Database combinations to retrieve systematic reviews in overviews of reviews: a methodological study
Bibliographic record
Abstract
BACKGROUND: When conducting an Overviews of Reviews on health-related topics, it is unclear which combination of bibliographic databases authors should use for searching for SRs. Our goal was to determine which databases included the most systematic reviews and identify an optimal database combination for searching systematic reviews. METHODS: A set of 86 Overviews of Reviews with 1219 included systematic reviews was extracted from a previous study. Inclusion of the systematic reviews was assessed in MEDLINE, CINAHL, Embase, Epistemonikos, PsycINFO, and TRIP. The mean inclusion rate (% of included systematic reviews) and corresponding 95% confidence interval were calculated for each database individually, as well as for combinations of MEDLINE with each other database and reference checking. RESULTS: Inclusion of systematic reviews was higher in MEDLINE than in any other single database (mean inclusion rate 89.7%; 95% confidence interval [89.0-90.3%]). Combined with reference checking, this value increased to 93.7% [93.2-94.2%]. The best combination of two databases plus reference checking consisted of MEDLINE and Epistemonikos (99.2% [99.0-99.3%]). Stratification by Health Technology Assessment reports (97.7% [96.5-98.9%]) vs. Cochrane Overviews (100.0%) vs. non-Cochrane Overviews (99.3% [99.1-99.4%]) showed that inclusion was only slightly lower for Health Technology Assessment reports. However, MEDLINE, Epistemonikos, and reference checking remained the best combination. Among the 10/1219 systematic reviews not identified by this combination, five were published as websites rather than journals, two were included in CINAHL and Embase, and one was included in the database ERIC. CONCLUSIONS: MEDLINE and Epistemonikos, complemented by reference checking of included studies, is the best database combination to identify systematic reviews on health-related topics.
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 | 0.916 | 0.990 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.020 | 0.002 |
| Bibliometrics | 0.002 | 0.013 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.007 | 0.002 |
| Research integrity | 0.000 | 0.002 |
| Insufficient payload (model declined to judge) | 0.020 | 0.006 |
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".