients from The Ottawa Hospital. Models were externally validated in two independent advanced CKD cohorts from the Kingston General Hospital (N=724), and the Sunnybrook Health Sciences Center (N=323). All hospitals are in Ontario, Canada. Patients lost to follow-up, under conservative care management, or with <12 months of follow-up were excluded. Random forest classifiers were used to predict the 6- and 12-month probability of kidney failure for each patient at each follow-up visit. Input features included age, sex, and commonly available laboratory measurements and features characterizing their trajectory. The percentage of patients detected within clinically actionable timeframes of 3-15 months were computed for both models. Results: Internally, upon presentation, patients had a mean±SD age 66±15 years, and eGFR 18±7 mL/min/1.73m2, and median (IQR) ACR 164 (49, 333). Internal ROC-AUCs (95% CI) of 0.88 (0.87-0.88) and 0.86 (0.86-0.87) were achieved by the 6- and 12-month models, respectively. Models were well-calibrated. Internally, at 70% precision, patients requiring dialysis were correctly identified with at least 6 months advanced notice in 20% and 34% of cases using the 6- and 12-month models. Performance did not significantly differ at either external site. Conclusions: Machine learning-based short timeframe kidney failure risk prediction models accurately identify advanced CKD patients at high risk for imminent dialysis, including a substantial proportion destined for unplanned (or “crash”) dialysis initiation. Funding: Government Support - Non-U.S.
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 | 0.011 | 0.024 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.000 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| 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".