Articles: Background
Articles: Background
Articles: Background
Summary
Background Non-steroidal anti-inammatory drugs (NSAIDs) are the backbone of osteoarthritis pain management.
We aimed to assess the eectiveness of dierent preparations and doses of NSAIDs on osteoarthritis pain in a
network meta-analysis.
Methods For this network meta-analysis, we considered randomised trials comparing any of the following
interventions: NSAIDs, paracetamol, or placebo, for the treatment of osteoarthritis pain. We searched the Cochrane
Central Register of Controlled Trials (CENTRAL) and the reference lists of relevant articles for trials published
between Jan 1, 1980, and Feb 24, 2015, with at least 100 patients per group. The prespecied primary and secondary
outcomes were pain and physical function, and were extracted in duplicate for up to seven timepoints after the start
of treatment. We used an extension of multivariable Bayesian random eects models for mixed multiple treatment
comparisons with a random eect at the level of trials. For the primary analysis, a random walk of rst order was used
to account for multiple follow-up outcome data within a trial. Preparations that used dierent total daily dose were
considered separately in the analysis. To assess a potential doseresponse relation, we used preparation-specic
covariates assuming linearity on log relative dose.
Findings We identied 8973 manuscripts from our search, of which 74 randomised trials with a total of 58 556 patients
were included in this analysis. 23 nodes concerning seven dierent NSAIDs or paracetamol with specic daily dose of
administration or placebo were considered. All preparations, irrespective of dose, improved point estimates of pain
symptoms when compared with placebo. For six interventions (diclofenac 150 mg/day, etoricoxib 30 mg/day, 60 mg/day,
and 90 mg/day, and rofecoxib 25 mg/day and 50 mg/day), the probability that the dierence to placebo is at or below a
prespecied minimum clinically important eect for pain reduction (eect size [ES] 037) was at least 95%. Among
maximally approved daily doses, diclofenac 150 mg/day (ES 057, 95% credibility interval [CrI] 069 to 046) and
etoricoxib 60 mg/day (ES 058, 073 to 043) had the highest probability to be the best intervention, both with 100%
probability to reach the minimum clinically important dierence. Treatment eects increased as drug dose increased,
but corresponding tests for a linear dose eect were signicant only for celecoxib (p=0030), diclofenac (p=0031), and
naproxen (p=0026). We found no evidence that treatment eects varied over the duration of treatment. Model t was
good, and between-trial heterogeneity and inconsistency were low in all analyses. All trials were deemed to have a low
risk of bias for blinding of patients. Eect estimates did not change in sensitivity analyses with two additional statistical
models and accounting for methodological quality criteria in meta-regression analysis.
Interpretation On the basis of the available data, we see no role for single-agent paracetamol for the treatment of
patients with osteoarthritis irrespective of dose. We provide sound evidence that diclofenac 150 mg/day is the most
eective NSAID available at present, in terms of improving both pain and function. Nevertheless, in view of the safety
prole of these drugs, physicians need to consider our results together with all known safety information when
selecting the preparation and dose for individual patients.
Funding Swiss National Science Foundation (grant number 405340-104762) and Arco Foundation, Switzerland.
Introduction
Osteoarthritis is the most common form of joint disease
and the leading cause of pain in elderly people.1 Pain
symptoms associated with osteoarthritis result in
increased physical and walking disability, which in turn
increase the risk of all-cause mortality.13 Management of
osteoarthritis pain is based on a sequential hierarchical
approach, with non-steroidal anti-inammatory drugs
(NSAIDs) being the main form of treatment.4,5 In the
USA, about 65% of patients with osteoarthritis are
www.thelancet.com Vol 387 May 21, 2016
Articles
Methods
Selection criteria
Identication of trials
We searched the Cochrane Central Register of Controlled
Trials (CENTRAL) for eligible trials (appendix) from
Jan 1, 1980, to Feb 24, 2015, using the search terms:
osteoarthriti* OR osteoarthro* OR gonarthriti* OR
gonarthro* OR coxarthriti* OR coxarthro* OR arthros*
OR arthrot*. Although CENTRAL includes all
randomised trials indexed in Embase and MEDLINE,
indexing of trials from these databases in CENTRAL
might be delayed. To avoid missing relevant trials
because of this time lag, we also searched Embase and
MEDLINE from Jan 1, 2009, to Feb 24, 2015. Furthermore,
we screened for potential trials in an internal database of
musculoskeletal trials consisting of 721 trials. We then
screened reference lists of all obtained articles, including
relevant reviews, and searched ClinicalTrials.gov for
trials in progress. In the case of incomplete data, we
searched for additional data in ClinicalTrials.gov, WHOapproved trial registries, company-specic trial registries,
and documents available on the website of the US Food
2094
Endpoints
Our prespecied primary outcome was pain. If data for
more than one pain scale were provided for a trial, we
referred to a previously described hierarchy of pain
related outcomes and extracted data for the pain scale that
was highest on this list: (1) global pain score; (2) pain on
walking; (3) WOMAC osteoarthritis index pain subscore;
(4) composite pain scores other than WOMAC; (5) pain
on activities other than walking (such as stair climbing);
(6) WOMAC global score; (7) Lequesne osteoarthritis
index global score; (8) other algofunctional composite
scores; (9) patients global assessment; (10) physicians
global assessment.5,17 In this list, global pain takes
precedence over pain on walking and the Western Ontario
and McMaster Universities Osteoarthritis Index
(WOMAC) pain subscores. We extracted pain outcome
data at the following timepoints whenever available:
1 week (2 days), 2 weeks (2 days), 4 weeks (1 week),
6 weeks (1 week), 3 months (1 month), 6 months
(1 month), 12 months (1 month), and at the end of
treatment if not covered by the specic timepoints.
The secondary outcome was physical function. If data
for more than one physical function scale were provided
for a trial, we referred to a previously described hierarchy
of physical function related outcomes, in which global
measures take precedence over more complex scales,
and extracted data on the physical function scale that was
highest on this list: (1) global function score; (2) walking
disability; (3) WOMAC osteoarthritis index physical
function subscore; (4) composite physical function scores
other than WOMAC; (5) physical function on activities
other than walking (such as stair climbing); (6) WOMAC
global score; (7) Lequesne osteoarthritis index global
www.thelancet.com Vol 387 May 21, 2016
Articles
Statistical methods
For the analysis of eect sizes, we used an extension of
multivariable Bayesian random eects models for mixed
multiple treatment comparisons (appendix).19,20 It fully
preserves the direct randomised comparisons within
each trial, but allows the comparison of all available
treatments across trials, and accounts for multiple
comparisons in trials with more than two treatment
groups.21 The model includes a random eect at the level
of trials, and uses a random walk to account for
correlation of outcome data reported at various timepoints within a trial. The model assumes that, for any
trial, the outcome data recorded at a specied timepoint
are more similar to the outcome data recorded at adjacent
timepoints immediately before and after than at nonadjacent, more remote timepoints. In this sense, the
model borrows strength across timepoints for an
estimate. For all analyses reported herein, we xed the
timepoint at 6 weeks but did a sensitivity analysis with
the timepoint xed at week 1 (appendix).
To assess the robustness of the results obtained by the
primary model, we did two sensitivity analyses
(appendix). These sensitivity analyses investigated
dierent assumptions about the potential relation
between time and treatment eect. Furthermore, we
adjusted the results of the primary outcome for trial
characteristics (ie, concealment of allocation, therapist
blinding, completeness of outcome data, last-observationcarried-forward as imputation method, and whether
patients with knee, hip, or knee and hip osteoarthritis
were included in the analysis) by incorporating a
regression coecient in the model. Corresponding twosided p values for interaction between treatment eects
and trial characteristic were estimated from the posterior
distribution. To assess potential doseresponse relations,
we introduced preparation-specic covariates, assuming
linearity on log relative dose (appendix). To assess
whether treatment eects varied over time, we did
separate analyses per timepoint.
For all variables, minimally informative prior
distributions were chosen (appendix), and all estimates
reported are posterior medians with corresponding
95% credibility intervals (CrIs), unless stated otherwise.
Eect sizes were calculated by dividing the dierence in
mean values between treatment groups in a specic time
www.thelancet.com Vol 387 May 21, 2016
07
06
05
09
04
10
03
11
02
12
01
13
23
14
22
15
21
16
20
17
19
18
2095
Articles
Interventions
Intervention nodes
(intervention node
number)*
Number of
patients/
proportion
of women
(%)
LastRisk of bias
observationcarriedforward
Concealed
allocation
Patient
blinding/
investigator
binding
Incomplete
outcome
data
Altman et al (2007)
Placebo vs paracetamol
(650 mg/tid) vs paracetamol
(1300 mg/tid)
483/66%
12
62
Knee and
hip
Yes
Unclear
Low/unclear
Low
Baerwald et al (2010)
Placebo vs naproxen
(500 mg/bid)
810/66%
15
63
Hip
Yes
Unclear
Low/unclear
Low
Bensen et al (1999)
Placebo vs celecoxib
(50 mg/bid) vs celecoxib
(100 mg/bid) vs celecoxib
(200 mg/bid) vs naproxen
(500 mg/bid)
1004/72%
12
62
Knee
Yes
Unclear
Low/low
Low
Bin et al (2007)
703/85%
61
Knee
Yes
Unclear
Low/low
Low
Bingham et al (2007)
Placebo vs celecoxib
(200 mg/qid) vs etoricoxib
(30 mg/qid)
599/67%
26
63
Knee and
hip
No
Unclear
Low/low
High
Bingham et al (2007a)
Placebo vs celecoxib
(200 mg/qid) vs etoricoxib
(30 mg/qid)
608/66%
26
62
Knee and
hip
No
Unclear
Low/low
High
Birbara et al (2006)
Placebo vs celecoxib
(200 mg/qid) vs rofecoxib
(125 mg/qid)
395/72%
61
Knee
Yes
Unclear
Low/low
High
Birbara et al (2006a)
Placebo vs celecoxib
(200 mg/qid) vs rofecoxib
(125 mg/qid)
413/65%
61
Knee
Yes
Unclear
Low/low
High
Bocanegra et al (1998)
Placebo vs diclofenac
(75 mg/bid)
572/69%
63
Knee and
hip
Yes
Unclear
Low/unclear
Low
Boureau et al (2004)
222/73%
67
Knee and
hip
Yes
Unclear
Low/unclear
High
Cannon et al (2000)
784/67%
54
64
Knee and
hip
Yes
Low
Low/unclear
High
Caruso et al (1987)
Placebo vs naproxen
(125 mg/q6d)
734/74%
59
Knee and
hip
Yes
Unclear
Low/low
High
Conaghan et al (2013)
Placebo vs celecoxib
(100 mg/bid)
1399/66%
12
61
Knee
Yes
Unclear
Low/low
High
Dahlberg et al (2009)
925/70%
52
71
Knee and
hip
Yes
Low
Low/low
High
Day et al (2000)
Placebo vs ibuprofen
(800 mg/tid) vs rofecoxib
(125 mg/qid) vs rofecoxib
(25 mg/qid)
809/80%
64
Knee and
hip
Yes
Unclear
Low/low
High
DeLemos et al (2011)
Placebo vs celecoxib
(200 mg/qid)
1011/63%
13
60
Knee and
hip
Yes
Unclear
Low/low
High
Doherty et al (2011)
892/49%
13
61
Knee
Yes
Unclear
Low/low
High
Ehrich et al (2001)
Placebo vs rofecoxib
(125 mg/qid) vs rofecoxib
(25 mg/qid) vs rofecoxib
(50 mg/qid)
672/71%
62
Knee and
hip
Yes
Unclear
Low/low
High
Emery et al (2008)
249/
12
64
Hip
Yes
Unclear
Low/low
High
Essex et al (2012)
589/66%
26
60
Knee
Yes
Unclear
Low/low
High
42
2096
Articles
Interventions
Intervention nodes
(intervention node
number)*
Number of
patients/
proportion
of women
(%)
LastRisk of bias
observationcarriedforward
Concealed
allocation
Patient
blinding/
investigator
binding
Incomplete
outcome
data
Placebo vs celecoxib
(200 mg/qid) vs naproxen
(500 mg/bid)
322/80%
58
Knee
Yes
Unclear
Low/low
High
Essex et al (2014)
Placebo vs celecoxib
(200 mg/qid) vs naproxen
(500 mg/bid)
318/66%
60
Knee
Yes
Unclear
Low/low
High
1608/66%
Placebo (1) vs celecoxib
200 mg (18) vs lumiracoxib
200 mg (9) vs lumiracoxib
400 mg (10)
15
61
Knee
Yes
Unclear
Low/low
High
GAIT (2006)
Placebo vs celecoxib
(200 mg/qid)
1583/64%
24
59
Knee
Yes
Unclear
Low/low
Low
Gibofsky et al (2003)
Placebo vs celecoxib
(200 mg/qid) vs rofecoxib
(25 mg/qid)
477/67%
63
Knee
Yes
Unclear
Low/low
High
Gibofsky et al (2014)
Placebo vs diclofenac
(35 mg/bid) vs diclofenac
(35 mg/tid)
305/67%
13
62
Knee and
hip
Yes
Unclear
Low/unclear
High
617/72%
14
61
Knee
Yes
Low
Low/low
High
Hawkey et al (2000)
Placebo vs ibuprofen
(800 mg/tid) vs rofecoxib
(25 mg/qid) vs rofecoxib
(50 mg/qid)
775/75%
24
62
Knee and
hip
Yes
Unclear
Low/low
High
Herrero-Beaumont et al
(2007)
Placebo vs paracetamol
(1000 mg/tid)
325/86%
258
64
Knee
Yes
Low
Low/low
High
Hochberg et al (2011)
Placebo vs celecoxib
(200 mg/qid)
619/63%
12
62
Knee
Yes
Low
Low/low
High
Hochberg et al (2011a)
Placebo vs celecoxib
(200 mg/qid)
615/63%
12
62
Knee
Yes
Low
Low/low
High
Karlsson et al (2009)
Placebo vs rofecoxib
(25 mg/qid)
543/65%
62
Knee and
hip
Yes
Unclear
Low/low
High
Kivitz et al (2001)
Placebo vs celecoxib
(100 mg/qid) vs celecoxib
(200 mg/qid) vs celecoxib
(400 mg/qid) vs naproxen
(500 mg/bid)
1061/66%
63
Hip
Yes
Unclear
Low/unclear
High
Kivitz et al (2002)
Placebo vs naproxen
(500 mg/bid)
1019/65%
60
Knee
Yes
Unclear
Low/unclear
High
Kivitz et al (2004)
Placebo vs rofecoxib
(125 mg/qid)
1042/68%
63
Knee
Yes
Unclear
Low/low
High
Lehmann et al (2005)
Placebo vs celecoxib
(200 mg/qid) vs lumiracoxib
(100 mg/qid) vs lumiracoxib
(200 mg/qid)
1684/70%
Placebo (1) vs celecoxib
200 mg (18) vs lumiracoxib
100 mg (8) vs lumiracoxib
200 mg (9)
13
62
Knee
Yes
Low
Low/low
High
Leung et al (2002)
Placebo vs etoricoxib
(60 mg/qid) vs naproxen
(500 mg/bid)
501/78%
12
63
Knee and
hip
Yes
Unclear
Low/low
High
Lohmander et al (2005)
Placebo vs naproxen
(500 mg/bid)
970/73%
59
Knee and
hip
Yes
Unclear
Low/unclear
High
Makarowski et al (2002)
Placebo vs naproxen
(500 mg/bid)
467/68%
12
62
Hip
Yes
Unclear
Low/unclear
High
2097
Articles
Interventions
Intervention nodes
(intervention node
number)*
Number of
patients/
proportion
of women
(%)
LastRisk of bias
observationcarriedforward
Concealed
allocation
Patient
blinding/
investigator
binding
Incomplete
outcome
data
Placebo vs celecoxib
(100 mg/bid) vs diclofenac
(50 mg/tid)
600/65%
62
Knee
Yes
Unclear
Low/unclear
High
Miceli-Richard et al
(2004)
Placebo vs paracetamol
(1000 mg/qid)
779/75%
70
Knee
Yes
Unclear
Low/unclear
Low
Novartis (2005)
Placebo vs celecoxib
(200 mg/bid) vs lumiracoxib
(200 mg/qid) vs lumiracoxib
(400 mg/qid)
408/
66
Knee
Yes
Unclear
Low/low
High
Novartis (2005a)
Placebo vs celecoxib
(200 mg/qid) vs lumiracoxib
(100 mg/qid) vs lumiracoxib
(200 mg/qid)
1551/
Placebo (1) vs celecoxib
200 mg (18) vs lumiracoxib
100 mg (8) vs lumiracoxib
200 mg (9)
13
61
Knee
Yes
Unclear
Low/low
High
Novartis (2006)
Placebo vs celecoxib
(200 mg/qid) vs lumiracoxib
(100 mg/qid) vs lumiracoxib
(200 mg/qid)
1684/
Placebo (1) vs celecoxib
200 mg (18) vs lumiracoxib
100 mg (8) vs lumiracoxib
200 mg (9)
13
62
Knee
Yes
Unclear
Low/low
Low
Novartis (2006a)
703/
61
Knee
Yes
Unclear
Low/low
Low
Novartis (2007)
Placebo vs celecoxib
(200 mg/qid) vs lumiracoxib
(100 mg/qid)
1262/62%
Placebo (1) vs celecoxib
200 mg (18) vs lumiracoxib
100 mg (8)
13
62
Hip
Yes
Unclear
Low/low
Low
PACES (2004)
Placebo vs celecoxib
(200 mg/qid) vs paracetamol
(1000 mg/qid)
524/63%
14
64
Knee and
hip
Yes
Unclear
Low/unclear
Low
PACESa (2004)
Placebo vs celecoxib
(200 mg/qid) vs paracetamol
(1000 mg/qid)
556/64%
14
64
Knee and
hip
Yes
Unclear
Low/unclear
Low
Prior et al (2014)
Placebo vs paracetamol
(1300 mg/tid)
542/74%
12
62
Knee and
hip
Yes
Low
Low/low
Low
Puopolo et al (2007)
Placebo vs etoricoxib
(30 mg/qid) vs ibuprofen
(800 mg/tid)
548/76%
12
63
Knee and
hip
Yes
Low
Low/low
High
Reginster et al (2007)
Placebo vs etoricoxib
(60 mg/qid) vs naproxen
(500 mg/bid)
997/72%
12
63
Knee and
hip
Yes
Unclear
Low/low
High
Rother et al (2007)
Placebo vs celecoxib
(100 mg/bid)
397/60%
63
Knee
Yes
Low
Low/unclear
Low
Saag et al (2000)
Placebo vs ibuprofen
(800 mg/tid) vs rofecoxib
(125 mg/qid) vs rofecoxib
(25 mg/qid)
736/74%
62
Knee and
hip
Yes
Unclear
Low/low
High
Saag et al (2000a)
693/80%
52
62
Knee and
hip
Yes
Unclear
Low/unclear
Low
Schnitzer et al (2004)
9511/77%
56
64
Knee and
hip
Yes
Low
Low/low
High
Schnitzer et al (2005)
Placebo vs naproxen
(500 mg/bid) vs rofecoxib
(25 mg/qid)
672/62%
60
Knee
Yes
Unclear
Low/low
High
2098
Articles
Interventions
Number of
patients/
proportion
of women
(%)
Intervention nodes
(intervention node
number)*
LastRisk of bias
observationcarriedforward
Concealed
allocation
Patient
blinding/
investigator
binding
Incomplete
outcome
data
Schnitzer et al (2009)
1578/67%
62
Knee
Yes
Unclear
Low/low
High
403/58%
60
Knee
Yes
Unclear
Low/low
High
Schnitzer et al (2010)
Placebo vs naproxen
(500 mg/bid)
918/70%
13
61
Knee
Yes
Unclear
Low/unclear
High
Schnitzer et al (2011)
Placebo vs celecoxib
(200 mg/qid) vs lumiracoxib
(100 mg/qid)
1262/62%
Placebo (1) vs celecoxib
200 mg (18) vs lumiracoxib
100 mg (8)
17
62
Hip
Yes
Unclear
Low/low
Low
Schnitzer et al (2011a)
Placebo vs naproxen
(500 mg/bid)
1020/70%
53
60
Knee
No
Unclear
Low/unclear
High
Sheldon et al (2005)
Placebo vs celecoxib
(200 mg/qid) vs lumiracoxib
(100 mg/qid) vs lumiracoxib
(200 mg/qid)
1551/62%
Placebo (1) vs celecoxib
200 mg (18) vs lumiracoxib
100 mg (8) vs lumiracoxib
200 mg (9)
13
60
Knee
Yes
Unclear
Low/low
Low
Smugar et al (2006)
Placebo vs celecoxib
(200 mg/qid) vs rofecoxib
(125 mg/qid) vs rofecoxib
(25 mg/qid)
1521/68%
62
Knee and
hip
Yes
Unclear
Low/low
High
Smugar et al (2006a)
Placebo vs celecoxib
(200 mg/qid) vs rofecoxib
(25 mg/qid)
1082/66%
62
Knee and
hip
Yes
Unclear
Low/low
High
Sowers et al (2005)
404/60%
12
63
Knee and
hip
Yes
Unclear
Low/unclear
High
Tannenbaum et al
(2004)
Placebo vs celecoxib
(200 mg/qid) vs lumiracoxib
(200 mg/qid) vs lumiracoxib
(400 mg/qid)
1702/69%
Placebo (1) vs celecoxib
200 mg (18) vs lumiracoxib
200 mg (9) vs lumiracoxib
400 mg (10)
13
64
Knee
Yes
Low
Low/low
Low
Weaver et al (2006)
Placebo vs rofecoxib
(125 mg/qid)
978/70%
63
Knee
Yes
Unclear
Low/low
High
Wiesenhutter et al
(2005)
Placebo vs etoricoxib
(30 mg/qid) vs ibuprofen
(800 mg/tid)
528/70%
12
62
Knee
Yes
Low
Low/low
High
Williams et al (2000)
Placebo vs celecoxib
(100 mg/bid) vs celecoxib
(200 mg/qid)
686/66%
63
Knee
Yes
Low
Low/low
High
Articles
Interventions
Intervention nodes
(intervention node
number)*
Number of
patients/
proportion
of women
(%)
LastRisk of bias
observationcarriedforward
Concealed
allocation
Patient
blinding/
investigator
binding
Incomplete
outcome
data
Placebo vs celecoxib
(100 mg/bid) vs celecoxib
(200 mg/qid)
718/70%
62
Knee
Yes
Unclear
Low/low
High
Wittenberg et al (2006)
Placebo vs celecoxib
(200 mg/bid) vs lumiracoxib
(400 mg/qid)
364/58%
65
Knee
Yes
Unclear
Low/low
Low
Yoo et al (2014)
239/90%
12
63
Knee
Yes
Unclear
Low/low
High
Zacher et al (2003)
516/80%
63
Knee and
hip
Yes
Unclear
Low/low
High
Zhao et al (1999)
Placebo vs celecoxib
(50 mg/bid) vs celecoxib
(100 mg/bid) vs celecoxib
(200 mg/bid) vs naproxen
(500 mg/bid)
1004/72%
12
62
Knee
Yes
Unclear
Low/unclear
High
Full references for all trials are given in the appendix. bid=twice a day. tid=three times a day. qid=four times a day. q6d=six times a day. *Intervention node number corresponds to those given in the legend of
gure 1.Industry funded trial. Unclear whether industry funded.
Paracetamol <2000 mg
Paracetamol 3000 mg
Rofecoxib 125 mg
Rofecoxib 25 mg*
Rofecoxib 50 mg
Lumiracoxib 100 mg
Lumiracoxib 400 mg
Etoricoxib 30 mg
Etoricoxib 60 mg*
Etoricoxib 90 mg
Diclofenac 70 mg
Diclofenac 100 mg
Celecoxib 100 mg
Celecoxib 200 mg
Naproxen 750 mg
Ibuprofen 1200 mg
125
10
075
050
025
025
p=083
p=046
p=024
p=077
p=0031
p=0030
p=0026
p=050
050
Favours placebo
Figure 2: Estimates of the treatment eects on pain for dierent daily doses of NSAIDs and paracetamol
compared with placebo
Analysis considers data from all timepoints as available. Area between dashed lines shows the treatment eect
estimates below the minimum clinically important dierence. Two-sided p values are derived from tests of linear
doseeect. NSAID=non-steroidal anti-inammatory drug. CrI=credibility interval. *Maximum approved daily dose.
2100
the report. All authors had full access to all the data in the
study and the corresponding author had nal
responsibility for the decision to submit for publication.
Results
We identied 8973 reports, of which 74 randomised clinical
trials investigating seven dierent NSAIDs and paracetamol
were described and included in the analysis (appendix).
23 nodes were included in our network meta-analysis. Each
of the nodes concerned dierent interventions with specic
daily dose of administration, or placebo (gure 1). Celecoxib
200 mg/day was the most frequently investigated
intervention (39 trials), whereas four interventions were
investigated by only one trial (table).
Across trials, the mean age of patients ranged from
58 to 71 years, the percentage of female patients ranged
from 49% to 90%, and the median follow-up was 12 weeks
(range 152 weeks). In total, 58 556 patients were
included in our primary analysis of osteoarthritis pain.
The interventions with the most randomly assigned
patients were celecoxib 200 mg/day (11 411 patients) and
naproxen 1000 mg/day (8195 patients), whereas the
interventions with the fewest randomly assigned patients
were diclofenac 70 mg/day (104 patients) and etoricoxib
90 mg/day (112 patients).
All trials were judged to have a low risk of bias for
blinding of patients, 73% for blinding of therapists, 26%
for incomplete outcome data, and 19% for concealment
of allocation. None of the trials was thought to have a
www.thelancet.com Vol 387 May 21, 2016
Articles
Median rank
(95% CrI)
Intervention
Probability
to reach MID
SUCRA (%)
Rofecoxib 50 mg
200 (1001100)
98
92
Etoricoxib 90 mg
200 (1001400)
95
89
300 (100800)
100
91
Etoricoxib 60 mg*
300 (100800)
100
91
Rofecoxib 25 mg*
500 (200900)
100
82
Etoricoxib 30 mg
600 (2001200)
98
78
Rofecoxib 125 mg
900 (5001400)
93
65
900 (4001600)
83
65
Lumiracoxib 400 mg
900 (3001600)
79
65
1100 (7001500)
78
58
Celecoxib 200 mg
1300 (9001600)
41
50
Diclofenac 100 mg
1400 (3002000)
40
48
Lumiracoxib 100 mg
1400 (8001800)
28
45
1400 (8001800)
27
44
1400 (9001800)
25
44
Ibuprofen 1200 mg
1500 (1002300)
40
46
Diclofenac 70 mg
1700 (3002300)
23
34
Paracetamol 3000 mg
1900 (3002300)
21
29
1900 (16002100)
22
Celecoxib 100 mg
1900 (15002200)
21
Paracetamol <2000 mg
2100 (10002300)
17
Naproxen 750 mg
2100 (10002300)
16
Placebo
2200 (19002300)
1
2 3 4
5 6
Reference
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Rank
Figure 3: Median rank, probability of reaching MID, and SUCRA values of competing interventions and daily doses
MID=minumum clinically important dierence. SUCRA=surface under the cumulative ranking curve. CrI=credibility interval. *Maximum daily dose
Articles
Intervention
Paracetamol <2000 mg
Paracetamol 3000 mg
Rofecoxib 125 mg
Rofecoxib 25 mg*
Rofecoxib 50 mg
Lumiracoxib 100 mg
Lumiracoxib 400 mg
Etoricoxib 30 mg
Etoricoxib 60 mg*
Diclofenac 70 mg
Diclofenac 100 mg
Celecoxib 100 mg
Celecoxib 200 mg
Naproxen 750 mg
Ibuprofen 1200 mg
10
075
050
025
025
050
Favours placebo
Figure 4: Estimates of the treatment eects on physical function for dierent daily doses of NSAIDs and
paracetamol compared with placebo
Analysis considers data from all timepoints as available. Area between dashed lines shows treatment eect
estimates below the minimum clinically important dierence. NSAID=non-steroidal anti-inammatory drug.
CrI=credibility interval. *Maximum approved daily dose.
Discussion
In this network meta-analysis comparing the eectiveness
of dierent treatment regimens of NSAIDs, paracetamol,
or placebo, diclofenac 150 mg/day seemed to be the most
eective in terms of pain and physical function. The
magnitude of treatment eect estimates varied greatly
across dierent NSAIDs and doses. Whereas paracetamol
2102
Articles
Articles
5
6
10
11
12
13
14
Declaration of interests
LN and ST are aliated with CTU Bern, University of Bern, which has a
sta policy of not accepting honoraria or consultancy fees. However,
CTU Bern is involved in design, conduct, or analysis of clinical studies
funded by Abbott Vascular, Ablynx, Amgen, AstraZeneca, Biosensors,
Biotronik, Boehringer Ingelheim, Eisai, Eli Lilly, Exelixis, Geron, Gilead
Sciences, Nestl, Novartis, Novo Nordisc, Padma, Roche, ScheringPlough, St Jude Medical, and Swiss Cardio Technologies. PJ has received
research grants to the institution from AstraZeneca, Biotronik,
Biosensors International, Eli Lilly, and the Medicines Company, and
serves as an unpaid member of the steering group of trials funded by
AstraZeneca, Biotronik, Biosensors, St Jude Medical, and The Medicines
Company. SW is now an employee of Novartis Pharma AG, Biometrics
and Data Management, Oncology. SW was previously an employee of
and currently holds shares in Cogitars GmbH Switzerland (in
liquidation). All other authors declare no competing interests.
15
Acknowledgments
This study was funded by the Swiss National Science Foundation
(grant number 405340-104762) and by a grant from the Arco Foundation,
Switzerland. We thank Kali Tal for proofreading the manuscript and
Toshi Furukawa for support with the development of the search strategy.
21
References
1
Altman R, Brandt K, Hochberg M, et al. Design and conduct of
clinical trials in patients with osteoarthritis: recommendations from
a task force of the Osteoarthritis Research Society. Results from a
workshop. Osteoarthritis Cartilage 1996; 4: 21743.
2
Rosemann T, Wensing M, Joest K, Backenstrass M, Mahler C,
Szecsenyi J. Problems and needs for improving primary care of
osteoarthritis patients: the views of patients, general practitioners
and practice nurses. BMC Musculoskelet Disord 2006; 7: 48.
2104
16
17
18
19
20
22
23
24
25
Articles
26
27
28
29
30
31
32
33
34
35
36
37
38
2105