Development of a search filter for identifying studies completed in primary care
Bibliographic record
Abstract
BACKGROUND: Identifying articles relevant to primary care is challenging for busy clinicians. Setting specific search strategies can be used to help clinicians find pertinent studies in a timely fashion. OBJECTIVES: To develop search filters for identifying research studies of relevance to primary care in MEDLINE (OvidSP). METHODS: We conducted a search of MEDLINE (OvidSP) for articles published in five core medical journals at five yearly intervals. We identified a gold standard set of primary care relevant articles which was divided into two subsets. The first subset was used to identify frequently occurring words and phrases through textual analysis. Search filters were developed from these words and phrases and internally validated against records in the second subset. We evaluated the filters performance in a search for articles on two common primary care conditions in MEDLINE (OvidSP). RESULTS: Of the 12 045 articles retrieved, 9028 records were reviewed, of which 371 articles were relevant to primary care (gold standard). When the search filters generated from textual analysis were internally validated, filter specificity peaked at 99% with 60% sensitivity, 67% precision and 97% accuracy. When evaluated against a set of articles on two common primary care conditions, the best performing combination search filter specificity maximized at 99.7% with sensitivity reaching 15% (precision 90%; accuracy 89%). CONCLUSION: The best performing combination search filter works well in reducing the number of irrelevant papers retrieved in a MEDLINE (OvidSP) search if a busy clinician needs to focus on research relevant to primary care.
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 | 0.167 | 0.461 |
| Meta-epidemiology (narrow) | 0.004 | 0.002 |
| Meta-epidemiology (broad) | 0.007 | 0.009 |
| Bibliometrics | 0.063 | 0.023 |
| Science and technology studies | 0.003 | 0.001 |
| Scholarly communication | 0.006 | 0.006 |
| Open science | 0.004 | 0.004 |
| Research integrity | 0.005 | 0.002 |
| Insufficient payload (model declined to judge) | 0.007 | 0.002 |
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; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
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".