British Medical Journal 1990 (Jun 2); 300 (6737): 1431–1437 ~ FULL TEXT
NOTE: You may need to fill out the Free BMJ Registration to access this article
Meade TW, Dyer S, Browne W, Townsend J, Frank AO
MRC Epidemiology and Medical Care Unit, Northwick Park Hospital, Harrow, Middlesex
OBJECTIVE: To compare chiropractic and hospital outpatient treatment for managing low back pain of mechanical origin.
DESIGN: Randomised controlled trial. Allocation to chiropractic or hospital management by minimisation to establish groups for analysis of results according to initial referral clinic, length of current episode, history, and severity of back pain. Patients were followed up for up two years.
SETTING: Chiropractic and hospital outpatient clinics in 11 centres.
PATIENTS: 741 Patients aged 18-65 who had no contraindications to manipulation and who had not been treated within the past month.
INTERVENTIONS: Treatment at the discretion of the chiropractors, who used chiropractic manipulation in most patients, or of the hospital staff, who most commonly used Maitland mobilisation or manipulation, or both. MAIN OUTCOME MEASURES--Changes in the score on the Oswestry pain disability questionnaire and in the results of tests of straight leg raising and lumbar flexion.
RESULTS: Chiropractic treatment was more effective than hospital outpatient management, mainly for patients with chronic or severe back pain. A benefit of about 7% points on the Oswestry scale was seen at two years. The benefit of chiropractic treatment became more evident throughout the follow up period. Secondary outcome measures also showed that chiropractic was more beneficial.
CONCLUSIONS: For patients with low back pain in whom manipulation is not contraindicated chiropractic almost certainly confers worthwhile, long term benefit in comparison with hospital outpatient management. The benefit is seen mainly in those with chronic or severe pain. Introducing chiropractic into NHS practice should be considered.
From the FULL TEXT Article:
ECONOMIC IMPLICATIONS
The potential economic, resource, and policy
implications of our results are extensive. The average
cost of chiropractic investigation and treatment at
1988-9 prices was £165 per patient compared with £111
for hospital treatment. Some 300 000 patients are
referred to hospital for back pain each year," of whom
about 72000 would be expected to have no contraindications
to manipulation.'2 If all these patients were
referred for chiropractic instead of hospital treatment
the annual cost would be about £4m. Our results
suggest that there might be a reduction of some 290 000
days in sickness absence during two years, saving
about £13m in output and £2-9m in social security
payments. As it was not clear, however, that the
improvement in those treated by chiropractic was
related to the number of treatments the cost of essential
chiropractic treatment might be substantially less than
£4m. The possibility that patients treated in hospital
would need more treatment during the second year
than those treated by chiropractic (see above and table
VI) also has to be borne in mind. There is, therefore,
economic support for use of chiropractic in low back
pain, though the obvious clinical improvement in pain
and disability attributable to chiropractic treatment is
in itself an adequate reason for considering the use of
chiropractic.
Discussion
Though many randomised controlled trials of treatments
for back pain have been carried out, there have
so far been no clear indications in favour of any
particular method. The place of manipulation in back
pain has been reviewed by Jayson," who concluded
that any minor benefits seemed to be confined to those
with acute pain of recent onset, that there was no
evidence that manipulation helped those with severe or
chronic back problems, and that it did not reduce long
term complications or prevent recurrences. For
chiropractic our findings suggest otherwise. The
difficulties of clinical trials in low back pain have been
discussed.'4 Our trial had the combined advantages of
considerably larger numbers and a longer follow up
period than other trials comparing orthodox treatments
or, less often, orthodox and alternative treatments.
We studied only patients who had no contraindications
to manipulation. Although this group represents
a substantial proportion of all patients with back pain,
the findings cannot be automatically applied to all
patients with back pain. With this proviso the results
leave little doubt that chiropractic is more effective
than conventional hospital outpatient treatment. The
confidence intervals for the differences in Oswestry
scores were wide, but the degree of improvement
recorded for many of the secondary outcome measures
(table VI) suggests that chiropractic has appreciable
benefit. The effects of chiropractic seem to be long
term, as there was no consistent evidence of a return to
pretreatment Oswestry scores during the two years of
follow up, whereas those treated in hospital may have
begun to deteriorate after six months or a year.
Chiropractic was particularly effective in those with
fairly intractable pain-that is, those with a history and
severe pain. Although we have discussed the results in
terms of differences at the various follow up intervals,
the full effects of treatment are better thought of as an
integrated benefit throughout the two year follow up
period, represented by the area between the curves for
the two treatments. The greater proportions of patients
treated by chiropractic who were satisfied and relieved
at six weeks, when 90% of patients had provided follow
up data, strengthens the likelihood that the differences
in Oswestry scores and other variables later on,
when fewer patients have provided data, were true
differences.
The results from the secondary outcome measures
(table VI) suggest that the advantage of chiropractic
starts soon after treatment begins. The reason for the
much larger advantage later on is not obvious. Part of
the explanation could be that hospital treatment is
effective in the short term but not the longer term,
perhaps because it is not given for as long as chiropractic.
The undoubted difficulties under which some
of the participating physiotherapy departments were
working during the trial almost certainly meant that
they were unable to give all the specific treatment they
would have wished to all patients.
A central question is the extent to which the results
could be due to biases and placebo effects. Patients
were deliberately sent follow up Oswestry questionnaires
at home to minimise any chance that their
answers might be affected by actual or perceived
influence by their therapist. Ideally, straight leg raising
and lumbar flexion should have been measured by an
assessor who was blind to the treatment allocation. The
nurse coordinators, however, did not have the initial
results available at the time of the follow up measurements
at six weeks. In addition nearly all the other
subsidiary measures suggested greater improvement
among those treated by chiropractic.
The consequences of biased outcome measures or of
a placebo effect associated with chiropractic would
almost certainly have been more evident when treatment
was still in progress or just afterwards. In fact,
the main difference between hospital and chiropractic
treatment was seen from six months or a year onwards,
well after treatment and contact with therapists had
ended.
The fact that chiropractic treatment tended to be
more effective in those initially presenting to the
chiropractors than in those presenting to hospital
raises the possibility that the self assessment by the
patients who presented to chiropractors may have been
influenced by their expectation that chiropractic would
be effective. The results in all patients who had been
followed up for two years, however, indicate a similar
effect of chiropractic in both referral groups (table V).
There were several differences between the two referral
groups that may have influenced response to treatment
(these will be reported in detail elsewhere). For
example, a significantly higher proportion of patients
initially attending the chiropractors had had previous
episodes of back pain. Those initially attending chiropractors
had also waited much less time for appointments
for the current episode and scored significantly
less on questionnaires for depressive and inappropriate
symptoms and for somatic awareness than the patients
initially attending hospital. In addition, the analyses
among the (non-clinic) subgroups prespecified in the
minimisation procedure were balanced for referral
clinic, there being similar proportions initially presenting
to chiropractors and to hospital in each of the
randomised treatment groups. Yet the tendency for
chiropractic to be more effective was not universalfor
example, the absence of clear benefit in those with
no previous history of back pain. Finally, the self
exclusion of many patients who initially presented to
the chiropractors probably resulted in only a few of
these patients who might automatically have assessed
themselves as better after chiropractic or worse after
hospital treatment being included. In summary, it is
unlikely that the benefits of chiropractic are the result
of biased outcome assessments or of a placebo effect.
Centres where chiropractic was more effective at six
weeks and six months and those where there was less
difference between the two treatments at that stage
contributed to the results to about the same extent at a
year and two years. The sustained effect of chiropractic
was therefore probably not due to a disproportionate
contribution from individual centres where there was
an obvious early benefit from chiropractic.
In the absence of any clear relation between the
number of treatment sessions and outcome, specific
components ofchiropractic responsible for its effectiveness
have to be considered. An obvious possibility is
the use of high velocity, low amplitude manipulation in
virtually all the patients treated by chiropractic.
Another is that chiropractic was given for a longer
period than hospital treatment. Whatever the explanation
for the difference between the two approaches,
however, this pragmatic comparison of two types of
treatment used in day to day practice shows that
patients treated by chiropractors were not only no
worse off than those treated in hospital but almost
certainly fared considerably better and that they
maintained their improvement for at least two years.
Ifour results are more widely applicable the practical
implications are far reaching. Consideration should be
given to recognising appropriately trained and experienced
chiropractors and to providing chiropractic
within the NHS, either in hospitals or by purchasing
chiropractic treatment in existing clinics. Further
trials to identify the specific component(s) responsible
for the effectiveness of chiropractic should be undertaken.
Whether the results of this trial can also be
applied to other heterodox regimens of manipulation is
an open question.
We thank the nurse coordinators, medical staff, physiotherapists,
and chiropractors in the 11 centres for their work,
and Mr Alan Breen of the British Chiropractic Association for
his help. The centres were in Harrow, Taunton, Plymouth,
Bournemouth and Poole, Oswestry, Chertsey, Liverpool,
Chelmsford, Birmingham, Exeter, and Leeds. Without the
assistance of many staff members in each the trial could
not have been completed. The study was supported by
the Medical Research Council, the National Back Pain
Association, the European Chiropractors Union, and the
King Edward's Hospital Fund for London.
ADDENDUM- In view of the long term benefit apparently
due to chiropractic we initiated a three year follow
up, sending multiple reminders to those initially not
responding. By mid April 1990-beyond the closing
date for the earlier results-data were available for 113
patients, representing a 79% response. At three years
the mean fall in Oswestry score for those treated by
chiropractic was 9-6% points more than for those
treated in hospital (p=0-01). The fall was greater
(13 8% p=00003) among those presenting with current
episodes of more than a month's duration than for
those presenting with episodes of less than a month
(5 3%, NS). Among those with a previous history of
back pain, the improvement in Oswestry score at three
years was 9-7% points greater in patients treated by
chiropractic than those treated in hospital (p=002).
A similar difference between the two forms of
treatment (9-4%) was found among those with no
previous history of back pain, but numbers in this
group were smaller and the difference was not
significant.
Return to ChiroZine ARTICLES
Return to the LOW BACK PAIN Section
Return to the COST-EFFECTIVENESS Section
Since 8-11-1998
|
