Bounding the Implications of Noncompliance in Randomized Controlled Trials in Orthopaedics: An Example in Arthroscopic Surgery
Bibliographic record
Abstract
INTRODUCTION: Randomized controlled trials (RCTs) are not impervious to bias especially when there are substantial numbers of patients who cross over from the treatment assigned by randomization to another treatment group, leading to loss of confidence in study results. The goals of this study were to (1) quantify the effects of crossovers on RCTs, (2) describe the specific effects of crossovers on RCTs for arthroscopic meniscectomy for osteoarthritis of the knee (APM/OAK), and (3) assess the confidence in APM/OAK in which there have been substantial numbers of patients crossing over to another treatment group than that assigned. METHODS: Studies were included that were RCTs of APM/OAK with intention-to-treat (ITT) analysis and illustrated the problem of crossovers on confidence in the analysis. Studies were excluded if they consisted of APM for conditions other than OAK or had unavailability of data needed for the analysis. For eligible RCTs, the ITT effect was calculated; bounds for the average treatment effect (ATE) and the complier ATE were assessed by estimating confidence intervals for the bound through robust Bayesian analysis. RESULTS: The eligible studies had different comparators and, therefore, were analyzed individually. Data were not pooled. The most extreme point estimates (with 95% confidence interval) for ITT ranged from -0.01 to 0.04 (-0.16 to 0.16); for ATE with no assumptions, 0.38 (-0.58 to 0.43) to 0.62 (0.56 to 0.70); for ATE with minimum assumptions, -0.50 (-0.22 to 0.10) to 0.61 (0.53 to 0.57); and for complier ATE, -0.01 to 0.07 (-0.22 to 0.24). DISCUSSION: These data suggest large bounds, crossing the threshold of "no effect," which indicates a high degree of uncertainty and low confidence in the RCTs studied. The results demonstrate that when there are crossovers, ITT analyses do not estimate the ATE and confidence in the results of these RCTs is low. DATA AVAILABILITY: All analyzed data are provided in the article. LEVEL OF EVIDENCE: Level I (therapeutic study = RCT).
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 | 0.654 | 0.866 |
| Meta-epidemiology (narrow) | 0.004 | 0.002 |
| Meta-epidemiology (broad) | 0.010 | 0.011 |
| Bibliometrics | 0.006 | 0.008 |
| Science and technology studies | 0.003 | 0.023 |
| Scholarly communication | 0.012 | 0.015 |
| Open science | 0.006 | 0.008 |
| Research integrity | 0.013 | 0.023 |
| Insufficient payload (model declined to judge) | 0.003 | 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; the direct Gemma label and the distilled Codex classifier agree on what is shown here.
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".