Protocol for the validation of four search strategies for retrieval of Clinical Practice Guidelines in MEDLINE, Embase and PubMed
Bibliographic record
Abstract
Background Guidelines are systematically developed recommendations to assist practitioner and patient decisions about treatments for clinical conditions. Researchers, healthcare professionals and policy makers need to be able to retrieve clinical practice guidelines (CPGs) efficiently and quickly from the literature. Despite the widespread use of CPGs in practice and policy formulation, no filter for retrieval of guidelines has been validated to date. The use of a validated search filter for CPGs would make their retrieval from major bibliographic databases more efficient. Objectives We aim to fill this gap by validating search filters for use in the systematic retrieval of CPGs and measure their performance according to sensitivity and precision. Methods We found four search filters for retrieval of CPGs (two CADTH, PubMed and University of Texas filters) which we will validate in three databases (MEDLINE, Embase and PubMed). We will derive a test set of CPGs from a search of the TRIP and Epistemonikos databases. The citations retrieved will be randomly sorted and screened sequentially by two reviewers until at least 100 CPGs are included. We will include CPGs that provide at least two explicit recommendations for the treatment of any clinical condition, and that are produced by a group or organization (i.e., not authored by one person) . We will translate the filters into Ovid MEDLINE, Embase, and PubMed syntax as appropriate. Then, we will run the strategies and assess whether the filters retrieved the citations in our test set. We will calculate and compare the sensitivity and precision of the four filters in each database. The limitations of the CADTH, PubMed and University of Texas search filters for each database will be assessed by examining the keywords in the titles and abstracts of the citations not found by the search filters. Discussion Decision makers, healthcare providers and researcher will be able to choose the most precise and sensitive search strategy among the four available, which will enable them to more efficiently identify relevant clinical practice guidelines.
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 | 0.316 | 0.458 |
| Meta-epidemiology (narrow) | 0.007 | 0.010 |
| Meta-epidemiology (broad) | 0.016 | 0.016 |
| Bibliometrics | 0.027 | 0.023 |
| Science and technology studies | 0.007 | 0.007 |
| Scholarly communication | 0.009 | 0.010 |
| Open science | 0.007 | 0.008 |
| Research integrity | 0.012 | 0.010 |
| Insufficient payload (model declined to judge) | 0.124 | 0.033 |
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".