The ADHERE classification and regression tree model overestimates mortality rates in clinical trials: results from REVIVE I & II
Bibliographic record
Abstract
Blood urea nitrogen (BUN), systolic blood pressure (SBP), and serum creatinine (Cr) were significant predictors of inhospital mortality by classification and regression tree (CART) analysis of ADHERE. REVIVE I & II (REVIVE) compared levosimendan with placebo, in addition to standard-of-care (SOC), in patients with acute decompensated heart failure. We hypothesized that mortality in REVIVE would be similar to ADHERE in all CART-defined risk subgroups. REVIVE ( n = 700) mortality data were mapped using the same variables/cut-points as the ADHERE CART analysis. Compared with ADHERE, proportionately more patients in REVIVE had SBP <115 mmHg (56.4% vs 18.6%; P < 0.001) with more patients (3.0% vs 1.9%; P < 0.05) in the highest mortality risk subgroup (SBP <115 mmHg, BUN ≥ 43 mg/dl, and Cr ≥ 2.75 mg/dl). For the total population and for every CART-defined subgroup, REVIVE inhospital mortality rates were lower than those from ADHERE. See Figure 1 . Mortality rates from REVIVE for subgroups defined by the ADHERE classification and regression tree model. Clinical trials (REVIVE) may enroll proportionately more patients at increased risk of mortality in comparison with the general population (ADHERE). Despite the predicted increased mortality risk, mortality rates were lower in REVIVE than in ADHERE for the total population and for every CART-defined risk subgroup. Differences in SOC or additional risk factors, such as age or other comorbid conditions, may contribute to the poorer prognosis in nontrial populations.
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 | 0.091 | 0.140 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.003 | 0.008 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.002 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.003 |
| 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".