Table 2.
The meta-regression analysis in the binary classification of variable data using the odds ratio (OR)
| LogOR | Coef. | Std. Err. | t | P>|t| | [95% Conf. Interval] | |
|---|---|---|---|---|---|---|
| Regulation mode | -1.302928 | 0.492039 | -2.65 | 0.016 | -2.336663 | -0.2691922 |
| miRNA profiling | 1.869125 | 0.7728873 | 2.42 | 0.026 | 0.2453491 | 3.492901 |
| Sample size | -0.9457466 | 0.841248 | -1.12 | 0.276 | -2.713143 | 0.8216499 |
| Internal reference types | 0.7083823 | 0.525466 | 1.35 | 0.194 | -0.3955807 | 1.812345 |
| Specimen types | 0.5501442 | 0.8642539 | 0.64 | 0.532 | -1.265586 | 2.365874 |
| Ethnicity | 1.472494 | 1.404431 | 1.05 | 0.308 | -1.478106 | 4.423094 |
| Control-CHB | 1.596359 | 0.7393835 | 2.16 | 0.043 | 0.0587269 | 3.133991 |
| Control-LC | 1.99139 | 0.7978941 | 2.50 | 0.021 | 0.332078 | 3.650701 |
| Control-CHB+LC | 0.9694282 | 0.8318447 | 1.17 | 0.257 | -0.7604876 | 2.699344 |
LogOR was used as response variables as well as regulation modes, miRNA profiling, sample size, internal reference types, specimen types, ethnicity, and source of controls group were as covariates. Estimate of between-study variance tau2 = 0.6265. Residual variation due to heterogeneity: I-squared_res = 66.64%. Proportion of between-study variance explained: Adj R-squared = 47.42%. Joint test for all covariates with Knapp-Hartung modifcation: Prob > F = 0.0712.