Multivariable meta-regressions
| Moderators | P value | Marginal R2 | Within-articles variance | Between-articles variance | ||
| τ2 | Adjusted I2 | τ2 | Adjusted I2 | |||
| Without moderators | 3.6% | 56.6% | 2.8% | 42.9% | ||
| TMZ concentration and treatment duration | <0.001 | 42.1% | 1.7% (1.5% to 1.9%) | 30.9% | 3.7% (2.7% to 5.2%) | 68.5% |
| TMZ concentration, treatment duration and mediums’ glucose level | <0.001 | 45.4% | 1.7% (1.5% to 1.9%) | 31.9% | 3.5% (2.6% to 5.0%) | 67.4% |
| TMZ concentration, treatment duration and articles reporting quality | <0.001 | 44.1% | 1.7% (1.5% to 1.9%) | 31.6% | 3.6% (2.6% to 5.0%) | 67.8% |
| TMZ concentration, treatment duration, mediums’ glucose level and articles reporting quality | <0.001 | 45.9% | 1.7% (1.5% to 1.9%) | 32.0% | 3.5% (2.5% to 5.0%) | 67.4% |
Multivariable random-effects three-level meta-regressions with the raw data the effects were calculated with as first level, the reported effects as second level and the articles the effects were reported in as third level. P value is given for the test of the moderator. Marginal R2 indicates the regression model fit;69 τ2: estimator of the variance of true effects; adjusted I2: proportion of within-articles and between-articles variance of true effects, respectively, of the total observed variance including sampling error with the indicated moderators. τ2 estimator: restricted-maximum likelihood. For all multivariable meta-regressions, 644 effects in 101 articles were included.
TMZ, temozolomide.