Table 1.
Characteristics of included study
| Author | Singh et al34 |
| Country | USA |
| Indication | Total knee replacement >6 months. Chronic pain >3 months (≥6 points on 10-point VAS scale). Unsuccessful treatment with oral pain medication, not surgical candidate or infection identified. Mean pain duration 4.5 years |
| Number of patients | 49 patients with 60 total knee replacements (30 intervention: 30 control) |
| Age | Mean: intervention 67.1 years; control 66.8 years |
| Sex | Female: intervention 22%; control 12% |
| Approach | Standardised medial or lateral |
| Intervention | Intra-articular injection of 100 units botulinum toxin A diluted in 5 mL sterile normal saline |
| Control | Intra-articular injection of 5 mL sterile normal saline |
| Follow-up interval | Up to 6 months |
| Outcome measures | Proportion of responders at 2 months (≥2 point VAS reduction) Physicians’ global assessment of change Onset and duration of pain (20 point WOMAC pain decrease) WOMAC function Timed-stands test Timed-up-and-go Active knee flexion Medical Outcomes Study Short-Form 36 (SF-36) Short-form McGill Pain Questionnaire Changes in analgesic medications Side effects and adverse outcomes |
| Economic evaluation | None reported |
| Risk of bias | Overall: low |
| Random sequence | Independent |
| Concealment | Syringes prepared independently |
| Blinding | Patients, surgeon, investigators, statistician all blind to group allocation |
| Blind outcome | Assessment bind to group allocation |
| Complete data | Low losses to follow-up at primary outcome intervals |
| Selective reporting | Appropriate range of outcomes reported |
| Other bias | None apparent |
| Losses to follow-up | 2 (1:1) lost to 2 month follow-up. 7 (3:4) lost to 6 month follow-up |
| Power calculation | Reported to be powered for significant improvement on WOMAC scale |
| Results summary | Pain severity reduced in 71% of intervention patients compared with 35% in placebo group at 2 months. Benefit also at 3 months but not at 4 months. Duration of meaningful pain relief was 39.6 (SD=50.4) days in intervention group compared with 15.7 (SD=22.6) days in placebo group |
VAS, visual analogue scale; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index.