SuHeSS | FreeWoSS | Intercept of FreeWoSS and SuHeSSa | |
---|---|---|---|
Performance indicators (with 95% CI) of the search strategy for obtaining articles used for full-text estimation in comparison with the ‘gold standard’ set | |||
Sensitivity | 85.3 (81.3–88.7) | 94.5 (91.7–96.5) | 79.8 (75.4–83.7) |
Specificity | 55.9 (55.3–56.5) | 20.9 (20.5–21.4) | 76.8 (76.3–77.3) |
Precision | 2.5 (2.2–2.7) | 1.5 (1.4–1.7) | 4.3 (3.8–4.8) |
Accuracy | 56.3 (55.7–56.8) | 21.9 (21.4–22.4) | 76.8 (76.4–77.3) |
Number of bibliographic records excluded by the search strategy from the overall PubMed ‘gold standard’ set at a given step | |||
Initial search | 16,313 | 6108 | 22,421 |
Potentially useful abstracts | 612 | 250 | 862 |
Selected for full-text extraction | 66 | 26 | 92 |
Obtained as a full-text and data extracted | 56 | 21 | 77 |
Percentage of bibliographic records excluded by the search strategy from the overall PubMed ‘gold standard’ set at a given step | |||
Initial search | 55.4 | 20.7 | 76.1 |
Potentially useful abstracts | 29.8 | 12.2 | 41.9 |
Selected for full-text extraction | 14.7 | 5.8 | 20.4 |
Obtained as a full text and data extracted | 14.7 | 5.5 | 20.2 |
Extracted data rows excluded by the search strategy from the overall data rows extracted from PubMed full-text articles | |||
Number of data rows excluded by the strategy | 1222 | 117 | 1339 |
Percentage of data rows excluded by the strategy | 15.8 | 1.5 | 17.3 |