Wednesday May 8 2013
Treating back pain costs the NHS more than a billion pounds a year
News that back pain could be "cured" by antibiotics has prompted wide media coverage, with The Independent reporting that "Half a million sufferers of back pain 'could be cured with antibiotics'."
The headlines are based on research into chronic lower back pain that showed some cases may be caused by a bacterial infection. Researchers found evidence that antibiotic treatment of a specific type of chronic lower back pain was more effective than placebo pills at reducing back pain and disability one year after treatment began.
Although antibiotics may also be effective for other types of back pain, this was not established by this study. While the results appear genuinely encouraging, this study involved patients with a very specific type of lower back pain. This means the results may be different in other people with different types of back pain.
There is also the potential risk this research could lead to the indiscriminate use of antibiotics in the hope of curing all back pain. This has negative consequences for both the individual and the community, as bacteria become resistant to antibiotics over time.
The study authors themselves say they do not support the idea that all patients with lower back pain should have a trial course of antibiotics, and state that excessive use of antibiotics should be avoided.
The findings of this research are positive but, as the study authors acknowledge, they will need to be confirmed in bigger studies in more diverse populations.
Where did the story come from?
The study was carried out by researchers from university hospitals in Denmark and was funded by the Danish Rheumatism Association and other foundations.
It was published in the peer-reviewed European Spine Journal.
A related study in the same journal and by the same researchers, also discussed in the media, provided good quality evidence that some types of lower back pain are associated with bacterial infection.
The media coverage was overexcited and may have generally overstated the significance of this research, implying that most people with lower back pain may benefit from antibiotics. The research was in a subgroup of lower back pain sufferers, so any treatment developed further down the line would not be suitable for all.
While the researchers had no conflicts of interest, much of the UK media's reporting of the study included quotes from neurosurgeon Peter Hamlyn, who claimed the research was "the stuff of Nobel Prizes…[and] is going to require us to rewrite the textbooks".
Hamlyn was not involved in the research, but he was reported to have funded a website that promotes the type of treatment used in the study, Modic Antibiotic Spinal Treatment. Only The Independent highlighted this potential conflict of interest.
What kind of research was this?
The research was a double-blind randomised control trial (RCT) looking at how well an antibiotic worked at treating long-term back pain in a subgroup of lower back pain sufferers.
The specific subgroup of lower back pain sufferers being studied had signs of bone swelling in their lower back that can only be detected through an MRI scan. The exact medical term for these changes is Modic type 1 changes, or bone oedema.
The theory was that, in some cases, this bone swelling might be caused by a bacterial infection. This meant that treatment with antibiotics represented a new avenue for the researchers to explore in an effort to treat this type of lower back pain.
A randomised control trial is an appropriate study design to test this hypothesis.
What did the research involve?
The researchers recruited 162 adult patients who had lower back pain for more than six months following a spinal disc herniation, commonly known as a slipped disc. To take part in the study they also had to have disease-related changes in the vertebrae next to the previous slipped disc site, so-called Modic type 1 changes, or bone oedema. These were confirmed through multiple MRI scans.
This select group of patients was randomised to receive either 100 days of antibiotic treatment with amoxicillin clavulanate tablets three times a day, or 100 days of an identical placebo.
Their health was evaluated at the start of the study, with participants unaware of which group they would be randomised to. Further evaluation of their health was also done without them knowing whether they had received placebo or antibiotics. Evaluation took place at the end of the 100-day treatment and again one year from the start of the study.
The researchers focused on changes in disease-specific disability and back pain. Disease-specific disability was measured using the Roland Morris Disability questionnaire (RMDQ). This is a 23-item questionnaire where the patient answers 23 yes or no questions related to the impact back pain has on their daily activities and quality of life. The questionnaire results in a scale score from zero to 23, with higher scores being worse.
Back pain was also measured using a patient-completed rating scale. Clinically relevant improvements in both measures were defined in advance of the results of the study – for instance, a 30% reduction for the RMDQ.
They also recorded changes in leg pain, number of hours with pain in the last four weeks, perceived health, days with sick leave, "bothersomeness", constant pain and disease-related changes observed under MRI.
The statistical analysis of the results was appropriate, and compared pain and disease changes in the group given antibiotics with those given placebo.
What were the basic results?
Of the 162 patients enrolled at the start, 147 (90.7%) completed the end-of-treatment questionnaires after 100 days, and 144 (88.9%) completed the one-year follow-up MRI scans, questionnaires and physical health checks.
Patients assigned to the placebo and antibiotic groups generally had similar characteristics at the start of the study.
The key results were:
- The group given antibiotics improved their disease-specific disability and back pain scores after treatment (100 days) and showed even larger improvements at the one-year time point.
- Back pain assessed by the RMDQ improved from 15 in the antibiotic group to 11.5 at 100 days and seven at one year, compared with a fall from 15 in the placebo group to 14 at 100 days remaining unchanged at 14 after one year.
- In comparison with the placebo group, the improvements observed using antibiotics were statistically significantly better.
- The magnitude of the improvement in disease-specific disability and back pain scores after treatment with antibiotics was also deemed clinically important using criteria defined before the start of the study.
- Patients reported that pain relief and improvement in disability started gradually – for most patients, six to eight weeks after starting antibiotic tablets, and for some at the end of the treatment period (100 days).
- Improvements reportedly continued long after the end of the treatment period – at least for another six months – and some patients reported continuing improvement at one-year follow-up.
- Less disease-related changes were detected in the vertebrae of the spine in patients given antibiotics than those given placebo. Changes were assessed from the start of the study to the one-year time point, with statistically significant differences seen.
- Overall, side effects were more common in the antibiotic group (65%) compared with the placebo group (23%). However, these side effects were described by the researchers as generally minor and were related to stomach upsets such as loose bowel movements, flatulence (farting) and burping.
How did the researchers interpret the results?
The researchers concluded that, "The antibiotic protocol [treatment] in this study was significantly more effective for this group of patients than placebo in all the primary and secondary outcomes."
They highlighted how "for the primary outcome measures, disease-specific disability and lumbar pain, the effect magnitude was also clinically significant."
This well-designed double-blind RCT shows that the antibiotic treatment of chronic lower back pain caused by swelling of the spinal vertebrae is more effective than placebo at reducing back pain and disease-related disability
The study had many strengths, including its randomised double-blind design, adequate sample size and one-year follow-up point.
However, it did have some limitations, including the fact that:
- Patients varied at the start of the study. More people in the placebo group had lower grades of vertebrae change. This is hard to explain if allocation to the two groups had been completely concealed and fair, although it may have favoured improvements in the placebo group and therefore may not have influenced the results.
- Blinding of the participants may have been broken unintentionally. As this antibiotic caused such predictable bowel side effects in 65% of people taking the active treatment, it is possible that the participants knew they were taking an active treatment and therefore may have reported the subjective scores differently form the placebo group. The researchers did not report any testing for fidelity of the blinding, such as asking participants if they could guess which group they were in.
As strong as this research it is, it is not definitive. Further research, most likely with larger numbers of people in the study, will be needed to confirm these findings before any treatment is likely to be approved and licensed for routine use in UK. There will also need to be extensive safety investigations.
Crucially, the study recruited a very select group of lower back pain sufferers who showed small changes in their vertebrae next to the site of a previous slipped disc. This select group therefore is not representative of all lower back pain sufferers.
This research certainly does not advocate giving antibiotics to all lower back pain sufferers. However, if the results are confirmed in subsequent studies and this form of treatment is deemed safe, it may provide a new treatment option for this type of lower back pain in the future. This is cause for much optimism.
The researchers' estimate that approximately 35-40% of long-term back pain sufferers experience excess fluid in the spinal vertebrae and could potentially benefit from this type of treatment in the future. However, it is unclear how accurate this estimate is, and may indeed be an overestimate.
Even if all these hurdles are overcome, media talk of a "back pain cure" could still be premature. Antibiotics may help relieve symptoms, but there is currently no conclusive evidence that they can correct the underlying causes of chronic back pain.
Analysis by Bazian. Edited by NHS Choices. Follow Behind the Headlines on Twitter.