GRADE evidence profile for PICO 4
| Certainty assessment | № of patients | Effect | Certainty | Importance | ||||||||
| № of studies | Study design | Risk of bias | Inconsistency | Indirectness | Imprecision | Other considerations | Selection of reperfusion therapy (MT or IVT) candidates based on a specific clinical presentation (eg, NIHSS cut-off - or coma on admission) | patient selection irrespective of clinical presentation | Relative (95% CI) | Absolute (95% CI) | ||
| mRS 0–3 RCT and NIHSS 0–9 | ||||||||||||
| 2 | Randomized trials | Serious* | Not serious | Serious† | Serious‡ | None | 26/37 (70.3%) | 30/41 (73.2%) | RR 0.67 (0.23 to 1.99) | 241 fewer per 1000 (from 563 fewer to 724 more) | ⨁◯◯◯Very low§ | CRITICAL |
| mRS 0–3 RCT and NIHSS 10- | ||||||||||||
| 3 | Randomized trials | Serious* | Not serious | Serious† | Serious‡ | None | 197/453 (43.5%) | 77/326 (23.6%) | RR 2.44 (1.77 to 3.36) | 340 more per 1000 (from 182 more to 557 more) | ⨁◯◯◯Very low§ | CRITICAL |
| mRS 0–3, NRSI and NIHSS 0–9 | ||||||||||||
| 2 | Non-randomized studies | Serious¶ | Not serious | Serious† | Not serious | None | RR 0.96 (0.78 to 1.19) | One fewer per 1000 (from one fewer to one fewer) | ⨁◯◯◯Very low | CRITICAL | ||
| mRS 0–3, NRSI and NIHSS 10- | ||||||||||||
| 2 | Non-randomized studies | Serious¶ | Not serious | Serious† | Not serious | None | RR 2.19 (0.84 to 5.76) | Two fewer per 1000 (from 6 fewer to one fewer) | ⨁◯◯◯Very low | CRITICAL | ||
| sICH RCT and NIHSS 0–9 | ||||||||||||
| 2 | Randomized trials | Serious* | Not serious | Serious† | Serious‡ | None | 3/37 (8.1%) | 0/41 (0.0%) | RR 6.78 (0.37 to 125.95) | 0 fewer per 1000 (from 0 fewer to 0 fewer) | ⨁◯◯◯Very low | IMPORTANT |
P=0.01 for interaction among proximal, middle, and distal locations.
* Risk of bias was assessed as serious due to high risk of bias detected in all RCTs.
†No study specifically tested efficacy in high versus low NIHSS scores.
‡Serious imprecision due to low optimal information size. The total number of patients included is less than the number of patients generated by a conventional size sample calculation for a single adequately powered clinical trial.
§P<0.03 for interaction between NIHSS 0-9 vs. NIHSS ≥ 10.
¶Risk of bias was assessed as serious using ROBINS-I tool.
CI, confidence interval; IVT, give details; mRS, modified Rankin Scale; MT, give details; NIHSS, give details; NRSI, non-randomized studies of intervention; RCT, randomized controlled trials; RR, risk ratio; sICH, give details.