How to Measure the Generalizability of Clinical Trials
Bibliographic record
Abstract
Randomized controlled trials are widely regarded as the gold standard in clinical research and public health. However, they have been criticized for potentially lacking generalizability, as trial participants may not fully represent the target patient population due to the inability to obtain a truly random sample for enrollment. Assessing and evaluating the generalizability of randomized controlled trials is an important issue that has not been addressed adequately in literature. Additionally, although the importance of describing clinical trial generalizability is recognized by clinical trial reporting guidelines (e.g., CONSORT), it provides no clear guidance on statistical tests or estimation procedures. In this paper, we compare five generalizability indexes, including Standardized Mean Difference, C-Statistic, β-Index, Kolmogorov-Smirnov Distance, and Lévy Distance. We simulate a patient population with a treatment effect size of 0.5 (Cohen's d ) and seven covariates that include gender, health insurance, race, baseline symptoms, comorbidity, age, and motivation. We then evaluate the performance of the five generalizability indexes using selected nonrandom and random clinical trial samples under different number of covariates and sample sizes. Our work supports the use of -index and C-statistic due to their strong statistical performance, ease of interpretation and ability to clearly categorize generalizability into levels such as very high, high, medium or low. A -index value between 1 and 0.8 (inclusive) or a C-statistic value between 0.5 and 0.8 (inclusive) indicates that the trail sample is very highly or highly representative of the patient 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 | 0.634 | 0.894 |
| Meta-epidemiology (narrow) | 0.003 | 0.002 |
| Meta-epidemiology (broad) | 0.007 | 0.012 |
| Bibliometrics | 0.011 | 0.010 |
| Science and technology studies | 0.002 | 0.011 |
| Scholarly communication | 0.011 | 0.013 |
| Open science | 0.005 | 0.007 |
| Research integrity | 0.008 | 0.010 |
| 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; 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".