External Validation of Prediction Models for Unilateral Primary Aldosteronism
Bibliographic record
Abstract
Abstract Primary aldosteronism (PA) is the most common cause of remediable hypertension. Treatment is informed by establishing whether disease is unilateral (localized to one adrenal gland) or bilateral. Adrenalectomy is the guideline-recommended treatment of choice for unilateral PA. However, the currently recommended subtyping test, adrenal vein sampling (AVS), is often limited in accessibility. Thus, prediction models have been developed to diagnose unilateral PA and therefore bypass AVS. However, their generalizability remains unknown. In this retrospective study, we aimed to externally validate the performance of prediction models for unilateral PA in a large population of PA patients at a Canadian referral center who underwent AVS during 2006–2018. The presence of unilateral disease was indicated by a lateralization index of >3 on AVS. We identified 6 clinical prediction models from the literature. The discrimination and calibration of each model were systematically evaluated. For the original models, the derivation cohorts were based out of Japan, France, Italy, and England, with mean age between 46–54 years and 43–56% being male. The derivation cohorts were generally small, with 4 of the 6 studies reporting less than 50 people with unilateral PA. Common variables reported to be predictive of unilateral PA included male sex, hypokalemia, elevated aldosterone-renin ratio, and the presence of a unilateral adrenal nodule on imaging. The validation cohort included 342 PA patients who underwent successful AVS (average age, 52.1 years; 58.8% male). Among them, 186 (54.4%) demonstrated unilateral disease, and the remaining 156 (45.6%) were considered to have bilateral disease. The baseline characteristics of the validation cohort were broadly similar to those of the derivation cohorts, except for potential differences in ethnicity. When applying the models to the validation cohort, subjects were excluded if any candidate variables were missing. All 6 models demonstrated poor discrimination in the validation set (C-statistics; range, 0.59–0.72), representing a marked decrease compared to the derivation sets where they were reported (range, 0.80–0.87). Assessment of calibration by comparing observed and predicted probabilities of the unilateral subtype revealed significant miscalibration. Calibration-in-the-large for every model was >0 (range, 0.36–2.23), signifying systematic underprediction of unilateral PA. Calibration slopes were all <1 (range, 0.35–0.85), indicating poor performance at the extremes of risk. These results suggest that the original models were optimistic due to overfitting in the derivation cohorts and therefore lack generalizability. This is primarily because these models were developed in small data sets. In conclusion, clinical assessment with prediction models for unilateral PA cannot be readily used to bypass AVS in the general PA population.
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 | 0.059 | 0.123 |
| Meta-epidemiology (narrow) | 0.002 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.003 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 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) | 0.002 | 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".