Effects of Analgesics on Self-Reported Physical Function and Walking Ability in People With Hip or Knee Osteoarthritis: A Systematic Review and Meta-Analysis.

Certainty was low for acetaminophen and moderate for NSAIDs and opioids, so none of these results is settled. Ten of the 33 trials had a high risk of attrition bias, mainly from high dropout rates analysed per protocol, and allocation concealment was unclear in most trials, so the measured benefits may be inflated. Most trials ran around 12 weeks, so there is no evidence here about months or years of use. Physical function was self-reported in most trials, and the questionnaires include items about pain, so some of the improvement in function scores may reflect less pain rather than people actually doing more. The search covered only three databases, and nearly half of the included studies were found by hand-searching reference lists rather than by the search itself, so relevant trials may have been missed. Two outcomes named in the registered protocol, physical activity level and physiological response to exercise, could not be reviewed because too few studies had measured them.

The paper states that no funders contributed to the design, execution or interpretation of the results. No other funding or author declaration is given in the text provided.

Do not read these small improvements as evidence that analgesics should replace exercise for hip or knee osteoarthritis. The effects were small, the evidence was low to moderate quality, most trials lasted around 12 weeks so nothing is known about long-term use, and 10 of the 33 trials had high dropout analysed per protocol. The authors themselves put exercise first, with analgesics as a possible short-term support.