Retrieving randomized controlled trials from <scp>medline</scp>: a comparison of 38 published search filters
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
BACKGROUND: People search medline for trials of healthcare interventions for clinical decisions, or to produce systematic reviews, practice guidelines, or technology assessments. Finding all relevant randomized controlled trials (RCTs) with little extraneous material is challenging. OBJECTIVE: To provide comparative data on the operating characteristics of search filters designed to retrieve RCTs from medline. METHODS: We identified 38 filters. The testing database comprises handsearching data from 161 clinical journals indexed in medline. Sensitivity, specificity and precision were calculated. RESULTS: The number of terms and operating characteristics varied considerably. Comparing the retrieval against the single term 'randomized controlled trials.pt.' (sensitivity for retrieving RCTs, 93.7%), 24 of 38 filters had statistically higher sensitivity; 6 had a sensitivity of at least 99.0%. Four other filters had specificities (non retrieval of non-RCTs) that were statistically not different or better than the single term (97.6%). Precision was poor: only two filters had precision (proportion of retrieved articles that were RCTs) statistically similar to that of the single term (56.4%)-all others were lower. Filters with more search terms often had lower specificity, especially at high sensitivities. CONCLUSION: Many RCT filters exist (n = 38). These comparative data can direct the choice of an RCT filter.
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.
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 | 0.416 | 0.480 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.021 | 0.005 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.008 | 0.008 |
| Open science | 0.002 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.007 | 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 it