Table 8.
Inconsistency revealed in the subgroup NMA node-splitting tests.
| Outcomes of the NMAs | Groups | Inconsistency was observed between direct and indirect evidences in these comparisons |
|---|---|---|
| LDL-C | Group 1 | Placebo vs. fluvastatin; atorvastatin vs. rosuvastatin; atorvastatin vs. lovastatin; rosuvastatin vs. lovastatin |
| Group 2 | Placebo vs. rosuvastatin | |
| Group 3 | None (no indirect comparison involved) | |
|
| ||
| HDL-C | Group 1 | None (no indirect comparison involved) |
| Group 2 | Placebo vs. rosuvastatin | |
| Group 3 | None (no indirect comparison involved) | |
|
| ||
| TC | Group 1 | None (no indirect comparison involved) |
| Group 2 | None (no indirect comparison involved) | |
| Group 3 | None (no indirect comparison involved) | |
|
| ||
| TG | Group 1 | None (no indirect comparison involved) |
| Group 2 | None (no indirect comparison involved) | |
| Group 3 | None (no indirect comparison involved) | |
|
| ||
| ApoA1 | Group 1 | NA∗∗ |
| Group 2 | NA∗∗ | |
| Group 3 | NA∗∗ | |
|
| ||
| ApoB | Group 1 | NA∗∗ |
| Group 2 | NA∗∗ | |
| Group 3 | NA∗∗ | |
NA∗∗: too few articles included in the subgroup analysis, data contain no potential source of heterogeneity.