Influence of Mortality on Estimating the Risk of Kidney Failure in People with Stage 4 CKD
Bibliographic record
Abstract
Significance Statement Risk calculators are widely used to predict kidney failure in people with CKD. These tools represent major advances compared with reporting eGFR alone because they provide objective risk assessment. These calculators are based on analysis methods that censor for death, which may result in overestimation of the risk of kidney failure. By analyzing data on a large population of patients with stage 4 CKD, the authors show that kidney failure risk estimates that censor for death consistently exceed the estimates from analyses that treat death as a competing risk, by 1%–27% at 5 years. Risk overestimation with methods that censor for death increases over time and is higher in people with more comorbidities. Not treating death as a competing risk in the risk calculator leads to overestimates of the risk of kidney failure in people with stage 4 CKD, which could have negative psychological effects on patients or contribute to overtreatment. Background Most kidney failure risk calculators are based on methods that censor for death. Because mortality is high in people with severe, nondialysis-dependent CKD, censoring for death may overestimate their risk of kidney failure. Methods Using 2002–2014 population-based laboratory and administrative data for adults with stage 4 CKD in Alberta, Canada, we analyzed the time to the earliest of kidney failure, death, or censoring, using methods that censor for death and methods that treat death as a competing event factoring in age, sex, diabetes, cardiovascular disease, eGFR, and albuminuria. Stage 4 CKD was defined as a sustained eGFR of 15–30 ml/min per 1.73 m 2 . Results Of the 30,801 participants (106,447 patient-years at risk; mean age 77 years), 18% developed kidney failure and 53% died. The observed risk of the combined end point of death or kidney failure was 64% at 5 years and 87% at 10 years. By comparison, standard risk calculators that censored for death estimated these risks to be 76% at 5 years and >100% at 7.5 years. Censoring for death increasingly overestimated the risk of kidney failure over time from 7% at 5 years to 19% at 10 years, especially in people at higher risk of death. For example, the overestimation of 5-year absolute risk ranged from 1% in a woman without diabetes, cardiovascular disease, or albuminuria and with an eGFR of 25 ml/min per 1.73 m 2 (9% versus 8%), to 27% in a man with diabetes, cardiovascular disease, albuminuria >300 mg/d, and an eGFR of 20 ml/min per 1.73 m 2 (78% versus 51%). Conclusions Kidney failure risk calculators should account for death as a competing risk to increase their accuracy and utility for patients and providers.
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 | 0.048 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.003 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.002 |
| 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".