Table 5.
bivariate random-effect meta-regression with DL/ML as covariate
| Point Estimate | SE | z value | p value | 95% CI lower | 95% CI upper | |
|---|---|---|---|---|---|---|
| All studies (n = 16) | ||||||
| DOR (higher is better) | 0.464 | 0.334 | 1.387 | 0.165 | −0.192 | 1.119 |
| Sens (higher is better) | 0.068 | 0.242 | 0.281 | 0.779 | −0.406 | 0.542 |
| Fpr (lower is better) | −0.419 | 0.195 | 2.149 | 0.032* | −0.801 | −0.037 |
| ASD (n = 4) | ||||||
| DOR (higher is better) | 0.347 | 0.128 | 2.721 | 0.007** | 0.097 | 0.597 |
| Sens (higher is better) | 0.181 | 0.203 | 0.888 | 0.374 | −0.218 | 0.579 |
| Fpr (lower is better) | −0.162 | 0.183 | −0.884 | 0.377 | −0.520 | 0.197 |
| SZ (n = 9) | ||||||
| DOR (higher is better) | 0.601 | 0.331 | 1.814 | 0.070 | −0.048 | 1.250 |
| Sens (higher is better) | 0.086 | 0.328 | 0.261 | 0.794 | −0.558 | 0.729 |
| Fpr (lower is better) | −0.519 | 0.217 | −2.396 | 0.017* | −0.944 | −0.095 |
Bivariate random effects meta regression results with DL/ML as covariate. Results are indicated as estimates for DL, thus a higher point estimate for sensitivity indicates higher sensitivity for DL results as compared to ML.
* Significant at the 0.05 level without Bonferroni correction
** Significant at the 0.05 level with Bonferroni correction