Table 1.
Characteristics of the 25 large, placebo controlled trials*
| Author (year) | No of randomised participants | Instrument | Reported SAEs | Publication status | Funding independent of industry | Molecular weight | Structure | No of injections (cycles)† | |||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Visco | Placebo | Pain | Function | ||||||||
| Shichikawa (1983)34 | 114 | 114 | Global pain (VAS) | — | No | Published | No | Unclear | Non-cross linked | 5 (1) | |
| Puhl (1993)35 | 102 | 107 | Global pain (VAS) | Lequesne index | No | Published | No | Low | Non-cross linked | 5 (1) | |
| Lohmander (1996)36 | 120 | 120 | Global pain (VAS) | Lequesne index | No | Published | No | Low | Non-cross linked | 5 (1) | |
| Altman and Moskowitz (1998)37 | 164 | 168 | Pain on walking (VAS) | WOMAC function | Yes | Published | No | Low | Non-cross linked | 5 (1) | |
| Brandt (2001)38‡ | 114 | 112 | — | — | Yes | Published | No | Intermediate | Non-cross linked | 3 (1) | |
| Seikagaku [UK] (2001)39 | 116 | 115 | Lesquene index | Lequesne index | No | Unpublished | No | Low | Non-cross linked | 5 (1) | |
| Jubb (2003)40 | 208 | 200 | Pain on walking (VAS) | — | Yes | Published | No | Low | Non-cross linked | 3 (3) | |
| Altman (2004)41 | 173 | 174 | WOMAC pain | WOMAC function | Yes | Published | No | High | Cross linked | 1 (1) | |
| Day (2004)42 | 116 | 124 | WOMAC pain | WOMAC function | No | Published | No | Low | Non-cross linked | 5 (1) | |
| Pham (2004)43 | 131 | 85 | Global pain (VAS) | Lequesne index | No | Published | No | Intermediate | Unclear | 3 (3) | |
| Altman (2009)44 | 293 | 295 | Pain on walking (VAS) | WOMAC function | Yes | Published | No | Intermediate | Non-cross linked | 3 (1) | |
| Baltzer (2009)45 | 135 | 107 | Global pain (VAS) | WOMAC function | No | Published | Yes | Low | Non-cross linked | 3 (1) | |
| Chevalier (2010)46 | 124 | 129 | WOMAC pain | WOMAC function | Yes | Published | No | High | Cross linked | 1 (1) | |
| Jørgensen (2010)47 | 167 | 170 | Pain on walking (VAS) | Lequesne index | No | Published | No | Low | Non-cross linked | 5 (1) | |
| Huang (2011)48 | 100 | 100 | Pain on walking (VAS) | WOMAC function | Yes | Published | No | Low | Non-cross linked | 5 (1) | |
| Strand (2012)49 | 251 | 128 | WOMAC pain | WOMAC function | Yes | Published | No | Unclear | Cross linked | 1 (1) | |
| NCT00988091 (2012)50 | 298 | 298 | Pain on walking (VAS) | WOMAC function | Yes | Unpublished | No | Unclear | Unclear | 1 (1) | |
| Arden (2014)51 | 108 | 110 | WOMAC pain | WOMAC function | No | Published | No | High | Cross linked | 1 (1) | |
| NCT01372475 (2015)52 | 400 | 400 | WOMAC pain | — | No | Unpublished | No | Unclear | Non-cross linked | 2 (1) | |
| NCT01934218 (2017)53§ | 404 | 410 | Pain on walking (VAS) | — | Yes | Unpublished | No | Unclear | Cross linked | 1 (1) | |
| Hangody (2017)54 | 150 | 69 | WOMAC pain | WOMAC function | Yes | Published | No | Intermediate | Cross linked | 1 (1) | |
| Petterson and Plantcher (2018)55 | 184 | 185 | Patient global assessment (VAS) | WOMAC function | Yes | Published | No | Intermediate | Cross linked | 1 (1) | |
| NCT02495857 (2018)56¶ | 400 | 199 | WOMAC pain | WOMAC function | Yes | Unpublished | No | Intermediate | Non-cross linked | 3 (1) | |
| Ke (2021)57 | 220 | 220 | WOMAC pain | — | Yes | Published | No | High | Cross linked | 1 (1) | |
| Migliore (2021)58 | 347 | 345 | Global pain (VAS) | Lequesne index | Yes | Published | No | High and low | Unclear | 1 (1) | |
When available, molecular weight was categorised as low (<1500 kDa), intermediate (≥1500 and <6000 kDa) and high (≥6000 kDa).
SAE=serious adverse event; WOMAC=Western Ontario and McMaster Universities Arthritis Index; VAS=visual analogue scale.
In all 25 large, placebo controlled trials, 100% of the 9424 randomised participants had clinically or radiologically confirmed osteoarthritis of the knee. There were no quasi randomised trials among the 25 large trials.
Number of injections per cycle.
Pain and function only reported for a subgroup of patients, but serious adverse events reported for all randomised participants.
Only partially published as a subgroup analysis in Takamura et al.60
Trial NCT02495857 was a three arm trial (200 participants in viscosupplementation group 1, 200 participants in viscosupplementation group 2, and 199 in the placebo group). Pain and function estimates were not reported for viscosupplementation group 1, resulting in a total of 399 participants for efficacy estimates and 599 for serious adverse events (data from all three groups reported). Among the 25 large, placebo controlled trials, trial NCT02495857 was the only trial requiring combination of viscosupplementation groups. The remaining large, placebo controlled trials had one viscosupplementation group only.