acupuncture and the placebo question: evidence on msk pain ... › assets ›...

Acupuncture and the placebo question:

evidence on MSK pain, osteoarthritis and

headache

Hugh MacPherson

University of York, UK

2

“... any discipline whose practitioners make

specific claims ... to treat specific conditions

should have evidence ... above and beyond the

placebo effect“ (HoL Report 2000*)

*House of Lords, Science and Technology, Sixth Report, 2000 www/parliament.the-stationarey-office/co.uk/pa

The need to answer the

placebo question

Colquhoun D, Novella SP. Acupuncture Is Theatrical Placebo.

Anesthesia & Analgesia. 2013 Jun;116(6):1360–3.

Evidence hierarchy

Individual Patient Data

Meta-analysis

• A meta-analysis in which the analysis is by

patient not just by trial

– Increases statistical power

– Allows standardisation of analyses across trials

– Facilitates sub-group analyses

– Has been used for cancer and diabetes

– Requires effective research collaboration

What is an Individual Patient Data (IPD)

Meta-analysis?

Members of the

Acupuncture Trialists Collaboration

Claire Allen (patient rep.) Dominik Irnich Karen Sherman*

Mac Beckner Dr Wayne B. Jonas Hans Trampisch

Benno Brinkhaus Kai Kronfeld Jorge Vas

Hans-Christoph Diener Lixing Lao Andrew J. Vickers**

Brian Berman George Lewith Norbert Victor

Remy Coeytaux Klaus Linde* Peter White

Angel M. Cronin Dieter Melchart Lyn Williamson

Nadine Foster Albrecht Molsberger Stefan Willich

Michael Haake Hugh MacPherson* Claudia M. Witt*

Richard Hammerschlag Eric Manheimer

**Chair

*Steering Group

www.acupuncturetrialistscollaboration.org

Question 1: is acupuncture better

than sham (placebo)

acupuncture?

Question 2: is acupuncture better

than standard care, usual care,

waitlist, etc?

Key research questions asked by the

Acupuncture Trialists Collaboration

Inclusion criteria

• Randomised controlled trials (RCTs) of

acupuncture for chronic pain:

Headache/migraine

Osteoarthritis

Back & neck pain

• High quality: unambiguously concealed

random allocation

PRISMA flow diagram

ATC results

Raw data obtained from 29 trials with 17,922 patients:

20 trials with sham controls (5,230

patients)

18 trials with non-acupuncture controls

(14,597 patients)

Results for research question 1

Is acupuncture better than sham

Acupuncture varies across trials

Sham acupuncture controls vary:

Non-needle sham (e.g. inactive TENS)

Needle-based sham Penetrating needles

Non-penetrating needle

At acupuncture points

At non-acupuncture points

Acupuncture vs. Sham

Headache/migraine

Favours control ←→ Favours acupuncture

Acupuncture vs Sham (including outlier Vas trial)

Osteoarthritis pain


Acupuncture vs Sham (including outlier Vas trials)

Musculo-skeletal pain (back and neck)


Indication Effect size

(Fixed effects)

Acupuncture vs. Sham controls (excluding outliers)

Migraine/headache -0.15 (-0.24, -0.07) P<0.001

Osteoarthritis -0.16 (-0.25, -0.07) P<0.001

LBP & Neck Pain -0.23 (-0.33, -0.13) P<0.001

Negative values represent better outcomes

Values in parentheses are 95% confidence intervals

Individual patient data meta-analysis Acupuncture vs. Sham controls (n= 5,230)

Interpretation of effect sizes:

0.8 = LARGE

0.5 = MODERATE

0.3 = SMALL

Results for research question 2

Is acupuncture better than non-acupuncture controls

Acupuncture varies across trials

Non-acupuncture controls vary:

No treatment

Wait list

Attention control

Rescue medication

Usual care

Other standard treatment

Acupuncture vs. Non-acupuncture controls

Headache/migraine



Osteoarthritis pain


Acupuncture vs. Non-acupuncture controls:

Musculo-skeletal pain


Individual patient data meta-analysis

Acupuncture vs. Sham controls, and

Acupuncture vs. Non-acupuncture controls (n= 14,597 )

Indication Effect size

(Fixed effects)

Acupuncture vs. Sham controls (excluding outliers)



Back & Neck Pain -0.23 (-0.33, -0.13) P<0.001




Back & Neck Pain -0.55 (-0.58, -0.51) P<0.001

Negative values represent better outcomes

Values in parentheses are 95% confidence intervals

Effect sizes: 0.8 = LARGE

0.5 = MODERATE

0.3 = SMALL

Vickers et al. Archives of Internal Medicine,

2012;172(19):1444-1453

…. “significant difference between true and

sham acupuncture indicate that

acupuncture is more than a placebo”

…. “Acupuncture is effective for the

treatment of chronic pain”

“It hardly matters whether it is correct that acupuncture

is better than sham ….

What really matters is that Vickers et al showed that the

difference is far too small to be of the slightest clinical

interest.”

Reference: David Colquhoun, BMJ 2012

http://www.bmj.com/content/345/bmj.e6060?tab=responses

Criticism: What about the size of the

clinical effect?

NSAIDs are the commonly prescribed for

chronic pain

NSAIDs vs. placebo for pain reduction have

similar effect sizes: 0.23 (0.15 to 0.31)

(Bjordal et al BMJ 2004)

NSAIDs have worse safety profile

[in UK 2,000 deaths a year from people

taking NSAIDS for more than 2 months]

(Tramer et al Pain 2000)

Criticism: What about the size of the

clinical effect?

Interpretation for sample sizes

needed for trials in population

*Number needed to show a difference, if one existed,

based on 90% power and 0.05 significance

Estimated

effect size

Approx sample

size* for a two-

arm trial

Medium 0.5 170

Small 0.3 468

True vs. sham

acupuncture

0.2 1,052

Acu style 1 vs.

Acu style 2

0.1 4,204

ATC sub-studies:

1. What characteristics of acupuncture are

associated with better/worse outcome

2. How does the choice of control impact on

effectiveness of acupuncture

3. Do some patients respond better than others,

and if so what are their characteristics?

4. Is there a subset of patients who are “super-

responders” to acupuncture?

5. What is the time course of acupuncture effects?

How long are benefits sustained?

6. How much variation in outcome is there between

practitioners?

MacPherson et al. (2013) PLoS ONE 8(10): e77438.

Sub-study 1: Variation by

acupuncture characteristic

Sub-study 2: Variation with different

types of control

MacPherson et al. (2014) PLoS ONE 9(4): e93739

Individual patient data

All trials “high quality”

Large patient numbers

Limited by the available studies

Limited data collected (?unknown unknowns)

Similarity of analysed studies (e.g. weekly

sessions)

Not enough trials

Limitations & Strengths

Final conclusion

• Acupuncture outperforms sham

– small effect size of ~0.2,

– statistically significant at p<0.001

– similar effect size to NSAIDs vs. placebo (and

safer)

• Acupuncture is effective for chronic pain

– Moderate effect size of ~0.5 (p<0.001) and

which is clinically relevant

Acknowledgments

The Acupuncture

Trialists’ Collaboration

is funded by an R21

(AT004189I) from the

National Center for

Complementary and

Alternative Medicine

(NCCAM) at the

National Institutes of

Health (NIH) to Dr

Vickers and by a grant

from the Samueli

Institute.

This research is

supported in part by the

National Institute for

Health Research (NIHR)

under Programme

Grants for Applied

Research (Grant No. RP-

PG-0707-10186). The

views expressed in this

presentation are those of

the author(s) and not

necessarily those of the

NHS, the NIHR or the

Department of Health.

1. Vickers AJ, Cronin AM, Maschino AC, Lewith G, MacPherson H, Victor N, et al.

Individual patient data meta-analysis of acupuncture for chronic pain: protocol of the

Acupuncture Trialists’ Collaboration. Trials. 2010;11:90.

2. Vickers AJ, Cronin AM, Maschino AC, Lewith G, MacPherson H, Foster NE, et al.

Acupuncture for Chronic Pain: Individual Patient Data Meta-analysis. ArchInternMed. 2012

Sep 10;172(19):1444–53.

3. Vickers AJ, Maschino AC, Lewith G, MacPherson H, Sherman KJ, Witt CM, et al.

Responses to the Acupuncture Trialists’ Collaboration individual patient data meta-analysis.

Acupunct Med. 2013 Mar;31(1):98–100.

4. MacPherson H, Maschino AC, Lewith G, Foster NE, Witt C, Vickers AJ, et al.

Characteristics of acupuncture treatment associated with outcome: an individual patient meta-

analysis of 17,922 patients with chronic pain in randomised controlled trials. PLoS ONE.

2013;8(10):e77438.

5. MacPherson H, Vertosick E, Lewith G, Linde K, Sherman KJ, Witt CM, et al.

Influence of Control Group on Effect Size in Trials of Acupuncture for Chronic Pain: A

Secondary Analysis of an Individual Patient Data Meta-Analysis. PLoS ONE. 2014 Apr

4;9(4):e93739.

6. Vickers AJ, Linde K. Acupuncture for chronic pain. JAMA. 2014 Mar 5;311(9):955–6.

Publications to date

acupuncture and the placebo question: evidence on msk pain ... › assets ›...

Documents