Articles: Background

Articles
Eectiveness of non-steroidal anti-inammatory drugs

for the treatment of pain in knee and hip osteoarthritis:
a network meta-analysis
Bruno R da Costa*, Stephan Reichenbach*, Noah Keller, Linda Nartey, Simon Wandel, Peter Jni, Sven Trelle
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.
Lancet 2016; 387: 2093105

Published Online
March 17, 2016
http://dx.doi.org/10.1016/
S0140-6736(16)30002-2
See Comment page 2065
*Both authors contributed
equally to this work
Institute of Primary Health
Care (BIHAM) (B R da Costa PhD,
N Keller MMed, Prof P Jni MD),
Institute of Social and
Preventive Medicine
(S Reichenbach MD, S Trelle MD),
and CTU Bern, Department of
Clinical Research (L Nartey MD,
S Trelle), University of Bern,
Bern, Switzerland; Department
of Rheumatology, Immunology
and Allergology, University
Hospital and University of Bern
(S Reichenbach); Cogitars GmbH
(in Liq), Wangen an der Aare,
Switzerland (S Wandel PhD);
and Applied Health Research
Centre, Li Ka Shing Knowledge
Institute of St Michaels
Hospital (Prof P Jni) and
Department of Medicine
(Prof P Jni), University of
Toronto, Toronto, Canada
Correspondence to:
Dr Sven Trelle, CTU Bern,
University of Bern,
Finkenhubelweg 11, 3012 Bern,
Switzerland
sven.trelle@ctu.unibe.ch
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
prescribed NSAIDs, making them one of the most widely

used drugs in this patient population.6
When prescribing NSAIDs, clinicians are faced with a
myriad of dierent preparations and dosages, which
poses a challenge to clinical decision making. Analyses of
routine data suggest that initial treatment is characterised
by switching between drugs or complete discontinuation.7
Inadequate pain control is probably a major reason for
this approach, which is usually what patients perceive as
the main treatment target.8,9 Several guidelines and
2093
Articles
systematic reviews have investigated the eectiveness of

NSAIDs for treatment of osteoarthritis pain.10,11 However,
these reviews report only the eect of NSAIDs on pain
reduction as compared with placebo and therefore are
only of restricted use for clinical practice. A few systematic
reviews have looked at the comparative eectiveness of
dierent NSAIDs, but considered only direct evidence
and did not address dierent drug preparations or drug
doses.12,13 Network meta-analysis allows an integrated
analysis of all randomised controlled trials that compare
dierent doses of NSAIDs head to head or with placebo
while fully respecting randomisation.14 We assessed the
eectiveness of dierent preparations and doses of
NSAIDs for osteoarthritis pain by integrating all available
direct and indirect evidence in a network meta-analysis.
Methods
Selection criteria
See Online for appendix
We considered large-scale randomised controlled trials of

patients with knee or hip osteoarthritis, comparing any
of the following interventions: NSAIDs, paracetamol
(acetaminophen), or placebo, for the treatment of
osteoarthritis pain. Trials of NSAIDs or treatment groups
within trials of NSAIDs that did not have enough data to
be included in our previously published safety
assessment analysis were not considered (see appendix
for protocol).15 Trials that included patients with diseases
other than osteoarthritis of the knee or hip must either
have reported results separately for the dierent patient
populations or have at least 80% of the included patients
with knee or hip osteoarthritis. Trials must have
randomly assigned, on average, at least 100 patients per
group to minimise bias due to small study eects.16 We
excluded trials published only as abstracts (with no
additional data available from other sources). No
language restrictions were applied.
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
and Drug Administration). The search was last updated

on Feb 24, 2015.
Selection of studies and data extraction

Two investigators independently assessed all trials for
eligibility and extracted data with the help of a translator
for non-English reports. In the case of disagreements,
consensus was reached through discussion. We screened
trials for eligibility, extracted data, and reached
consensus using a standardised, piloted web-based data
management tool for systematic reviews, accompanied
by a codebook. We extracted trial design; trial size;
details of intervention including dose and treatment
duration; patient characteristics such as mean age, sex,
and mean duration of symptoms; duration of follow-up;
type and source of nancial support; type of outcome
(pain or function); and outcome data for each timepoint
of interest. For crossover trials, we extracted data from
the rst period only, because of possible carryover
eects. Whenever necessary, we approximated means
and measures of dispersion from gures in the reports
as previously described.17 We extracted results from
intention-to-treat analyses whenever possible.
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
Articles
score; (8) other algofunctional composite scores;

(9) patients global assessment; (10) physicians global
assessment.5,17 We extracted physical function outcome
data at the following timepoints whenever available:
1 week (2 days), 2 weeks (2 days), and at the end of
treatment if not covered by the specic timepoints.
Risk of bias assessment

We assessed methodological quality of included trials
using a slightly adapted version of the risk of bias
approach of the Cochrane Collaboration (appendix).18
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
window by the median pooled SD recorded across all

timepoints in a trial.22 If SDs were not provided, we
calculated them from SEs or CIs as described elsewhere.17,23
We assessed the goodness of t of the model to the data
by calculating the number of means of standardised
node-based residuals within 196 of the standard normal
distribution; visually inspecting the distribution of
residuals on QQ plots; calculating the heterogeneity of
treatment eects estimated from the posterior median
between trial variance ;24 and calculating the consistency
of the network (determined by the dierence in eect
sizes derived from direct and indirect comparisons).25 To
examine the data for small study eects, we constructed
comparison-adjusted funnel plots.26
To facilitate interpretation of estimated treatment
eects, we calculated several metrics for each intervention.
First, the median rank and 95% CrIs; an intervention
with a median rank of 1 would be interpreted as having
the most benecial eect. Second, the probability for the
eect of the experimental intervention to reach the
minimum clinically important dierence of 037 SD
units, with high probabilities favouring the active
08
07
06
05
09
04
10
03
11
02
12
01
13
23
14
22
15
21
16
20
17
19
18
Figure 1: Network of comparisons included in the analyses

The size of every circle is proportional to the number of randomly assigned patients and indicates the sample size.
The width of the lines corresponds to the number of trials. 01=placebo. 02=paracetamol <2000 mg.
03=paracetamol 3000 mg. 04=paracetamol 39004000 mg. 05=rofecoxib 125 mg. 06=rofecoxib 25 mg.
07=rofecoxib 50 mg. 08=lumiracoxib 100 mg. 09=lumiracoxib 200 mg. 10=lumiracoxib 400 mg.
11=etoricoxib 30 mg. 12=etoricoxib 60 mg. 13=etoricoxib 90 mg. 14=diclofenac 70 mg. 15=diclofenac 100 mg.
16=diclofenac 150 mg. 17=celecoxib 100 mg. 18=celecoxib 200 mg. 19=celecoxib 400 mg. 20=naproxen 750 mg.
21=naproxen 1000 mg. 22=ibuprofen 1200 mg. 23=ibuprofen 2400 mg.
2095
Articles
Interventions
Intervention nodes
(intervention node
number)*
Number of
patients/
proportion
of women
(%)
Follow-up Mean Joint

(weeks)
age
(years)
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)
Placebo (1) vs paracetamol

<2000 mg (2) vs
paracetamol
39004000 mg (4)
483/66%
12
62
Knee and
hip
Yes
Unclear
Low/unclear
Low
Baerwald et al (2010)
Placebo vs naproxen
(500 mg/bid)
Placebo (1) vs naproxen

1000 mg (21)
810/66%
15
63
Hip
Yes
Unclear
Low/unclear
Low
Bensen et al (1999)
Placebo vs celecoxib
(50 mg/bid) vs celecoxib
(200 mg/bid) vs naproxen
(500 mg/bid)
Placebo (1) vs celecoxib

100 mg (17) vs celecoxib
400 mg (19) vs naproxen
1000 mg (21)
1004/72%
12
62
Knee
Yes
Unclear
Low/low
Low
Bin et al (2007)
Celecoxib (200 mg/qid) vs

lumiracoxib (200 mg/qid)
Celecoxib 200 mg (18) vs

lumiracoxib 200 mg (9)
703/85%
61
Knee
Yes
Unclear
Low/low
Low
Bingham et al (2007)
(200 mg/qid) vs etoricoxib
(30 mg/qid)

200 mg (18) vs etoricoxib
30 mg (11)
599/67%
26
63
Knee and
hip
No
Unclear
Low/low
High
Bingham et al (2007a)
(30 mg/qid)

30 mg (11)
608/66%
26
62
Knee and
hip
No
Unclear
Low/low
High
Birbara et al (2006)
(200 mg/qid) vs rofecoxib
(125 mg/qid)

200 mg (18) vs rofecoxib
125 mg (5)
395/72%
61
Knee
Yes
Unclear
Low/low
High
Birbara et al (2006a)
(125 mg/qid)

125 mg (5)
413/65%
61
Knee
Yes
Unclear
Low/low
High
Bocanegra et al (1998)
Placebo vs diclofenac
(75 mg/bid)
Placebo (1) vs diclofenac

150 mg (16)
572/69%
63
Knee and
hip
Yes
Unclear
Low/unclear
Low
Boureau et al (2004)
Ibuprofen (200 mg/q6d) vs

paracetamol (500 mg/q6d)
Ibuprofen 1200 mg (22) vs

paracetamol 3000 mg (3)
222/73%
67
Knee and
hip
Yes
Unclear
Low/unclear
High
Cannon et al (2000)
Diclofenac (50 mg/tid) vs

rofecoxib (125 mg/qid) vs
rofecoxib (25 mg/qid)
Diclofenac 150 mg (16) vs

rofecoxib 125 mg (5) vs
rofecoxib 25 mg (6)
784/67%
54
64
Knee and
hip
Yes
Low
Low/unclear
High
Caruso et al (1987)
Placebo vs naproxen
(125 mg/q6d)

750 mg (20)
734/74%
59
Knee and
hip
Yes
Unclear
Low/low
High
Conaghan et al (2013)
(100 mg/bid)

200 mg (18)
1399/66%
12
61
Knee
Yes
Unclear
Low/low
High
Dahlberg et al (2009)

diclofenac (50 mg/bid)

diclofenac 100 mg (15)
925/70%
52
71
Knee and
hip
Yes
Low
Low/low
High
Day et al (2000)
Placebo vs ibuprofen
(800 mg/tid) vs rofecoxib
(25 mg/qid)
Placebo (1) vs ibuprofen

25 mg (6)
809/80%
64
Knee and
hip
Yes
Unclear
Low/low
High
DeLemos et al (2011)
(200 mg/qid)

200 mg (18)
1011/63%
13
60
Knee and
hip
Yes
Unclear
Low/low
High
Doherty et al (2011)
Ibuprofen (400 mg/tid) vs

paracetamol (1000 mg/tid)
Ibuprofen 1200 mg (22) vs

paracetamol 3000 mg (3)
892/49%
13
61
Knee
Yes
Unclear
Low/low
High
Ehrich et al (2001)
Placebo vs rofecoxib
(50 mg/qid)
Placebo (1) vs rofecoxib

50 mg (7)
672/71%
62
Knee and
hip
Yes
Unclear
Low/low
High
Emery et al (2008)

diclofenac (50 mg/tid)

diclofenac 150 mg (16)
249/
12
64
Hip
Yes
Unclear
Low/low
High
Essex et al (2012)

naproxen (500 mg/bid)

naproxen 1000 mg (21)
589/66%
26
60
Knee
Yes
Unclear
Low/low
High
42
(Table continues on next page)
2096
Articles
Interventions
Intervention nodes
(intervention node
number)*
Number of
patients/
proportion
of women
(%)

(weeks)
age
(years)
LastRisk of bias
Concealed
allocation
Patient
blinding/
investigator
binding
Incomplete
outcome
data
(Continued from previous page)

Essex et al (2012a)
(200 mg/qid) vs naproxen
(500 mg/bid)

1000 mg (21)
322/80%
58
Knee
Yes
Unclear
Low/low
High
Essex et al (2014)
(500 mg/bid)

1000 mg (21)
318/66%
60
Knee
Yes
Unclear
Low/low
High
Fleischmann et al (2006) Placebo vs celecoxib

(200 mg/qid) vs lumiracoxib
(400 mg/qid)
1608/66%
200 mg (18) vs lumiracoxib
400 mg (10)
15
61
Knee
Yes
Unclear
Low/low
High
GAIT (2006)
(200 mg/qid)

200 mg (18)
1583/64%
24
59
Knee
Yes
Unclear
Low/low
Low
Gibofsky et al (2003)
(25 mg/qid)

25 mg (6)
477/67%
63
Knee
Yes
Unclear
Low/low
High
Gibofsky et al (2014)
Placebo vs diclofenac
(35 mg/bid) vs diclofenac
(35 mg/tid)
Placebo (1) vs diclofenac

70 mg (14) vs diclofenac
100 mg (15)
305/67%
13
62
Knee and
hip
Yes
Unclear
Low/unclear
High
Gottesdiener et al (2002) Placebo vs etoricoxib

(90 mg/qid)
Placebo (1) vs etoricoxib

90 mg (13)
617/72%
14
61
Knee
Yes
Low
Low/low
High
Hawkey et al (2000)
(50 mg/qid)

50 mg (7)
775/75%
24
62
Knee and
hip
Yes
Unclear
Low/low
High
Herrero-Beaumont et al
(2007)
(1000 mg/tid)

3000 mg (3)
325/86%
258
64
Knee
Yes
Low
Low/low
High
Hochberg et al (2011)
(200 mg/qid)

200 mg (18)
619/63%
12
62
Knee
Yes
Low
Low/low
High
Hochberg et al (2011a)
(200 mg/qid)

200 mg (18)
615/63%
12
62
Knee
Yes
Low
Low/low
High
Karlsson et al (2009)
(25 mg/qid)

25 mg (6)
543/65%
62
Knee and
hip
Yes
Unclear
Low/low
High
Kivitz et al (2001)
(100 mg/qid) vs celecoxib
(200 mg/qid) vs celecoxib
(500 mg/bid)

1000 mg (21)
1061/66%
63
Hip
Yes
Unclear
Low/unclear
High
Kivitz et al (2002)
Placebo vs naproxen
(500 mg/bid)

1000 mg (21)
1019/65%
60
Knee
Yes
Unclear
Low/unclear
High
Kivitz et al (2004)
(125 mg/qid)

125 mg (5)
1042/68%
63
Knee
Yes
Unclear
Low/low
High
Lehmann et al (2005)
(200 mg/qid)
1684/70%
200 mg (9)
13
62
Knee
Yes
Low
Low/low
High
Leung et al (2002)
Placebo vs etoricoxib
(500 mg/bid)

1000 mg (21)
501/78%
12
63
Knee and
hip
Yes
Unclear
Low/low
High
Lohmander et al (2005)
Placebo vs naproxen
(500 mg/bid)

1000 mg (21)
970/73%
59
Knee and
hip
Yes
Unclear
Low/unclear
High
Makarowski et al (2002)
Placebo vs naproxen
(500 mg/bid)

1000 mg (21)
467/68%
12
62
Hip
Yes
Unclear
Low/unclear
High
2097
Articles
Interventions
Intervention nodes
(intervention node
number)*
Number of
patients/
proportion
of women
(%)

(weeks)
age
(years)
LastRisk of bias
Concealed
allocation
Patient
blinding/
investigator
binding
Incomplete
outcome
data

McKenna et al (2001)
(100 mg/bid) vs diclofenac
(50 mg/tid)

200 mg (18) vs diclofenac
150 mg (16)
600/65%
62
Knee
Yes
Unclear
Low/unclear
High
Miceli-Richard et al
(2004)
(1000 mg/qid)

39004000 mg (4)
779/75%
70
Knee
Yes
Unclear
Low/unclear
Low
Novartis (2005)
(200 mg/bid) vs lumiracoxib
(400 mg/qid)

400 mg (10)
408/
66
Knee
Yes
Unclear
Low/low
High
Novartis (2005a)
(200 mg/qid)
1551/
200 mg (9)
13
61
Knee
Yes
Unclear
Low/low
High
Novartis (2006)
(200 mg/qid)
1684/
200 mg (9)
13
62
Knee
Yes
Unclear
Low/low
Low
Novartis (2006a)

lumiracoxib (200 mg/qid)

lumiracoxib 200 mg (9)
703/
61
Knee
Yes
Unclear
Low/low
Low
Novartis (2007)
(100 mg/qid)
1262/62%
100 mg (8)
13
62
Hip
Yes
Unclear
Low/low
Low
PACES (2004)
(200 mg/qid) vs paracetamol
(1000 mg/qid)

200 mg (18) vs
paracetamol
39004000 mg (4)
524/63%
14
64
Knee and
hip
Yes
Unclear
Low/unclear
Low
PACESa (2004)
(200 mg/qid) vs paracetamol
(1000 mg/qid)

200 mg (18) vs
paracetamol
39004000 mg (4)
556/64%
14
64
Knee and
hip
Yes
Unclear
Low/unclear
Low
Prior et al (2014)
(1300 mg/tid)

39004000 mg (4)
542/74%
12
62
Knee and
hip
Yes
Low
Low/low
Low
Puopolo et al (2007)
(30 mg/qid) vs ibuprofen
(800 mg/tid)

30 mg (11) vs ibuprofen
2400 mg (23)
548/76%
12
63
Knee and
hip
Yes
Low
Low/low
High
Reginster et al (2007)
(500 mg/bid)

1000 mg (21)
997/72%
12
63
Knee and
hip
Yes
Unclear
Low/low
High
Rother et al (2007)
(100 mg/bid)

200 mg (18)
397/60%
63
Knee
Yes
Low
Low/unclear
Low
Saag et al (2000)
(25 mg/qid)

25 mg (6)
736/74%
62
Knee and
hip
Yes
Unclear
Low/low
High
Saag et al (2000a)


rofecoxib 25 mg (6)
693/80%
52
62
Knee and
hip
Yes
Unclear
Low/unclear
Low
Schnitzer et al (2004)
Lumiracoxib (400 mg/qid) vs

naproxen (500 mg/bid)
Lumiracoxib 400 mg (10)

vs naproxen 1000 mg (21)
9511/77%
56
64
Knee and
hip
Yes
Low
Low/low
High
Placebo vs naproxen
(500 mg/bid) vs rofecoxib
(25 mg/qid)

25 mg (6)
672/62%
60
Knee
Yes
Unclear
Low/low
High
2098
Articles
Interventions
Number of
patients/
proportion
of women
(%)
Intervention nodes
(intervention node
number)*

(weeks)
age
(years)
LastRisk of bias
Concealed
allocation
Patient
blinding/
investigator
binding
Incomplete
outcome
data

Schnitzer et al (2005a)

paracetamol (1000 mg/qid vs
1578/67%
62
Knee
Yes
Unclear
Low/low
High
Paracetamol (1300 mg/tid) vs Paracetamol

39004000 mg (4) vs
rofecoxib 25 mg (6)
403/58%
60
Knee
Yes
Unclear
Low/low
High
Placebo vs naproxen
(500 mg/bid)

1000 mg (21)
918/70%
13
61
Knee
Yes
Unclear
Low/unclear
High
(100 mg/qid)
1262/62%
100 mg (8)
17
62
Hip
Yes
Unclear
Low/low
Low
Schnitzer et al (2011a)
Placebo vs naproxen
(500 mg/bid)

1000 mg (21)
1020/70%
53
60
Knee
No
Unclear
Low/unclear
High
Sheldon et al (2005)
(200 mg/qid)
1551/62%
200 mg (9)
13
60
Knee
Yes
Unclear
Low/low
Low
Smugar et al (2006)
(25 mg/qid)

25 mg (6)
1521/68%
62
Knee and
hip
Yes
Unclear
Low/low
High
Smugar et al (2006a)
(25 mg/qid)

25 mg (6)
1082/66%
62
Knee and
hip
Yes
Unclear
Low/low
High
Sowers et al (2005)

naproxen (500 mg/bid) vs

naproxen 1000 mg (21) vs
rofecoxib 25 mg (6)
404/60%
12
63
Knee and
hip
Yes
Unclear
Low/unclear
High
Tannenbaum et al
(2004)
(400 mg/qid)
1702/69%
400 mg (10)
13
64
Knee
Yes
Low
Low/low
Low
Weaver et al (2006)
(125 mg/qid)

125 mg (5)
978/70%
63
Knee
Yes
Unclear
Low/low
High
Wiesenhutter et al
(2005)
(30 mg/qid) vs ibuprofen
(800 mg/tid)

30 mg (11) vs ibuprofen
2400 mg (23)
528/70%
12
62
Knee
Yes
Low
Low/low
High
Williams et al (2000)
(200 mg/qid)

200 mg (18)
686/66%
63
Knee
Yes
Low
Low/low
High

paracetamol
39004000 mg (4) vs
rofecoxib 25 mg (6)
treatment. This threshold of 037 SD units is based on

the median minimum clinically important dierence
reported in studies in patients with osteoarthritis.27 An
eect size of 037 corresponds to a dierence of 9 mm on
a 100 mm visual analogue scale. Third, the surface under
the cumulative ranking (SUCRA) line; an intervention
with a SUCRA value of 100 is certain to be the best,
whereas an intervention with 0 is certain to be the worst.28
Analyses were done with Stata (StataCorp LP 2005.

Stata Statistical Software: Release 12. College Station, TX,
USA) and WinBUGS (MRC Biostatistics Unit 2007.
Version 1.4.3 Cambridge, UK).
Role of the funding source

The funder of the study had no role in study design, data
collection, data analysis, data interpretation, or writing of
2099
Articles
Interventions
Intervention nodes
(intervention node
number)*
Number of
patients/
proportion
of women
(%)

(weeks)
age
(years)
LastRisk of bias
Concealed
allocation
Patient
blinding/
investigator
binding
Incomplete
outcome
data

Williams et al (2001)
(200 mg/qid)

200 mg (18)
718/70%
62
Knee
Yes
Unclear
Low/low
High
Wittenberg et al (2006)
(200 mg/bid) vs lumiracoxib
(400 mg/qid)

400 mg (10)
364/58%
65
Knee
Yes
Unclear
Low/low
Low
Yoo et al (2014)

etoricoxib (30 mg/qid)

etoricoxib 30 mg (11)
239/90%
12
63
Knee
Yes
Unclear
Low/low
High
Zacher et al (2003)

etoricoxib (60 mg/qid)

etoricoxib 60 mg (12)
516/80%
63
Knee and
hip
Yes
Unclear
Low/low
High
Zhao et al (1999)
(200 mg/bid) vs naproxen
(500 mg/bid)

1000 mg (21)
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.
Table: Characteristics of included trials

Intervention
Eect size (95% CrI)
Paracetamol <2000 mg
008 (041 to 027)
Paracetamol 3000 mg
017 (066 to 032)
Paracetamol 39004000 mg*
017 (027 to 006)
Rofecoxib 125 mg
043 (050 to 035)
Rofecoxib 25 mg*
051 (058 to 043)
Rofecoxib 50 mg
062 (084 to 039)
Lumiracoxib 100 mg
034 (044 to 024)
Lumiracoxib 200 mg*
034 (043 to 025)
Lumiracoxib 400 mg
043 (057 to 028)
Etoricoxib 30 mg
049 (061 to 037)
Etoricoxib 60 mg*
058 (073 to 043)
Etoricoxib 90 mg
061 (090 to 033)
Diclofenac 70 mg
022 (061 to 016)
Diclofenac 100 mg
034 (058 to 010)
Diclofenac 150 mg*
057 (069 to 046)
Celecoxib 100 mg
016 (029 to 003)
Celecoxib 200 mg
036 (041 to 032)
Celecoxib 400 mg*
033 (045 to 021)
Naproxen 750 mg
005 (042 to 032)
Naproxen 1000 mg*
040 (047 to 033)
Ibuprofen 1200 mg
030 (084 to 025)
Ibuprofen 2400 mg*
043 (055 to 031)
125
10
075
050
Favours active treatment
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
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
Diclofenac 150 mg*
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
Ibuprofen 2400 mg*
900 (4001600)
83
65
Lumiracoxib 400 mg
900 (3001600)
79
65
Naproxen 1000 mg*
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
Celecoxib 400 mg*
1400 (8001800)
27
44
Lumiracoxib 200 mg*
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
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
high risk of bias for any of the methodological quality

items assessed, except for incomplete outcome data,
since 55 (74%) of the 74 trials analysed had excluded at
least one of the randomly assigned patients from their
analysis. 50 (68%) of the 74 trials analysed reported the
use of last-observation-carried-forward for imputation of
missing data, four (5%) did not need to use imputation,
and 20 (27%) did not do an intention-to-treat analysis.
None of the trials reported the use of multiple imputation.
68 (92%) of the trials analysed received nancial funding
from a commercial body; source of funding was unclear
for the other six (8%; table).
Pooled eect sizes suggested that all interventions,
irrespective of dose, improved osteoarthritic pain
symptoms when compared with placebo (gure 2).
For ve interventions (paracetamol <2000 mg/day
and 3000 mg/day, diclofenac 70 mg/day, naproxen
750 mg/day, and ibuprofen 1200 mg/day), not enough
statistical evidence was available to support superiority
when compared with placebo (ie, estimates were also
compatible with a null eect). For ten interventions, eect
sizes exceeded the prespecied minimum clinically
important dierence of 037. 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),
sucient statistical evidence was available to support a
minimum clinically important eect (ie, the probability
that the dierence from placebo is at or below the

prespecied threshold of 037 was at least 95%; gure 3).
Visual inspection of point estimates in gure 2 suggests
that treatment eects increased as drug doses increased,
but CrIs of dierent doses within one preparation
overlapped widely and a corresponding test for a linear
dose eect was only signicant for celecoxib (p=0030),
diclofenac (p=0031), and naproxen (p=0026). Figure 3
shows the median rank of interventions according to their
treatment eect, the probability for the treatment eect of
a specic intervention to reach the minimum clinically
important dierence, and the SUCRA. The top four ranked
interventions were etoricoxib 90 mg/day, rofecoxib
50 mg/day, diclofenac 150 mg/day, and etoricoxib
60 mg/day. Among these interventions, only diclofenac
150 mg/day and etoricoxib 60 mg/day had a 100%
probability to reach the minimum clinically important
dierence. No evidence existed that treatment eects
varied with the duration of treatment (appendix).
Eect sizes suggest that for 20 (95%) of 21 interventions
physical function was improved when compared with
placebo (no data for physical function were available for
etoricoxib 90 mg; gure 4). However, for six interventions
(paracetamol <2000 mg/day or 3000 mg/day, rofecoxib
50 mg/day, diclofenac 70 mg/day, naproxen 750 mg/day,
and ibuprofen 1200 mg/day), not enough evidence exists
(p>005) to support superiority when compared with
2101
Articles
Intervention
Eect size (95% CrI)
004 (030 to 037)
Paracetamol 3000 mg
030 (078 to 019)
014 (025 to 004)
Rofecoxib 125 mg
038 (047 to 029)
Rofecoxib 25 mg*
048 (056 to 040)
Rofecoxib 50 mg
030 (068 to 007)
Lumiracoxib 100 mg
035 (047 to 024)
Lumiracoxib 200 mg*
037 (047 to 027)
Lumiracoxib 400 mg
041 (057 to 025)
Etoricoxib 30 mg
045 (059 to 031)
Etoricoxib 60 mg*
052 (070 to 035)
Diclofenac 70 mg
022 (062 to 017)
Diclofenac 100 mg
039 (072 to 006)
Diclofenac 150 mg*
051 (065 to 038)
Celecoxib 100 mg
021 (036 to 006)
Celecoxib 200 mg
035 (040 to 030)
Celecoxib 400 mg*
032 (047 to 018)
Naproxen 750 mg
003 (039 to 033)
Naproxen 1000 mg*
041 (049 to 033)
Ibuprofen 1200 mg
051 (107 to 007)
Ibuprofen 2400 mg*
038 (051 to 026)

125
10
075
050
Favours active treatment
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.
placebo. In ten interventions, eect sizes exceeded the

prespecied minimum clinically important dierence of
037. For two interventions (diclofenac 150 mg/day and
rofecoxib 25 mg/day), enough evidence exists to support
a minimum clinically important treatment eect.
The model t was good for both pain and physical
function outcomes (appendix). estimates suggest low
statistical heterogeneity for both pain (0011, 95% CrI
00070017) and physical function (0010, 00070015).
No relevant inconsistency was seen for both pain and
physical function (appendix). Analyses with alternative
models for network meta-analysis yielded much the
same results (appendix).
No evidence exists for an interaction between any of
the trial characteristics assessed and the treatment eect
(all p values 014; appendix). An analysis adjusted for
patient blinding was not possible since all trials had low
risk of bias for this item. Comparison-adjusted funnel
plots show no evidence of asymmetry (appendix).
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
had nearly a null eect on pain symptoms at various

doses (eect size of 017, corresponding to 4 mm
dierence on a 100 mm visual analogue scale), diclofenac
150 mg/day had a moderate to large eect size of 057,
corresponding to dierence on a 100 mm visual analogue
scale of 14 mm. This is 15 times the minimum clinically
important dierence for chronic pain of 037. Our
results also suggest that a typical patient with only
osteoarthritis has 100% probability of having a minimum
clinically important improvement when taking diclofenac
150 mg/day, etoricoxib 60 mg/day, or rofecoxib 25 mg/day.
Finally, these ndings are corroborated by similar
treatment eects on pain and physical function within
each of the interventions.
Although our ndings suggest that some NSAIDs have
a clinically relevant treatment eect on osteoarthritis pain,
their benet has to be weighed against their potential
harmful eects.29 For example, previous analyses suggest
that diclofenac increases the risk for cardiovascular events,
especially cardiovascular death, but that the upper gastrointestinal complications are similar to cyclooxygenase-2
(COX-2) inhibitors.15,30 By contrast, naproxen does not
seem to increase cardiovascular risk but substantially
increases the likelihood of upper gastrointestinal
complications. Appropriate drug selection is a major
challenge in patients with osteoarthritis, who are often
elderly with polypharmacy.31 Our study will help to put the
available safety data into perspective.
The length of follow-up in most of the included trials
was short to intermediate, with durations of 3 months or
less. We believe this is an accurate representation of
clinical practice. Nowadays, NSAIDs, in view of their well
established gastrointestinal and cardiovascular harm,15,30
will typically be prescribed on an as-needed basis, with
intermittent short-term to mid-term use in doses as
required, rather than a long-term xed dose. Our network
meta-analysis provides data for the short-term to midterm analgesic eectiveness of dierent NSAIDs and
paracetamol at dierent doses that are relevant to present
practice. However, we acknowledge that some people
might call for additional long-term trials that directly
compare head-to-head, not only the most eective
preparation-dose combinations identied in our network
meta-analysis, but also NSAIDs on a continuous xeddose versus NSAIDs on an as-needed basis, before coming
to denite conclusions. These measures could be achieved
with a two-by-two factorial design, in which diclofenac
150 mg/day is compared with etoricoxib 60 mg/day, either
on a xed or as-needed basis for at least 12 months.
The available safety data do not allow the same extent of
resolution according to dierent doses. The most pertinent
analyses of safety data are by the Coxib and traditional
NSAID Trialists Collaboration,30 who reported on both
gastrointestinal and cardiovascular safety, and by our
group, who reported cardiovascular safety only.15 The Coxib
and traditional NSAID Trialists Collaboration showed that
all NSAIDs are associated with increased gastrointestinal
Articles
risk compared with placebo. Both groups recorded the

same results for cardiovascular toxic eects, with naproxen
being the only exception. Because of the low number of
events per each specic dose, neither group could
distinguish between dierent doses of a specic NSAID.
The quality of our analysis is limited by the quality of the
underlying data. Whether investigators in some trials were
properly blinded was unclear, and most trials had a high
risk of incomplete outcome data biasbecause they used
the inappropriate approach of last-observation carriedforward to impute missing data. However, by disregarding
small studies, we minimised the risk of biases due to small
study eects.16 The generally high quality of included trials
could be viewed as reassuring for this approach. Moreover,
results were much the same after we controlled for trial
characteristics that could potentially confound our results.
Although the overall number of included patients was
fairly large in view of the continuous nature of our
outcomes, the number of individual trials assessing
individual doses was still low. Although not a limitation for
statistical power, the small number of studies included
could be of concern for the generalisability of our results.
Conrmation of clinical trial results is required by
regulatory authorities to substantiate claims of clinical
eectiveness of drugs, and the importance of independent
validation of research results is now also acknowledged
in the scientic community.32,33 Most of the included
trials were multicentric and therefore replication might
not be an issue.34 Various doses of individual preparations
were assessed in the available trials. We chose to separate
the various doses by analysing them according to daily
doses. This approach hampers statistical eciency by
introducing additional parameters in the statistical
model but allows exible modelling and exploration of
any dose eects. However, data were too sparse to explore
eects of dierent schedules, which might also aect
treatment eectiveness because of dierences in the
pharmacokinetics of various preparations.35
We used a comprehensive search strategy and searched
pertinent sources to retrieve potentially eligible
randomised controlled trials. We therefore believe that it
is unlikely that we missed any relevant trials. We used
various statistical models to analyse the available data
and to increase robustness of our results. All analytical
approaches provided qualitatively the same results.
Moreover, results were also consistent across the two
outcomes of pain and function. The robustness and
accuracy of the results are further corroborated by the
low between-trial heterogeneity, absence of inconsistency,
and good model t. However, the main limitations of our
primary model need to be acknowledged. First, since
study-specic covariance estimates are rarely reported,
the dependency of outcome data over time within
participants is only approximately represented through
the random walk. Second, if strong temporal patterns
such as a linear trend can be identied in the data, the
random walk model used in our analysis cannot account
for it properly. Consequently, the primary model does not

allow assessment of whether treatment eects varied
over time. Study-specic covariance estimates would be
needed but were not available. The CrIs of estimates per
timepoint presented in the appendix are therefore
probably too narrow. However, the analysis still allows
the conclusion that variation over time was minimum.
One of the most comprehensive systematic reviews13 so
far included 273 studies of NSAIDs, irrespective of the
study design, and did not do a meta-analysis. On the basis
of a qualitative assessment, the authors concluded that the
dierent preparations of NSAIDs had no clear dierence
between them. Another large systematic review,12 including
93 randomised clinical trials, compared non-selective and
COX-2-selective NSAIDs and concluded that no dierence
existed between the two classes.12 Machado and colleagues36
focused on the eects of paracetamol, and included only
placebo-controlled trials in a standard, pair-wise metaanalysis. On the basis of ten trials that used a dose similar
to our highest dose group for paracetamol, they showed an
eect that was similar to that in our analysis.36
As opposed to these previous studies, our network
meta-analysis integrates all available high-quality
randomised evidence on the eectiveness of NSAIDs to
treat osteoarthritis pain in one analysis while fully
preserving randomisation. The integration of direct and
indirect comparisons results in a gain of statistical
precision compared with previous analyses and allows
for formal comparisons of NSAIDs with paracetamol
and placebo. This integration also facilitates
interpretation because it makes comparisons between
the dierent interventions explicit.
We are aware of two studies37,38 related to osteoarthritis
that also integrate direct and indirect evidence in one
network meta-analysis. Bannuru and colleagues37 did
not investigate all preparations considered in our
analysis. Notably, the latest COX-2 inhibitors were
missing in their analysis. Moreover, the network metaanalysis37 combined oral and intra-articular treatments
in one analysis, hampering clinical interpretation.
Finally, the number of patients in their analysis was
substantially lower than that in our analysis and no
consideration was given to the longitudinal nature of
the data. van Walsem and colleagues38 also did a
comprehensive network meta-analysis. However, their
analysis included fewer preparations than did our
analysis and only one dose per preparation. Their
analysis did not integrate data over time, instead
providing results only for 6 and 12 weeks separately.
Nevertheless, their results are qualitatively similar to
ours; diclofenac 150 mg/day is more eective than the
other preparations in the analysis except for etoricoxib
60 mg/day, which showed much the same eectiveness.
Thus, the eectiveness of NSAIDs varies substantially,
with evidence of dose dependency. Our analysis suggests
that paracetamol is clinically ineective and should not
be recommended for the symptomatic treatment of
2103
Articles
osteoarthritis, irrespective of the dose. Conversely,

diclofenac at the maximum daily dose of 150 mg/day is
most eective for the treatment of pain and physical
disability in osteoarthritis, and superior to the maximum
doses of frequently used NSAIDs, including ibuprofen,
naproxen, and celecoxib. Etoricoxib at the maximum
dose of 60 mg/day is as eective as diclofenac 150 mg/day
for treatment of pain, but its eect estimates on physical
disability are imprecise. Moreover, etoricoxib is not as
accessible as diclofenac because it has marketing
authorisation in fewer countries. Etoricoxib 90 mg/day
and rofecoxib 50 mg/day, which are both above the
approved maximum daily doses, might be more eective
for pain treatment, but estimates are imprecise. No
evidence exists that supramaximum doses of any of the
other NSAIDs further increase eectiveness. Since all
NSAIDs with available safety data were associated with
clinically relevant gastrointestinal and cardiovascular
harms,15,30 type and dose of NSAID should be chosen on
the basis of the analgesic eectiveness reported in this
analysis. In view of the well established harms of NSAIDs
and the treatment duration in almost all the trials
included in this analysis, intermittent short-term use of
NSAIDs in moderate to maximum doses as required
should be given preference over long-term xed doses.
Contributors
BRdC, SR, PJ, and ST were responsible for the conception and design of
the study. BRdC, SW, and ST did the analysis and interpreted the
analysis in collaboration with SR, NK, LN, and PJ. BRdC, NK, and LN
were responsible for the acquisition of data. BRdC and ST wrote the rst
draft of the Article. All authors critically revised the Article for important
intellectual content and approved the nal version. SR, PJ, and ST
obtained public funding.
For an up-to-date list of CTU

Berns conflicts of interest see
http://www.ctu.unibe.ch/
research/conflicts_of_interest/
index_eng.html
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
Nuesch E, Dieppe P, Reichenbach S, Williams S, I S, Juni P.

All cause and disease specic mortality in patients with knee or hip
osteoarthritis: population based cohort study. BMJ 2011; 342: d1165.
Porcheret M, Jordan K, Jinks C, Croft P. Primary care treatment of
knee paina survey in older adults. Rheumatology (Oxford) 2007;
46: 1694700.
Juni P, Reichenbach S, Dieppe P. Osteoarthritis: rational approach to
treating the individual. Best Pract Res Clin Rheumatol 2006; 20: 72140.
Gore M, Tai KS, Sadosky A, Leslie D, Stacey BR. Use and costs of
prescription medications and alternative treatments in patients with
osteoarthritis and chronic low back pain in community-based
settings. Pain Pract 2012; 12: 55060.
Gore M, Sadosky A, Leslie D, Tai KS, Seleznick M. Patterns of
therapy switching, augmentation, and discontinuation after
initiation of treatment with select medications in patients with
osteoarthritis. Clin Ther 2011; 33: 191431.
Arboleya LR, de la Figuera E, Soledad Garcia M, Aragon B.
Management pattern for patients with osteoarthritis treated with
traditional non-steroidal anti-inammatory drugs in Spain prior to
introduction of Coxibs. Curr Med Res Opin 2003; 19: 27887.
Cordero-Ampuero J, Darder A, Santillana J, Caloto MT, Nocea G.
Evaluation of patients and physicians expectations and attributes of
osteoarthritis treatment using Kano methodology.
Qual Life Res 2012; 21: 1391404.
McAlindon TE, Bannuru RR, Sullivan MC, et al. OARSI guidelines
for the non-surgical management of knee osteoarthritis.
Osteoarthritis Cartilage 2014; 22: 36388.
Bjordal JM, Klovning A, Ljunggren AE, Slordal L. Short-term
ecacy of pharmacotherapeutic interventions in osteoarthritic knee
pain: a meta-analysis of randomised placebo-controlled trials.
Eur J Pain 2007; 11: 12538.
Chen YF, Jobanputra P, Barton P, et al. Cyclooxygenase-2 selective
non-steroidal anti-inammatory drugs (etodolac, meloxicam,
celecoxib, rofecoxib, etoricoxib, valdecoxib and lumiracoxib) for
osteoarthritis and rheumatoid arthritis: a systematic review and
economic evaluation. Health Technol Assess 2008; 12: 1278.
Chou R, McDonagh MS, Nakamoto E, Grin J. Analgesics for
osteoarthritis: an update of the 2006 comparative eectiveness
review. Rockville, MD: Agency for Healthcare Research and Quality
(US), 2011.
Higgins JP, Whitehead A. Borrowing strength from external trials
in a meta-analysis. Stat Med 1996; 15: 273349.
Trelle S, Reichenbach S, Wandel S, et al. Cardiovascular safety of
non-steroidal anti-inammatory drugs: network meta-analysis.
BMJ 2011; 342: c7086.
Nuesch E, Trelle S, Reichenbach S, et al. Small study eects in
meta-analyses of osteoarthritis trials: meta-epidemiological study.
BMJ 2010; 341: c3515.
Reichenbach S, Sterchi R, Scherer M, et al. Meta-analysis:
chondroitin for osteoarthritis of the knee or hip. Ann Intern Med
2007; 146: 58090.
Higgins G, Green S, eds. Cochrane handbook for systematic
reviews of interventions version 5.1.0. Updated March, 2011.
London: The Cochrane Collaboration, 2011.
Smith TC, Spiegelhalter DJ, Thomas A. Bayesian approaches to
random-eects meta-analysis: a comparative study. Stat Med 1995;
14: 268599.
Lu G, Ades AE. Combination of direct and indirect evidence in
mixed treatment comparisons. Stat Med 2004; 23: 310524.
Cooper NJ, Sutton AJ, Lu G, Khunti K. Mixed comparison of stroke
prevention treatments in individuals with nonrheumatic atrial
brillation. Arch Intern Med 2006; 166: 126975.
Cohen J. Statistical power analysis for the behavioral sciences,
2nd edn. Hillsdale: Lawrence Earlbaum, 1988.
Follmann D, Elliott P, Suh I, Cutler J. Variance imputation for
overviews of clinical trials with continuous response.
J Clin Epidemiol 1992; 45: 76973.
Spiegelhalter DJ, Abrams KR, Myles JP. Bayesian approaches to
clinical trials and health care evaluation. Chichester: John Wiley &
Sons, 2004.
Dias S, Welton NJ, Caldwell DM, Ades AE. Checking consistency in
mixed treatment comparison meta-analysis. Stat Med 2010;
29: 93244.
Articles
26
27
28
29
30
31
32
Chaimani A, Higgins JP, Mavridis D, Spyridonos P, Salanti G.

Graphical tools for network meta-analysis in STATA.
PLoS One 2013; 8: e76654.
Wandel S, Juni P, Tendal B, et al. Eects of glucosamine,
chondroitin, or placebo in patients with osteoarthritis of hip or
knee: network meta-analysis. BMJ 2010; 341: c4675.
Salanti G, Ades AE, Ioannidis JP. Graphical methods and numerical
summaries for presenting results from multiple-treatment
meta-analysis: an overview and tutorial. J Clin Epidemiol 2011;
64: 16371.
Adebajo A. Non-steroidal anti-inammatory drugs for the treatment
of pain and immobility-associated osteoarthritis: consensus
guidance for primary care. BMC Fam Pract 2012; 13: 23.
Bhala N, Emberson J, Merhi A, et al. Vascular and upper
gastrointestinal eects of non-steroidal anti-inammatory drugs:
meta-analyses of individual participant data from randomised trials.
Lancet 2013; 382: 76979.
Cahir C, Fahey T, Teeling M, Teljeur C, Feely J, Bennett K.
Potentially inappropriate prescribing and cost outcomes for older
people: a national population study. Br J Clin Pharmacol 2010;
69: 54352.
Food and Drug Administration. Guidance for industry providing
clinical evidence of eectiveness for human drug and biological
products. Rockville, MD: Food and Drug Administration, 1998.
33
34
35
36
37
38
Wadman M. NIH mulls rules for validating key results.

Nature 2013; 500: 1416.
Johnson AF. Replication of multicenter clinical trials for ecacy?
J Chronic Dis 1985; 38: 26569.
Tacconelli S, Capone ML, Patrignani P. Clinical pharmacology of
novel selective COX-2 inhibitors. Curr Pharm Des 2004; 10: 589601.
Machado GC, Maher CG, Ferreira PH, et al. Ecacy and safety of
paracetamol for spinal pain and osteoarthritis: systematic review
and meta-analysis of randomised placebo controlled trials.
BMJ 2015; 350: h1225.
Bannuru RR, Schmid CH, Kent DM, Vaysbrot EE, Wong JB,
McAlindon TE. Comparative eectiveness of pharmacologic
interventions for knee osteoarthritis: a systematic review and
network meta-analysis. Ann Intern Med 2015; 162: 4654.
van Walsem A, Pandhi S, Nixon RM, Guyot P, Karabis A,
Moore RA. Relative benet-risk comparing diclofenac to other
traditional non-steroidal anti-inammatory drugs and
cyclooxygenase-2 inhibitors in patients with osteoarthritis or
rheumatoid arthritis: a network meta-analysis.
Arthritis Res Ther 2015; 17: 66.
2105

Articles: Background

Uploaded by

Copyright:

Available Formats

Articles: Background

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Articles: Background

Uploaded by

Copyright:

Available Formats

Articles

Eectiveness of non-steroidal anti-inammatory drugs

Lancet 2016; 387: 2093105

prescribed NSAIDs, making them one of the most widely

systematic reviews have investigated the eectiveness of

See Online for appendix

We considered large-scale randomised controlled trials of

and Drug Administration). The search was last updated

Selection of studies and data extraction

score; (8) other algofunctional composite scores;

Risk of bias assessment

window by the median pooled SD recorded across all

Figure 1: Network of comparisons included in the analyses

Follow-up Mean Joint

Placebo (1) vs paracetamol

Placebo (1) vs naproxen

Placebo (1) vs celecoxib

Celecoxib (200 mg/qid) vs

Celecoxib 200 mg (18) vs

Placebo (1) vs celecoxib

Placebo (1) vs celecoxib

Placebo (1) vs celecoxib

Placebo (1) vs celecoxib

Placebo (1) vs diclofenac

Ibuprofen (200 mg/q6d) vs

Ibuprofen 1200 mg (22) vs

Diclofenac (50 mg/tid) vs

Diclofenac 150 mg (16) vs

Placebo (1) vs naproxen

Placebo (1) vs celecoxib

Celecoxib (200 mg/qid) vs

Celecoxib 200 mg (18) vs

Placebo (1) vs ibuprofen

Placebo (1) vs celecoxib

Ibuprofen (400 mg/tid) vs

Ibuprofen 1200 mg (22) vs

Placebo (1) vs rofecoxib

Celecoxib (200 mg/qid) vs

Celecoxib 200 mg (18) vs

Celecoxib (200 mg/qid) vs

Celecoxib 200 mg (18) vs

(Table continues on next page)

www.thelancet.com Vol 387 May 21, 2016

Follow-up Mean Joint

(Continued from previous page)

Placebo (1) vs celecoxib

Placebo (1) vs celecoxib

Fleischmann et al (2006) Placebo vs celecoxib

Placebo (1) vs celecoxib

Placebo (1) vs celecoxib

Placebo (1) vs diclofenac

Gottesdiener et al (2002) Placebo vs etoricoxib

Placebo (1) vs etoricoxib

Placebo (1) vs ibuprofen

Placebo (1) vs paracetamol

Placebo (1) vs celecoxib

Placebo (1) vs celecoxib

Placebo (1) vs rofecoxib

Placebo (1) vs celecoxib