Randomized trials stopped early for benefit: a systematic review.
Bibliographic record
Abstract
CONTEXT: Randomized clinical trials (RCTs) that stop earlier than planned because of apparent benefit often receive great attention and affect clinical practice. Their prevalence, the magnitude and plausibility of their treatment effects, and the extent to which they report information about how investigators decided to stop early are, however, unknown. OBJECTIVE: To evaluate the epidemiology and reporting quality of RCTs involving interventions stopped early for benefit. DATA SOURCES: Systematic review up to November 2004 of MEDLINE, EMBASE, Current Contents, and full-text journal content databases to identify RCTs stopped early for benefit. STUDY SELECTION: Randomized clinical trials of any intervention reported as having stopped early because of results favoring the intervention. There were no exclusion criteria. DATA EXTRACTION: Twelve reviewers working independently and in duplicate abstracted data on content area and type of intervention tested, reporting of funding, type of end point driving study termination, treatment effect, length of follow-up, estimated sample size and total sample studied, role of a data and safety monitoring board in stopping the study, number of interim analyses planned and conducted, and existence and type of monitoring methods, statistical boundaries, and adjustment procedures for interim analyses and early stopping. DATA SYNTHESIS: Of 143 RCTs stopped early for benefit, the majority (92) were published in 5 high-impact medical journals. Typically, these were industry-funded drug trials in cardiology, cancer, and human immunodeficiency virus/AIDS. The proportion of all RCTs published in high-impact journals that were stopped early for benefit increased from 0.5% in 1990-1994 to 1.2% in 2000-2004 (P<.001 for trend). On average, RCTs recruited 63% (SD, 25%) of the planned sample and stopped after a median of 13 (interquartile range [IQR], 3-25) months of follow-up, 1 interim analysis, and when a median of 66 (IQR, 23-195) patients had experienced the end point driving study termination (event). The median risk ratio among truncated RCTs was 0.53 (IQR, 0.28-0.66). One hundred thirty-five (94%) of the 143 RCTs did not report at least 1 of the following: the planned sample size (n = 28), the interim analysis after which the trial was stopped (n = 45), whether a stopping rule informed the decision (n = 48), or an adjusted analysis accounting for interim monitoring and truncation (n = 129). Trials with fewer events yielded greater treatment effects (odds ratio, 28; 95% confidence interval, 11-73). CONCLUSIONS: RCTs stopped early for benefit are becoming more common, often fail to adequately report relevant information about the decision to stop early, and show implausibly large treatment effects, particularly when the number of events is small. These findings suggest clinicians should view the results of such trials with skepticism.
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 | 0.095 | 0.358 |
| Meta-epidemiology (narrow) | 0.003 | 0.003 |
| Meta-epidemiology (broad) | 0.015 | 0.009 |
| Bibliometrics | 0.020 | 0.020 |
| Science and technology studies | 0.001 | 0.003 |
| Scholarly communication | 0.005 | 0.008 |
| Open science | 0.003 | 0.002 |
| Research integrity | 0.005 | 0.004 |
| Insufficient payload (model declined to judge) | 0.005 | 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".