Table 2

Combination of Terms with the Best Sensitivity, Best Specificity, and Best Optimization of Sensitivity and Specificity for Detecting Studies of Diagnosis in EMBASE in 2000. Values are percentages (95% confidence intervals).

Search Strategyn OVID search*

Sensitivity (n = 97)

Specificity (n = 27672)

Precision†

Accuracy (n = 27769)


Best Sensitivity (keeping specificity ≥ 50%)

di.fs.

OR predict:.tw.

OR specificity.tw.

100.0 (100.0 to 100.0)

70.4 (69.8 to 70.9)

1.2 (0.9 to 1.4)

70.5 (69.9 to 71.0)

Small drop in sensitivity with a substantive gain in specificity

diagnos:.mp.

OR predict:.tw.

OR specificity.tw.

96.9 (93.5 to 100.0)

78.2 (77.7 to 78.7)

1.5 (1.2 to 1.8)

78.3 (77.8 to 78.8)

Best Specificity (keeping sensitivity ≥ 50%)

specificity.tw.

62.9 (53.5 to 72.5)

98.2 (98.1 to 98.4)

11.0 (8.4 to 13.6)

98.1 (97.9 to 98.3)

Small drop in specificity with a substantive gain in sensitivity

specificity.tw.

OR accurac:.tw.

73.2 (64.4 to 82.0)

97.4 (97.2 to 97.5)

8.8 (6.9 to 10.8)

97.3 (97.1 to 97.5)

Best Optimization of Sensitivity & Specificity‡

sensitiv:.tw.

OR diagnostic accuracy.sh.

OR diagnostic.tw.

89.7 (83.6 to 95.7)

91.6 (91.3 to 91.9)

3.3 (2.9 to 4.4)

91.6 (91.3 to 91.9)


*Search strategies are reported using Ovid's search engine syntax for EMBASE. †Denominator varies by row. ‡Based on the lowest possible absolute difference between sensitivity and specificity. di = diagnosis; fs = floating subheading; : = truncation; tw = textword (word or phrase appears in title or abstract); mp = multiple posting – term appears in title, abstract, or subject heading; sh = subject heading. Sensitivity = the proportion of high quality articles for that topic that are retrieved; specificity = the proportion of low quality articles not retrieved; precision = the proportion of retrieved articles that are of high quality; accuracy = the proportion of all articles that are correctly classified.

Wilczynski et al. BMC Medicine 2005 3:7   doi:10.1186/1741-7015-3-7

Open Data