Table 1 Characteristics of meta-analyses included in the review (n=53)
From: Assessing variability in results in systematic reviews of diagnostic studies
Characteristic | N or median | % or IQR |
|---|---|---|
Median number of primary studies | 14 | [9.5–18.5] |
Median number of patients in primary studies | 87 | [45–182] |
Type of test tests under study | ||
Laboratory tests | 15 | 28 % |
Image tests | 32 | 60 % |
Clinical examination | 6 | 11 % |
Meta-analyses looking at more than one test | 31 | 58 % |
Method(s) for conducting the meta-analysis† | ||
Univariate analysis only | 13 | 25 % |
SROC (Moses Littenberg): linear regression D on S | 24 | 45 % |
HSROC (Rutter and Gatsonis): accuracy, scale and threshold parameter | 5 | 9 % |
Bivariate random effects model (Reitsma): random effects sens & spec | 13 | 25 % |