#4486 VALIDATION OF A CKD PROGRESSION RISK PREDICTION MODEL IN THE FIDELITY TRIAL POPULATION
Bibliographic record
Abstract
Abstract Background and Aims Chronic kidney disease (CKD) is often underrecognised until later stages when most of the kidney function is lost and the therapeutic window for disease-modifying therapy is narrow [1]. We previously developed a lab-based risk prediction model to accurately predict CKD progression in adults at all stages of CKD [2]. Here, we describe a validation of our model in the clinical trial population of FIDELITY, a prespecified pooled analysis of the phase III FIDELIO-DKD (NCT02540993) and FIGARO-DKD (NCT02545049) trials for the nonsteroidal mineralocorticoid receptor antagonist finerenone [3]. Method We performed a post hoc analysis of all participants from the FIDELITY database, irrespective of estimated glomerular filtration rate (eGFR) or albuminuria stage. Baseline values for the underlying laboratory tests required for the model, Klinrisk, were extracted from the complete blood count, comprehensive metabolic panel and urine albumin-to-creatinine ratio (UACR). The predicted outcome was a ≥40% decline in eGFR or kidney failure. We calculated discrimination ability of the model and calibration using area under the curve (AUC), Brier scores and calibration plots in the overall population, and stratified by treatment assignment. Sensitivity analyses examined the accuracy of the models in predicting ≥57% decline in eGFR, as well as the change in risk score over time. Kidney Disease: Improving Global Outcomes (KDIGO) heat map categories were used as the reference standard. Results We included 13,026 participants with a mean age of 64.8 ± 9.5 years, mean eGFR of 57.6 ± 21.7 ml/min/1.73 m2, and median UACR of 58.2 mg/mmol (interquartile range 22.4–129.6). At time horizons of 2 and 4 years, 984 and 1795 patients experienced a primary outcome event, respectively. The Klinrisk model predicted progression accurately, with an AUC of 0.81 (95% confidence interval [CI] 0.79–0.82) at 2 years and 0.86 (95% CI 0.84–0.87) at 4 years, compared with the KDIGO heatmap categories (AUC of 0.59 [95% CI 0.58–0.60] at 2 years and 0.66 [95% CI 0.65–0.68] at 4 years). Calibration was appropriate (Brier score of 0.067 [95% CI 0.064–0.070] at 2 years and 0.115 [95% CI 0.109–0.120] at 4 years). Similar discrimination accuracy was seen for the ≥57% decline outcome (C-statistic 0.88, 95% CI 0.87–0.90) at 3 years. Conclusion Based on routinely collected lab data, our machine learning model (Klinrisk) accurately predicts CKD progression events in a well characterized global clinical trial population. Prospective implementation of the model in clinical trial enrolment as well as clinical care pathways is needed.
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 | 0.042 | 0.069 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.003 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.006 | 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".