Appraisal of Two Randomized Clinical Trials

1
Appraisal of Two Randomized Clinical Trials
2
Low Back Pain RCT

• Meade T, et al. Randomised comparison of
  chiropractic and hospital outpatient management
  for low back pain results from extended follow
  up. BMJ 1995 311(7001) 349-351.
• One of the strongest studies to date supporting
  chiropractic care of LBP

3
T. W. Meade

• Director of the Medical Research Council
  Epidemiology and Medical Care Unit, Wolfson
  Institute of Preventive Medicine, Medical College
  of St Bartholomew's Hospital, London

4
Introduction

• The authors previously reported that LBP patients
  treated with chiropractic did better than those
  receiving hospital outpatient management
  (followed for 6 months)
• This paper presented results for the same group,
  but three years later
• Oswestry questionnaires and pain scales were the
  outcome measures

5
Methods

• Patients presenting either to a chiropractic
  clinic or hospital were randomly allocated to
  either chiropractic or in hospital treatment
• Chiropractors used chiropractic manipulation in
  most patients
• Hospital staff most commonly used Maitland
  mobilization or manipulation

6
Methods Cont.

• 741 patients started treatment
• Progress was measured with the Oswestry
  questionnaire
• At six weeks, 95 of chiropractic and 89 of
  hospital were returned
• At three years by 77 and 70

7
Methods Cont.

• At the three year follow up patients were asked
  whether they thought their allocated treatment
  had helped their back pain
• Results were analyzed on an intention to treat
  basis
• All patients entering the study were included in
  the statistical analysis, even if they dropped out

Intention to Treat -
Whatever the
reason for people failing to complete follow up
or not adhering to the protocol, everyone should
be analyzed according to the group they were
initially allocated to, in other words the group
in which they were intended to remain
8
Methods Cont.

• Differences between group means were tested by
  unpaired t tests
• X2 was used to test for differences in
  proportions between the two treatment groups

9
Results Cont.

• Mean (SD) Oswestry scores before treatment were
  29.8 (14.2) in chiropractic and 28.5 (14.1) in
  hospital treatment group
• 20-40 moderate disability
• There was a 3.18 percentage point difference at
  three years a 29 greater improvement in
  patients treated with chiropractic

10
Results Cont.
11
Results Cont.

• Pain intensity before treatment and at the
  various follow up intervals were all positive
  (improved), but were all significantly greater in
  those treated by chiropractic
• Those with short current episodes, a history of
  back pain, and initially high Oswestry scores
  derived the most benefit from chiropractic

12
Results Cont.
TABLE III--Number (percentage) of patients at
three year follow up who considered allocated
trial treatment had helped their back
pain ---------------------------------------------
--------------------------------------------------
-- Hospital
treatment Chiropractic
treatment ----------------------------------------
--------------------------------------------------
------- Referral Help
No help Help No
help ---------------------------------------------
--------------------------------------------------
-- Hospital 71 (60.2) 47
(39.8) 103 (79.2) 27
(20.8) Chiropractic 76 (65.5) 40
(34.5) 127 (84.7) 23
(15.3) -------------------------------------------
--------------------------------------------------
---- For hospital referrals X210.7
P0.001. For chiropractic referrals X213.3
P 13
Discussion

• The results at six weeks and six months were
  identical with those in their first report
• The substantial benefit of chiropractic on
  intensity of pain is evident early on and then
  persists
• Larger proportions were lost to follow up in
  those treated in hospital than in those treated
  by chiropractic suggesting greater satisfaction
  with chiropractic

14
Discussion

• The authors indicated that there is now more
  support for the need to conduct rigorous trials
  focusing on specific components of management
• Meanwhile, the results of our trial show that
  chiropractic has a valuable part to play in the
  management of low back pain

15
RCT Question List

• Was the hypothesis stated clearly?
• H1 Chiropractic ? hospital outpatient treatment
  for managing low back pain
• Did the trial address a clearly focused issue?
  Yes
• Was there an adequate literature review? No. They
  relied too heavily on their previous article

To compare the effectiveness
over three years of chiropractic and hospital
outpatient management for low back pain
16
RCT Question List Cont.

• Were an adequate number of subjects used? Yes
• What were the exclusion and inclusion criteria?
  Not given
• Previous article
• Was the assignment of subjects to each group
  concealed? ?
• Were the subjects assigned to groups randomly? Yes

17
RCT Question List Cont.

• Were all of the subjects accounted for? Yes
• Were the groups similar at the start of the
  study? Yes
• Were the groups treated equally except for the
  intervention? No
• One group was treated in offices and the other in
  hospitals which may have made a huge difference

18
RCT Question List Cont.

• Were the methods adequately described and
  reproducible?
• No, but they referred to their 1990 article
• Treatment at the discretion of the chiropractors,
  who used chiropractic manipulation in most
  patients, or of the hospital staff, who most
  commonly used Maitland mobilization or
  manipulation, or both
• Could the differences have been related to
  patients preference for office vs. hospital?

19
RCT Question List Cont.

• What outcomes were measured? Oswestry, pain
  levels, and satisfaction
• Were all outcomes measured reported? Yes
• Were the appropriate outcomes assessed? Yes
• Were statistics calculated correctly? Yes
• But ANCOVA may have been a better test
• Mean scores before treatment were 29.8
  (chiropractic) and 28.5 (hospital)

20
RCT Question List Cont.

• Was the difference between groups statistically
  and clinically significant? Yes
• How are the results applicable outside of the
  study?
• Generalizable to the typical chiropractic setting

21
RCT Question List Cont.

• What do I think are the strengths of this
  article?
• Randomization
• The setting was similar to that of the average
  chiropractor
• What do I think are the weaknesses of this
  article?
• PTs using manipulation and in hospital

22
RCT Question List Cont.

• How would I improve this study?
• It would have been better to compare with a
  placebo
• Compare office-based care for both groups
• Use ANCOVA to further equalize groups
• May be able to use less than 741 subjects and get
  just as powerful results

23
Neck Pain RCT

• Cassidy, J.D., A.A. Lopes, and K. Yong-Hing, The
  immediate effect of manipulation versus
  mobilization on pain and range of motion in the
  cervical spine a randomized controlled trial.
  JMPT, 1992. 15(9) 570-5.
• One of a handful of studies supporting
  chiropractic care of neck pain

24
Cassidy, J.D.

• Department of Orthopaedic Surgery, Royal
  University Hospital, Saskatoon, Saskatchewan,
  Canada
• Involved in the QTF report on whiplash
• QTF was generated by an insurance company -
  Societe d'Assurance Automobile du Quebec (SAAQ)
• Was a very biased report

25
Introduction

• Adequate literature review pointing out that neck
  pain is common
• Affects 40-50 of the general population at some
  point in their lives
• Most patients with mechanical neck pain improve
  with time, but as many as one-third continue to
  have moderate or severe pain 15 yrs after the
  initial onset

26
Introduction Cont.

• Pointed out that 5 RCTs had been done, but 3 of
  them used mobilization
• Purpose of the study
• To compare the immediate results of manipulation
  and mobilization on pain and range of motion in
  patients with unilateral mechanical neck pain

27
Methods

• One hundred consecutive outpatients suffering
  from mechanical neck pain with radiation into the
  trapezius region
• Population ? 100 consecutive patients
• Inclusion/exclusion criteria
• Generally good health
• Unilateral neck pain aggravated by movement
• Without neurological deficit

28
Methods Cont.

• Each patient completed a questionnaire on the
  history of their neck pain, a Pain Disability
  Index, and ROM was tested
• Patients rated their pain on the NRS-101, a valid
  and reliable measure of pain
• The 101-point Numerical Rating Scale is a
  progressive numerical scaling method ranging from
  0-100, with 0 representing no pain and 100
  representing extreme pain

29
Methods Cont.

• All treatments were given once and were applied
  to the symptomatic side
• Within 5 min after treatment, the patients
  completed a post-test NRS-101 and ROM was
  retested
• ROM was tested with a goniometer

30
Methods Cont.

• Cervical manipulation involved
• Contacting the articular pillar on the painful
  side of the neck at the level of tenderness
• Passively rotating the neck away from the painful
  side as far as possible and applying a
  high-velocity, low-amplitude thrust in the same
  direction

31
Results

• 52 subjects were manipulated and 48 were
  mobilized without complication
• However, 6 of all subjects had increased pain
  after treatment
• The mean (SD) age was 34.5 (13.0) yr for the
  manipulated group and 37.7 (12.5) yr for the
  mobilized group
• The only demographics that were given

32
Results Cont.

• Pain intensity decreased and range of motion
  range of motion increased after treatment in both
  groups
• NRS-101 scores decreased 17.3 points in the
  manipulated group and 10.5 points in the
  mobilized group
• The manipulated group showed greater ROM gains
  but the differences were not significant

33
ROM Gain (Change) Scores
34
Gain (Change) Scores Graph
Error Bar
35
Discussion

• The results suggest that both treatments have the
  immediate effect of decreasing pain and
  increasing cervical ROM
• Overall pain improvement on the NRS-101 was more
  than 1.5 times greater in the manipulated
  patients
• A long-term trial was suggested to determine
  which treatment would give the best results over
  time

36
Discussion Cont.

• The mechanism by which manipulation works is not
  certain
• There is no evidence that it reduces subluxation
  or that minor positional misalignments are of
  clinical significance
• It is more likely that manipulation affects the
  pain reflex and/or muscle tension

37
Discussion Cont.

• Cervical spine manipulation has been reported to
  be associated with rare but serious complications
  
• Stroke in less than 1 in 1,000,000 manipulations
• Probably safer than most medical treatments for
  neck pain

38
Discussion Cont.

• Mobilization is usually very gentle and unlikely
  to harm patients, even if it has little or no
  therapeutic benefit
• Both risk and benefit are important concerns in
  the evaluation of treatment options

39
Conclusion

• This study demonstrates that a single
  manipulation is more effective than mobilization
  in decreasing pain in patients with mechanical
  neck pain
• Both treatments increase neck ROM to a similar
  degree

40
Comparative Risk of CMT-Related CAD Developing
CMT-related ICAD in US/cervical CMT
1601,145,000 Death from CMT related ICAD in
US/cervical CMT 03,606,870,000 Death by falling
aircraft/year 110,000,000 Death by lightning
strike/year 12,000,000 Developing
CMT-related VAD in US/ cervical CMT
11,000,000 Death by being struck by an
automobile/year 120,000 Death related to
regular NSAID use/ users 14000 Death related
to cervical spine surgery/procedures 1145
41
RCT Question List

• Was the hypothesis stated clearly? Yes
• H1 chiropractic ? mobilization for neck pain
• Did the trial address a clearly focused issue?
  Yes
• Was there an adequate literature review? Yes
• Were an adequate number of subjects used? Yes

42
RCT Question List Cont.

• What were the exclusion and inclusion criteria?
• Generally good health
• Unilateral neck pain aggravated by movement
• Without neurological deficit
• Was the assignment of subjects to each group
  concealed? ?
• Were the subjects assigned to groups randomly?
  Yes
• Were all of the subjects accounted for? Yes

43
RCT Question List Cont.

• Were the groups similar at the start of the
  study? Yes
• Were the groups treated equally except for the
  intervention? Yes
• Were the methods adequately described and
  reproducible?
• Yes
• What outcomes were measured? Pain level and ROM

44
RCT Question List Cont.

• Were all outcomes measured reported? Yes
• Were the appropriate outcomes assessed? Yes
• Were statistics calculated correctly? P0.05?
• Was the difference between groups statistically
  and clinically significant? Yes for pain, No for
  ROM

45
RCT Question List Cont.

• How are the results applicable outside of the
  study?
• Not generalizable to the typical chiropractic
  setting because multiple visits is the norm
• What do I think are the strengths of this
  article?
• Randomization
• Its nice to know that patients do better with
  manipulation immediately after treatment.