Multivariable model versus AJCC staging system: cancer-specific survival predictions in adrenocortical carcinoma
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
We developed a novel contemporary population-based model for predicting cancer-specific survival (CSS) in adrenocortical carcinoma (ACC) patients and compared it with the established 8th edition of the American Joint Committee on Cancer staging system (AJCC). Within the Surveillance, Epidemiology, and End Results database (2004-2020), we identified 1056 ACC patients. Univariable Cox regression model addressed CSS. Harrell's concordance index (C-index) quantified accuracy after 2000 bootstrap resamples for internal validation. The multivariable Cox regression model included the most informative, statistically significant predictors. Calibration and decision curve analyses (DCAs) tested the multivariable model as well as AJCC in head-to-head comparisons. Age at diagnosis (>60 vs ≤60 years), surgery, T, N, and M stages were included in the multivariable model. Multivariable model C-index for 3-year CSS prediction was 0.795 vs 0.757 for AJCC. Multivariable model outperformed AJCC in DCAs for the majority of possible CSS-predicted values. Both models exhibited similar calibration properties. Finally, the range of the multivariable model CSS predicted probabilities raged 0.02-75.3% versus only four single AJCC values, specifically 73.2% for stage I, 69.7% for stage II, 46.6% for stage III, and 15.5% for stage IV. The greatest benefit of the multivariable model-generated CSS probabilities applied to AJCC stage I and II patients. The multivariable model was more accurate than AJCC staging when CSS predictions represented the endpoint. Additionally, the multivariable model outperformed AJCC in DCAs. Finally, the AJCC appeared to lag behind the multivariable model when discrimination addressed AJCC stage I and II patients.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 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 it