Effect estimates, event rates, and NNTs of meta-analyses examining medical cannabinoids versus placebo for spasticity
| SYSTEMATIC REVIEW | OUTCOME | NO. OF RCTS (NO. OF PARTICIPANTS) | AUTHORS’ META-ANALYSIS RESULT (95% CI), HETEROGENEITY | META-ANALYSIS RE-ANALYZED (95% CI), HETEROGENEITY | CANNABINOID EVENT RATE | CONTROL EVENT RATE | NNT |
|---|---|---|---|---|---|---|---|
| Whiting et al, 20152 | ≥ 30% improvement in spasticity | 2 (519) | OR = 1.64 (0.95 to 2.83), I2 = 44% | RR = 1.43 (0.99 to 2.08), I2 = 35% | 35% | 24% | NS (approximately 10)* |
| Mean reduction on Ashworth spasticity scale | 5 (1244) | WMD = 0.12 (−0.01 to 0.24), I2 = 0% | NA | NR | NR | NA | |
| Change in spasticity (VAS-NRS scale) | 3 (698) | WMD = 0.76 (0.14 to 1.38), I2 = 73% | NA | NR | NR | NA | |
| Global impression of change | 3 (461) | OR = 2.09 (1.02 to 4.27),† I2 = 69% | RR = 1.57 (0.97 to 2.55), I2 = 73% | 49% | 32% | NS (approximately 6)* | |
| Wade et al, 201028 | ≥ 30% improvement in spasticity | 3 (652) | OR = 1.57 (1.11 to 2.23), I2 = NR | RR = 1.37 (1.07 to 1.76), I2 = 0% | 35% | 25% | 10 |
| Change in spasticity (VAS-NRS scale) | 3 (652) | Mean change in VAS-NRS of 0.31 (0.04 to 0.59), I2 = NR | NA | Started at about 6.2, decreased by 1.27 | Started at about 6.2, decreased by 0.95 | NA | |
| Global impression of change | 3 (605) | OR = 1.66 (1.19 to 2.30), I2 = NR | RR = 1.32 (1.10 to 1.58), I2 = NS | 51% | 38% | 8 | |
| Meza et al, 201729 | Spasticity (change in any scale) | 4 (1247) | SMD = 0.07 (−0.04 to 0.19), I2 = NR | NA | NR | NR | NS |
| All studies | Global impression of change | 4 (746) | NA | RR = 1.45 (1.08 to 1.95), I2 = 60% | 50% | 35% | 7 |
NA—not applicable, NNT—number needed to treat, NR—not reported, NRS—numerical rating scale, NS—not significant, OR—odds ratio, RCT—randomized controlled trial, RR—risk ratio, SMD—standardized mean difference, VAS—visual analogue scale, WMD—weighted mean difference.
↵* Confidence intervals suggest that benefit is likely, so estimated NNT provided.
↵† Whiting et al2 report an OR of 1.44 (95% CI 1.07 to 1.94), I2 = 0% but when we re-ran this meta-analysis we found the numbers presented in the table. We contacted the authors but did not hear back.