Certainty assessment | Patients (n) | Effect | Certainty | Importance | ||||||||
Studies (n) | Study design | Risk of bias | Inconsistency | Indirectness | Imprecision | Other considerations | To test the efficacy of sufentanil sublingual tablet | Placebo | Relative (95% CI) | Absolute (95% CI) | ||
Time-weighted summed pain intensity difference from baseline over the 12-hour study period | ||||||||||||
3 | Randomised trials | Not serious* | Not serious | Not serious | Not serious | Strong association, dose–response gradient | 314 | 226 | – | MD 12.33 lower (15.5 lower to 9.17 lower) | ⨁⨁⨁⨁ High | Critical |
Time-weighted summed pain intensity difference from baseline over the 48-hour study period | ||||||||||||
3 | Randomised trials | Not serious* | Not serious† | Not serious | Not serious | Strong association | 607 | 338 | – | MD 43.57 lower (58.65 lower to 28.48 lower) | ⨁⨁⨁⨁ High | Critical |
Time-weighted total pain relief over 12 hours | ||||||||||||
3 | Randomised trials | Not serious* | Not serious‡ | Not serious | Not serious | Dose–response gradient | 301 | 214 | – | MD 4.77 lower (6.28 lower to 3.27 lower) | ⨁⨁⨁⨁ High | Critical |
Pain intensity difference from baseline | ||||||||||||
4 | Randomised trials | Not serious* | Not serious | Not serious | Not serious | None | 714 | 392 | – | MD 0.73 lower (1.0 lower to 0.46 lower) | ⨁⨁⨁⨁ High | Important |
Patient Global Assessment at 24 hours | ||||||||||||
3 | Randomised trials | Not serious | Not serious | Not serious | Not serious | Strong association | 374/537 (69.6%) | 86/215 (40.0%) | OR 3.72 (2.65 to 5.21) | 313 more per 1000 (from 239 more to 376 more) | ⨁⨁⨁⨁ High | Critical |
Healthcare Professional Global Assessment at 24 hours | ||||||||||||
3 | Randomised trials | Not serious | Not serious | Not serious | Not serious | Strong association | 388/537 (72.3%) | 89/215 (41.4%) | OR 3.91 (2.79 to 5.48) | 320 more per 1000 (from 249 more to 381 more) | ⨁⨁⨁⨁ High | Important |
*All studies included in the analysis have low risk of bias except for one which has some concerns with concealment of randomisation and hence not downgraded for risk of bias.
†Melson et al’s22 study has active comparator as compared with other studies, hence heterogeneity is >40% and therefore not downgraded for heterogeneity.
‡Although I2 is >40%, the effect estimate in all subgroups showed effect estimate in favour of sufentanil, hence heterogeneity among studies is ignored and evidence not downgraded for heterogeneity.
GRADE, Grading of Recommendations, Assessment, Development and Evaluations; MD, mean difference.