2. Separating indirect from direct evidence: length of intensive care stay.
Treatment comparison | Proportion of direct evidence | RoR | P value |
Dobutamine versus levosimendan | 0.63 | 1.1501 | 0.7322 |
Dobutamine versus milrinone | 0.49 | 0.8695 | 0.7322 |
Dobutamine versus placebo, only indirect evidence | 0 | NA | NA |
Levosimendan versus milrinone | 0.76 | 1.2063 | 0.5537 |
Levosimendan versus placebo | 0.72 | 0.8302 | 0.6611 |
Milrinone versus placebo | 0.39 | 1.2046 | 0.6611 |
Separation of indirect from direct evidence using a back‐calculation method, random‐effects model RoR: ratio of ratios (direct/indirect), with P value of test for disagreement
Abbreviations NA: not applicable; RoR: ratio of ratios