Issue 28 / 4 August 2014

EVIDENCE-based medicine is the mantra of modern-day medical practice.

It underpins the treatment and advice given to patients and is the basis for mainstream medicine’s critique of non-evidence-based alternative therapies.

But what happens when the evidence changes?

A recent large Australian study of paracetamol use for acute lower back pain raises exactly that question.

Clinical guidelines around the world have long recommended paracetamol as the first-line analgesic for acute low-back pain, although there really wasn’t any evidence to support that advice.

The National Prescribing Service’s Back Pain Choices tool for management of low-back pain in primary care, for example, describes paracetamol as “a good first option for pain relief”.

“When taken regularly, paracetamol is as effective as other medications but with fewer side effects”, the tool says.

What’s missing from that advice is the fact that there was never much evidence that any medications were effective against low-back pain.

NHMRC guidelines that were rescinded last year acknowledged as much, calling for placebo-controlled trials of paracetamol, non-steroidal anti-inflammatory drugs and opioids.

We now have that evidence for paracetamol at least, with the large, well designed Australian study showing the drug offered no benefit above placebo in reducing recovery time. It also had no effect on pain, disability, function, global symptom change, sleep or quality of life.

So how did paracetamol become the recommended first-line treatment?

An editorial published in The Lancet alongside the study says: “In the absence of sound clinical evidence, these recommendations favouring paracetamol are based on indirect evidence from other pain specialties, consensus in guideline committees, and the relatively favourable safety profile of paracetamol compared with, for example, non-steroidal anti-inflammatory drugs (NSAIDs).”

Recommendations in guidelines will always be based on differing levels of evidence. And that’s fine as long as they’re transparent about it.

A drug like paracetamol is not a prime target for research: its excellent safety profile means it is seen as a low-risk treatment option and it offers little prospect of financial return for industry.

Which makes this potentially game-changing Australian research all the more welcome.

Interestingly, participants in all arms of the study did better than would normally be expected for patients with acute low-back pain. The median time to recovery was 16 days in the placebo group and 17 in the treatment groups, with nearly 85% of all participants recovering by 12 weeks.

The authors suggest a possible explanation for such good outcomes is that all patients were provided with good-quality advice and reassurance, “a feature that is often absent from usual care”.

It would be interesting to know, as the editorial asks, how patients would have fared if there had been an additional arm to the trial providing advice and reassurance only — without the placebo-triggering addition of either a real or sham pill.


Jane McCredie is a Sydney-based science and medicine writer.

6 thoughts on “Jane McCredie: Changing evidence

  1. Genevieve Freer says:

    Interesting article.

    Having read the abstract of the Lancet article, one problem I perceive is that the study is about “acute lower back pain”, which is not a definitive, rather is a differential diagnosis, including muscle strain, ligament strain, acute lumbosacral disc disease, lumbar or sacral fracture, or first presentation of a bony metastasis, etc. 

    While it is tempting to diagnose all acute LBP patients as having a soft-tissue injury, or nothing wrong at all, careful history regarding mechanism of injury, risk factors for osteoporosis, previous fractures,  previous malignancy, physical  examination , and diagnostic imaging where indicated by history and examination, may prove otherwise, so any study of treatment effectiveness needs to be based on a definitive diagnosis.


  2. Richard Shiell says:

     Presumably the article about Paracetamol applies only to acute back pain. For chronic arthritic pain I find Panadol-osteo quite wonderful.  Sure it doesn’t repair my 76 year old  Pagetic and arthritic spine, but taken 6 hourly, night and day, it has helped me throught the past eight years without resource to other analgesics and without side-effects. I am pretty sure that it is not placebo relief as when I forget to take the solitary tablet, my back certainly reminds me within an hour or two!!

  3. Sue Ieraci says:

    Two points of interest. First, paracetamol as a choice, when its anti-inflammatory effects are known to be inferior to the NSAIDs. Second – the time-to-full-recovery endpoint. WHy not assessment of the actual pain relief when a dose was taken? As the commenter above mentions, a single study should only invite critical analysis, not be taken as gospel.

  4. CG Merridew says:

    The study is valuable.

    It would be even more valuable with a no-treatment (no pills) arm, attempting to quantify the placebo response.


  5. Graham Row says:

    Jane, I am not one of the privileged few who are able to access the Lancet article without the privilege of paying for it and must make judgements based on the published summary.  My understanding is that Paracetamol made no difference to the outcome defined as time to recovery. Any Information on better control of pain or otherwise must be buried in the article itself.  I am not sure that Paracetamol was ever held out to be “curative”. The study does seem to confirm that Paracetamol is both safe and well tolerated.  It also seems to confirm the fairly well established fact that reassurance and encouragement of activity is both beneficial and necessary.

  6. David Champion says:

    There appears to be widespread assumption that the results of this clinical trial are valid. This study should be mainly regarded as a stimulus to conduct at least one optimally designed randomised control trial in acute low back pain and another in chronic low back pain.

Leave a Reply

Your email address will not be published. Required fields are marked *