Table 2.
Grade system for grading the quality of evidence.
| Quality assessment | No. of patients | Effect | Quality | Importance | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No. of studies | Design | Risk of bias | Inconsistency | Indirectness | Imprecision | Other considerations | Meta-analysis | Control | Relative (95% CI) | Absolute | ||
| ORR | ||||||||||||
| 17 | Randomised trials | Serious1 | No serious inconsistency | No serious indirectness | No serious imprecision | None | 416/735 (56.6%) | 270/710 (38%) | RR 1.45 (1.3 to 1.61) | 171 more per 1000 (from 114 more to 232 more) | ⊕⊕⊕Omoderate | Critical |
| 36.7% | 165 more per 1000 (from 110 more to 224 more) | |||||||||||
|
| ||||||||||||
| One-year survive rate | ||||||||||||
| 8 | Randomised trials | Serious1 | No serious inconsistency | No serious indirectness | No serious imprecision | None | 254/338 (75.1%) | 178/321 (55.5%) | RR 1.35 (1.21 to 1.51) | 194 more per 1000 (from 116 more to 283 more) | ⊕⊕⊕Omoderate | Critical |
| 52.5% | 184 more per 1000 (from 110 more to 268 more) | |||||||||||
|
| ||||||||||||
| Two-year survive rate | ||||||||||||
| 6 | Randomised trials | Serious1 | No serious inconsistency | No serious indirectness | No serious imprecision | None | 131/284 (46.1%) | 70/270 (25.9%) | RR 1.78 (1.42 to 2.22) | 202 more per 1000 (from 109 more to 316 more) | ⊕⊕⊕Omoderate | Critical |
| 35% | 273 more per 1000 (from 147 more to 427 more) | |||||||||||
|
| ||||||||||||
| Leukocyte toxicity | ||||||||||||
| 9 | Randomised trials | Serious1 | No serious inconsistency | No serious indirectness | Serious2 | None | 167/376 (44.4%) | 231/363 (63.6%) | RR 0.7 (0.62 to 0.79) | 191 fewer per 1000 (from 134 fewer to 242 fewer) | ⊕⊕OO low | Important |
| 66.7% | 200 fewer per 1000 (from 140 fewer to 253 fewer) | |||||||||||
|
| ||||||||||||
| Platelet toxicity | ||||||||||||
| 8 | Randomised trials | Serious1 | No serious inconsistency | No serious indirectness | Serious2 | None | 115/343 (33.5%) | 166/333 (49.8%) | RR 0.67 (0.57 to 0.79) | 165 fewer per 1000 (from 105 fewer to 214 fewer) | ⊕⊕OO low | Important |
| 50.1% | 165 fewer per 1000 (from 105 fewer to 215 fewer) | |||||||||||
|
| ||||||||||||
| Vomiting toxicity | ||||||||||||
| 14 | Randomised trials | No serious risk of bias | No serious inconsistency | No serious indirectness | No serious imprecision | Reporting bias3 | 260/588 (44.2%) | 321/574 (55.9%) | RR 0.79 (0.7 to 0.88) | 117 fewer per 1000 (from 67 fewer to 168 fewer) | ⊕⊕⊕O moderate | Not important |
| 61.4% | 129 fewer per 1000 (from 74 fewer to 184 fewer) | |||||||||||
|
| ||||||||||||
| CD3 + T cells (better indicated by lower values) | ||||||||||||
| 7 | Randomised trials | No serious risk of bias1 | No serious inconsistency | No serious indirectness | Serious2 | Reporting bias3 | 284 | 274 | — | SMD 1.42 higher (1.06 to 1.77 higher) | ⊕⊕OO low | Important |
|
| ||||||||||||
| CD4 + T cells (better indicated by lower values) | ||||||||||||
| 7 | Randomised trials | No serious risk of bias | No serious inconsistency | No serious indirectness | Serious2 | Reporting bias3 | 284 | 274 | — | SMD 1.52 higher (1.22 to 1.83 higher) | ⊕⊕OO low | Important |
|
| ||||||||||||
| CD4 + /CD8 + T-cell ratio (better indicated by lower values) | ||||||||||||
| 7 | Randomised trials | No serious risk of bias | No serious inconsistency | No serious indirectness | Serious2 | Reporting bias3 | 284 | 274 | — | SMD 1.34 higher (1.16 to 1.53 higher) | ⊕⊕OO low | Important |
1Risk of bias due to blind and random allocation concealment. 2Significant heterogeneity across studies. 3Small sample size.