Table 5.
Meta-analysis of diagnostic accuracy for patients with Parkinson’s disease vs multiple system atrophy summary statistics for the bivariate and hierarchal summary receiver operating characteristic (HSROC) models
| Model | Variable | Coefficient estimate ± SE (95% CI) |
|---|---|---|
| Summary statistic | Sensitivity | 0.845 ± 0.042 (0.743–0.912) |
| Specificity | 0.845 ± 0.027 (0.784–0.891) | |
| DOR | 29.71 ± 10.825 (14.54–60.68) | |
| posLR | 5.44 ± 0.959 (3.85–7.68) | |
| negLR | 0.183 ± 0.050 (0.107–0.312) | |
| 1/negLR | 5.46 ± 1.49 (3.20–9.32) | |
| Bivariate | Logit-transformed sensitivity | 1.70 ± 0.325 (1.06–2.33) |
| Logit-transformed sensitivity variance | 1.69 ± 0.207 (1.29–2.10) | |
| Logit-transformed specificity | 0.855 ± 0.448 (0.306–2.39) | |
| Logit-transformed specificity variance | 0.303 ± 0.171 (0.100–0.915) | |
| Correlation between sensitivity and specificity | -0.140 ± 0.394 (-0.729–0.569) | |
| AUC (partial AUC) | 0.903 (0.866) | |
| HSROC | Lambda (Λ) | 3.51 ± 0.385 (2.75–4.26) |
| Theta (Θ) | -0.443 ± 0.381 (-1.19–0.304) | |
| Beta (β) | -0.520 ± 0.382 (-1.36–0.230) | |
| Variance Λ | 0.875 ± 0.488 (0.293–2.61) | |
| Variance Θ | 0.290 ± 0.161 (0.098–0.859) |