Ablate and pace for patients with atrial fibrillation: a fragile option?
Bibliographic record
Abstract
This commentary refers to ‘AV junction ablation and cardiac resynchronization for patients with permanent atrial fibrillation and narrow QRS: the APAF-CRT mortality trial’, by M. Brignole et al., https://doi.org/10.1093/eurheartj/ehab569 and the discussion piece ‘Is APAF-CRT robust enough to change clinical practice?’, by M. Brignole et al., https://doi.org/10.1093/eurheartj/ehac092. In the Ablate and Pace for Atrial Fibrillation—cardiac resynchronization therapy (APAF-CRT) trial, Brignole et al. compared pharmacological rate control (drug) with ablation and biventricular pacing (ablation + CRT) in patients with severely symptomatic AF lasting >6 months, narrow QRS, and at least one prior heart failure hospitalization. In this multicentre, open-label trial, 140 patients were randomized and 7 were excluded from the primary analysis (2 withdrew consent and 5 were lost to follow-up). Of the 133 included in the final intention-to-treat analysis, 70 patients were allocated to pharmacological rate control and 63 to ablation + CRT. The Data Safety Monitoring Board stopped the trial at a median follow-up of 29 months. At the time of termination, the primary endpoint—all-cause mortality—had occurred in 7 patients in the ablation + CRT arm and 20 in the drug arm (hazard ratio: 0.26; 95% confidence interval: 0.10–0.65; P = 0.004). This was lower than the estimated 32 events required for 80% power in detecting a 70% reduction in hazards for mortality at a significance of 0.042.1 Randomization protects against confounding, but it cannot account for the play of chance. Randomized clinical trials (RCTs) in which only a small number of events occur are vulnerable to random error, and the results could change based on the status of a few participants. The fragility index is the minimum number of participants required to change status for a statistically significant result to lose significance. A low fragility index reduces confidence in the treatment effect, regardless of statistical significance.2 This is especially relevant for trials where loss to follow-up exceeds the fragility index.2 The fragility index of the APAF-CRT trial is 2, calculated by iteratively adding events and subtracting non-events from the ablation + CRT group until the P-value exceeded 0.05.1 This indicates that two additional deaths in the ablation + CRT group would have resulted in a loss of statistical significance. According to a review of 399 RCTs from five high-impact journals, this fragility index places the APAF-CRT trial below the 25th percentile among RCTs and calls into question the validity of its results.2 Regardless of methodological rigour, the results of similarly fragile RCTs have subsequently been contrasted by lower or no treatment effects when tested in larger samples. For example, the Leicester Intravenous Magnesium Intervention Trial (n = 2316) had a fragility index of 1, and its results were refuted in a subsequent larger trial and meta-analyses.3,4 Considering the global burden of AF and heart failure, studies that conclusively determine treatments effective to reduce mortality are important. Large RCTs that mitigate the challenges of the fragility associated with smaller studies are required to inform clinical practice. This study by Brignole et al. reinforces the importance of reporting the fragility index of RCTs to support informed clinical decision-making. Conflict of interest: Outside the submitted work, Dr McIntyre reports speaking fees from Servier, Bayer and Boehringer Ingelheim and Dr. Belley-Côté reports grants from Bayer, BMS-Pfizer, and Roche.
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 | 0.003 | 0.029 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.002 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.003 | 0.003 |
| Scholarly communication | 0.005 | 0.008 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.039 | 0.036 |
| Insufficient payload (model declined to judge) | 0.013 | 0.005 |
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".