Paracetamol 'not effective' for lower back pain or arthritis

"Paracetamol doesn't help lower-back pain or arthritis, study shows," The Guardian reports on a new review.

The review found no evidence that paracetamol had a significant positive effect, compared to placebo (dummy treatment) in relieving pain and disability in cases of acute lower back pain and was only minimally effective in osteoarthritis.

Before you start clearing out your medicine cabinet, the results of this review are not as clear-cut as reported.

The findings for lower back pain are based on three randomised controlled trials (RCTs), which, when grouped together, found no difference for pain relief, disability or quality of life between paracetamol and placebo. However, there are limitations in each of these studies. Two of the studies were small and the third only looked at acute lower back pain up to six weeks, when paracetamol may not be strong enough.

They did actually find that paracetamol slightly improved pain and disability from osteoarthritis of the hip or knee compared to placebo.

The study does not prove that paracetamol is no better than placebo for other types of back pain, such as chronic back pain (pain that persists for more than six weeks).

The National Institute for Health and Care Excellence (NICE) recommends that people with persistent back pain and recurrent back pain should stay physically active to manage and improve the condition.

Paracetamol is recommended as a first choice of painkiller because it has few side effects. NICE recommends that if this is not effective, stronger or different types of painkillers should be offered.

This guidance is currently under review, and this will take into account any new research such as the results of this study.

Where did the story come from?

The study was carried out by researchers from the University of Sydney, St Vincent’s Hospital and University of New South Wales and Concord Hospital in Sydney. It was funded by the National Health and Medical Research Council.

The study was published in the peer-reviewed British Medical Journal (BMJ) on an open-access basis so is free to read online (PDF 673kb).

The UK media reported the story accurately but did not explain any of the limitations of the study.

What kind of research was this?

This was a systematic review of all RCTs assessing the effectiveness of paracetamol for back pain and osteoarthritis of the hip or knee compared to placebo. The researchers also performed a meta-analysis. This is a statistical technique that combines the results of the RCTs to give an overall measure of effectiveness.

Pooling the results of multiple studies can help to give a better estimate of effectiveness, which is sometimes not seen in the individual studies, for example if they are too small.

This type of research is good at summarising all the research on a question and calculating an overall treatment effect, but relies on the quality and availability of the RCTs.

Paracetamol is currently recommended as the first line for pain relief for back pain and osteoarthritis of the hip and knee in clinical guidelines. The researchers wanted to assess whether this recommendation is backed up by the evidence.

What did the research involve?

A systematic review and meta-analysis was performed to identify and pool all RCTs that have assessed paracetamol compared to placebo for back pain and osteoarthritis of the hip and knee.

The following medical databases were searched for RCTs published up until December 2014: Medline, Embase, AMED, CINAHL, Web of Science, LILACS, International Pharmaceutical Abstracts, and Cochrane Central Register of Controlled Trials. A search was also made for unpublished studies, and authors were contacted for further information where required.

Three reviewers selected all relevant RCTs that reported on any of the following outcomes:

  • pain intensity
  • disability status
  • quality of life

Trials were excluded where a specific serious cause of the back pain had been identified, such as a tumour or infection, if they looked at post-operative pain and studies of people with rheumatoid arthritis.

The quality of each RCT was assessed using the standardised approach called a "risk of bias" assessment. The strength of the body of evidence as a whole was summarised using the internationally recognised GRADE approach (The Grading of Recommendations Assessment, Development and Evaluation).

A meta-analysis was then performed to pool the results of trials in people with the different conditions using appropriate statistical methods. This included an analysis of whether the RCTs were similar enough to be combined. The researchers also performed "secondary exploratory analysis", which looks at the effect various different factors may have had in biasing the results.

What were the basic results?

The systematic review included 13 moderate- to high-quality RCTs and 12 of them in the meta-analysis:

  • three trials investigated short-term use of paracetamol for lower back pain (including 1,825 people)
  • 10 trials assessed paracetamol compared to placebo for osteoarthritis of the knee or hip (including 3,541 people)
  • no trials were found for neck pain

No significant difference was found between paracetamol and placebo in the short term control of lower back pain in terms of:

  • pain intensity
  • disability
  • quality of life

Paracetamol slightly improved pain and disability from osteoarthritis of the hip or knee compared to placebo.

People experienced a similarly small number of side effects when taking paracetamol or placebo. However, people taking paracetamol were four times more likely to have abnormal liver function tests than those taking placebo. The review did not describe how abnormal the tests were or how quickly the tests returned to normal after stopping paracetamol.

How did the researchers interpret the results?

The researchers concluded that "paracetamol is ineffective in the treatment of lower back pain and provides minimal short term benefit for people with osteoarthritis". They call for "reconsideration of recommendations to use paracetamol for patients with lower back pain and osteoarthritis of the hip or knee in clinical practice guidelines".


This systematic review and meta-analysis suggests paracetamol may not be effective for some people with lower back pain and of limited help to people with osteoarthritis of the hip and knee.

Strengths of the study include:

  • the systematic review only contained the "gold standard" type of trials – RCTs
  • existing published RCTs comparing paracetamol with a placebo were likely to have been identified, as a large number of databases were searched from the beginning of their records up to December 2014. There were also two independent reviewers, which reduces the risk of any slipping through the net
  • they also searched for unpublished studies, reducing the risk of publication bias in their results (trials are less likely to be published if their results do not show a clear benefit)
  • the quality of evidence was appropriately assessed

However, as noted above, this type of research is reliant on the availability of relevant RCTs.

So while the review itself was well-conducted, the actual body of new evidence found about lower back pain was small.

In this case, the results for back pain were limited to three studies in specific populations. Non-specific lower back pain (i.e. back pain without an obvious cause) is complex in nature and these small studies may not be representative of all people who experience lower back pain.

First study

The first study was small, of 36 adults on strong (opioid) painkillers for at least six months for chronic back pain. While on these painkillers they did not find any difference in pain between an injection into the vein of either paracetamol, placebo or the non-steroidal anti-inflammatory drugs (NSAID) diclofenac and parecoxib.

Second study

The second study assessed the effect of paracetamol in acute back pain in 113 people after two and four days of use, compared to 20 people on placebo. The small study size limits the strength of the results. It may be that paracetamol was not a strong enough painkiller at this point in the course of the back pain, but may have been during the recovery phase.

Third study

The main outcome for the third study was whether paracetamol speeded up the time to recovery from acute lower back pain compared to placebo. How effective paracetamol was at pain relief was a secondary outcome so may not be as reliably assessed.

Some people will find paracetamol helps relieve the pain with relatively few side effects compared to other types of pain killers. The NICE guideline recommends paracetamol as a first line pain relief drug for lower back pain that has lasted for at least six weeks, along with other measures such as staying active. They recommend that if this does not provide adequate pain relief, then an NSAID should be offered.

NICE is currently updating its guidance on lower back pain and will take the results of this review into account.

NICE’s guidance also recommends paracetamol as a first line pain relief drug for osteoarthritis, however it does note that an evidence review suggested paracetamol may not be as effective for these people as originally thought. They are going to be reviewing this guidance (a draft is expected in 2016), and may revise their recommendations at that point, but for now have kept their existing guidance.

If you are finding that any prescribed treatment doesn’t seem to be working then you shouldn’t suddenly stop taking it (unless advised to). You do have the option of contacting your GP or doctor in charge of your care to discuss alternative drug (as well as non-drug) options.

NHS Attribution