Analysis · Health
Osteoarthritis pain: why beating placebo is only half the test
By AWEI · AI-compiled · Published · Analysis prepared · 1 source · doi.org
A pain medicine can outperform placebo without delivering a large enough average benefit to meet a clinically meaningful threshold. That distinction mattered in a 2021 review of knee and hip osteoarthritis treatments: some opioid regimens reduced pain statistically, but evidence for meaningful relief was weak, while treatment withdrawals due to adverse events were often increased.
Bruno R. da Costa and colleagues, publishing in The BMJ, combined 192 randomised trials involving 102,829 participants. Their network meta-analysis assessed 90 drug preparations or doses, keeping dose differences visible instead of grouping every version of a medicine together. It combined direct comparisons within trials with indirect comparisons across the network.
The main pain analysis focused on six weeks. Researchers used a threshold equivalent to a nine-point difference between treatment groups on a 100-point pain scale. Every opioid regimen assessed for pain had a probability of 53% or less of exceeding that threshold versus oral placebo. These probabilities describe uncertainty about average treatment effects, not the percentage of patients who would benefit.
Harms supplied another part of the comparison. Fifteen of 18 opioid regimens assessed for withdrawals showed evidence of more participants stopping because of adverse events than with oral placebo. Topical diclofenac showed more promising pain results without a clear increase in withdrawals, but its trials covered knee osteoarthritis only. Those findings cannot establish effectiveness for the hip.
The evidence also had important gaps. Median follow-up was 8.6 weeks, people with conditions increasing their risk of harm were underrepresented, and estimates for serious adverse events were often imprecise. Many trials had problems with missing outcome data. Short studies cannot establish long-term safety.
Randomisation supports estimates of treatment effects beyond observational associations, although indirect comparisons require assumptions about comparability across trials. The useful lesson is to read pain relief alongside tolerability, duration and the population studied. A favourable average result cannot identify the best treatment for an individual.
Sources used for this article (1)
Direct links to the publisher reports used to prepare this article.
- Source 1
- Effectiveness and safety of non-steroidal anti-inflammatory drugs and opioid treatment for knee and hip osteoarthritis: network meta-analysis ['Bruno R. da Costa', 'Tiago Pereira', 'Pakeezah Saadat', 'Martina Rudnicki', 'Samir Iskander', 'Nicolas S. Bodmer', 'Pavlos Bobos', 'Li Gao', 'Henry Dan Kiyomoto', 'Thaís Montezuma', 'Matheus Oliveir
