Osteoarthritis of the hip or knee is a progressive disabling disease affecting many people worldwide. Although paracetamol is widely used as a treatment option for this condition, recent studies have called into question how effective this pain relief medication is.
This review includes all trials published up to 3 October 2017.
We included randomised clinical trials (where people are randomly put into one of two or treatment groups) looking at the effects of paracetamol for people with hip or knee pain due to osteoarthritis against a placebo (a 'sugar tablet' that contains nothing that could act as a medicine). We found 10 trials with 3541 participants. On average, participants in the study were aged between 55 and 70 years, and most presented with knee osteoarthritis. The treatment dose ranged from 1.95 g/day to 4 g/day of paracetamol and participants were followed up between one and 12 weeks in all but one study, which followed people up for 24 weeks. Six trials were funded by companies that produced paracetamol.
Compared with placebo tablets, paracetamol resulted in little benefit at 12 weeks.
Pain (lower scores mean less pain)
Improved by 3% (1% better to 5% better), or 3.2 points (1 better to 5.4 better) on a 0- to 100-point scale.
• People who took paracetamol reported that their pain improved by 26 points.
• People who took placebo reported that their pain improved by 23 points.
Physical function (lower scores mean better function)
Improved by 3% (1% better to 5% better), or 2.9 points (1.0 better to 4.9 better) on a 0- to 100-point scale.
• People who took paracetamol reported that their function improved by 15 points.
• People who had placebo reported that their function improved by 12 points.
Side effects (up to 12 to 24 weeks)
No more people had side effects with paracetamol (3% less to 3% more), or 0 more people out of 100.
• 33 out of 100 people reported a side effect with paracetamol.
• 33 out of 100 people reported a side effect with placebo.
Serious side effects (up to 12 to 24 weeks)
1% more people had serious side effects with paracetamol (0% less to 1% more), or one more person out of 100.
• Two out of 100 people reported a serious side effect with paracetamol.
• One out of 100 people reported a serious side effect with placebo.
Withdrawals due to adverse events (up to 12 to 24 weeks)
1% more people withdrew from treatment with paracetamol (1% less to 3% more), or one more person out of 100.
• Eight out of 100 people withdrew from paracetamol treatment.
• Seven out of 100 people withdrew from placebo treatment.
Abnormal liver function tests (up to 12 to 24 weeks):
5% more people had abnormal liver function tests (meaning there was some inflammation or damage to the liver) with paracetamol (1% more to 10% more), or five more people out of 100.
• Seven out of 100 people had an abnormal liver function test with paracetamol.
• Two out of 100 people had an abnormal liver function test with placebo.
Quality of the evidence
High-quality evidence indicated that paracetamol provided only minimal improvements in pain and function for people with hip or knee osteoarthritis, with no increased risk of adverse events overall. None of the studies measured quality of life. Due to the small number of events, we were less certain if paracetamol use increased the risk of serious side effects, increased withdrawals due to side effects, and changed the rate of abnormal liver function tests. However, although there may be more abnormal liver function tests with paracetamol, the clinical implications are unknown.
Based on high-quality evidence this review confirms that paracetamol provides only minimal improvements in pain and function for people with hip or knee osteoarthritis, with no increased risk of adverse events overall. Subgroup analysis indicates that the effects on pain and function do not differ according to the dose of paracetamol. Due to the small number of events, we are less certain if paracetamol use increases the risk of serious adverse events, withdrawals due to adverse events, and rate of abnormal liver function tests.
Current clinical guidelines consistently recommend paracetamol as the first-line analgesic medication for hip or knee osteoarthritis, given its low absolute frequency of substantive harm. However, our results call for reconsideration of these recommendations.
Paracetamol (acetaminophen) is vastly recommended as the first-line analgesic for osteoarthritis of the hip or knee. However, there has been controversy about this recommendation given recent studies have revealed small effects of paracetamol when compared with placebo. Nonetheless, past studies have not systematically reviewed and appraised the literature to investigate the effects of this drug on specific osteoarthritis sites, that is, hip or knee, or on the dose used.
To assess the benefits and harms of paracetamol compared with placebo in the treatment of osteoarthritis of the hip or knee.
We searched the Cochrane Central Register of Controlled Trials, MEDLINE, Embase, AMED, CINAHL, Web of Science, LILACS, and International Pharmaceutical Abstracts to 3 October 2017, and ClinicalTrials.gov and the World Health Organization International Clinical Trials Registry Platform (ICTRP) portal on 20 October 2017.
We included randomised controlled trials comparing paracetamol with placebo in adults with osteoarthritis of the hip or knee. Major outcomes were pain, function, quality of life, adverse events and withdrawals due to adverse events, serious adverse events, and abnormal liver function tests.
Two review authors used standard Cochrane methods to collect data, and assess risk of bias and quality of the evidence. For pooling purposes, we converted pain and physical function (Western Ontario and McMaster Universities Osteoarthritis Index function) scores to a common 0 (no pain or disability) to 100 (worst possible pain or disability) scale.
We identified 10 randomised placebo-controlled trials involving 3541 participants with hip or knee osteoarthritis. The paracetamol dose varied from 1.95 g/day to 4 g/day, and the majority of trials followed participants for three months only. Most trials did not clearly report randomisation and concealment methods and were at unclear risk of selection bias. Trials were at low risk of performance, detection, and reporting bias.
At 3 weeks' to 3 months' follow-up, there was high-quality evidence that paracetamol provided no clinically important improvements in pain and physical function. Mean reduction in pain was 23 points (0 to 100 scale, lower scores indicated less pain) with placebo and 3.23 points better (5.43 better to 1.02 better) with paracetamol, an absolute reduction of 3% (1% better to 5% better, minimal clinical important difference 9%) and relative reduction of 5% (2% better to 8% better) (seven trials, 2355 participants). Physical function improved by 12 points on a 0 to 100 scale (lower scores indicated better function) with placebo and was 2.9 points better (0.95 better to 4.89 better) with paracetamol, an absolute improvement of 3% (1% better to 5% better, minimal clinical important difference 10%) and relative improvement of 5% (2% better to 9% better) (7 trials, 2354 participants).
High-quality evidence from eight trials indicated that the incidence of adverse events was similar between groups: 515/1586 (325 per 1000) in the placebo group versus 537/1666 (328 per 1000, range 299 to 360) in the paracetamol group (risk ratio (RR) 1.01, 95% confidence interval (CI) 0.92 to 1.11). There was less certainty (moderate-quality evidence) around the risk of serious adverse events, withdrawals due to adverse events, and the rate of abnormal liver function tests, due to wide CIs or small event rates, indicating imprecision. Seventeen of 1480 (11 per 1000) people treated with placebo and 28/1729 (16 per 1000, range 8 to 29) people treated with paracetamol experienced serious adverse events (RR 1.36, 95% CI 0.73 to 2.53; 6 trials). The incidence of withdrawals due to adverse events was 65/1000 participants in with placebo and 77/1000 (range 59 to 100) participants with paracetamol (RR 1.19, 95% CI 0.91 to 1.55; 7 trials). Abnormal liver function occurred in 18/1000 participants treated with placebo and 70/1000 participants treated with paracetamol (RR 3.79, 95% CI 1.94 to 7.39), but the clinical importance of this effect was uncertain. None of the trials reported quality of life.
Subgroup analyses indicated that the effects of paracetamol on pain and function did not differ according to the dose of paracetamol (3.0 g/day or less versus 3.9 g/day or greater).