Meta-Analyses of Diagnostic Accuracy in Imaging Journals: Analysis of Pooling Techniques and Their Effect on Summary Estimates of Diagnostic Accuracy
Bibliographic record
Abstract
Purpose To determine whether authors of systematic reviews of diagnostic accuracy studies published in imaging journals used recommended methods for meta-analysis, and to evaluate the effect of traditional methods on summary estimates of sensitivity and specificity. Materials and Methods Medline was searched for published systematic reviews that included meta-analysis of test accuracy data limited to imaging journals published from January 2005 to May 2015. Two reviewers independently extracted study data and classified methods for meta-analysis as traditional (univariate fixed- or random-effects pooling or summary receiver operating characteristic curve) or recommended (bivariate model or hierarchic summary receiver operating characteristic curve). Use of methods was analyzed for variation with time, geographical location, subspecialty, and journal. Results from reviews in which study authors used traditional univariate pooling methods were recalculated with a bivariate model. Results Three hundred reviews met the inclusion criteria, and in 118 (39%) of those, authors used recommended meta-analysis methods. No change in the method used was observed with time (r = 0.54, P = .09); however, there was geographic (χ(2) = 15.7, P = .001), subspecialty (χ(2) = 46.7, P < .001), and journal (χ(2) = 27.6, P < .001) heterogeneity. Fifty-one univariate random-effects meta-analyses were reanalyzed with the bivariate model; the average change in the summary estimate was -1.4% (P < .001) for sensitivity and -2.5% (P < .001) for specificity. The average change in width of the confidence interval was 7.7% (P < .001) for sensitivity and 9.9% (P ≤ .001) for specificity. Conclusion Recommended methods for meta-analysis of diagnostic accuracy in imaging journals are used in a minority of reviews; this has not changed significantly with time. Traditional (univariate) methods allow overestimation of diagnostic accuracy and provide narrower confidence intervals than do recommended (bivariate) methods. (©) RSNA, 2016 Online supplemental material is available for this article.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.092 | 0.522 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.042 | 0.013 |
| Bibliometrics | 0.006 | 0.005 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.003 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
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; both teacher heads 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".