Treatment response and cost-effectiveness analysis
COMPARATIVE EFFICACY AND ACCEPTABILITY OF FIRST- AND SECOND- GENERATION ANTIDEPRESSANTS IN THE ACUTE TREATMENT OF MAJOR DEPRESSION: A MULTIPLE TREATMENTS META-ANALYSIS
Andrea Cipriani, Corrado Barbui, Toshiaki Furukawa, Georgia Salanti, Stefan Leucht,
Eric Ruhe, Sarah Stockton, Julian Higgins, Guy Goodwin, John Geddes
Major depressive disorder is the most prevalent psychiatric disease in the general population,
affecting more than 16% of adults during their lifetime (Kessler et al., 2003). In 2000 the economic
burden of depressive disorders in the US was estimated to be around 80 billion dollars, with more
than 30% of these costs being attributable to direct medical expenses (Greenberg et al., 2003).
Pharmacotherapy plays an important role in the management of major depression. Before the late
1980s, pharmacologic treatment was limited to tricyclic antidepressants (TCAs) and monoamine
oxidase inhibitors (MAOIs). TCAs and MAOIs sometimes are referred to as traditional or first-
generation antidepressants. These drugs are often accompanied by multiple side effects that many
patients find intolerable. TCAs tend to cause anticholinergic effects including dry mouth and eyes,
urinary hesitancy, and sometimes retention and constipation and MAOIs have the potential to
produce hypertensive crises if taken along with certain foods or dietary supplements containing
tyramine. However, even though first-generation antidepressants are no longer agents of choice in
many circumstances, TCAs are still used worldwide, most of all in low and middle income countries
(according to the list of essential medicines issued by the World Health Organization amitriptyline is
one of the two available treatment options for major depression, along with an SSRI fluoxetine -WHO
2011). Newer antidepressant treatments include selective serotonin reuptake inhibitors (SSRIs),
serotonin and norepinephrine reuptake inhibitors (SNRIs), and other second-generation drugs. The
first of the second-generation drugs was introduced to the US market in 1985, when bupropion was
approved for the treatment of major depressive disorders. In 1987, the US Food and Drug
Administration (FDA) approved the first SSRI, fluoxetine. Since then, five other SSRIs have been
introduced in the market between 1991 and 2002: sertraline, paroxetine, citalopram, fluvoxamine and
escitalopram. The SNRIs were first introduced in 1993 with the approval of venlafaxine. In 1994,
nefazodone, which is essentially an SSRI with additional 5-hydroxytryptamine-2 (5-HT2) and 5-
hydroxytryptamine-3 (5-HT3) antagonist properties, was FDA-approved. Mirtazapine, a drug that
exhibits both noradrenergic and serotonergic activity with central autoreceptors, was added in 1996
and duloxetine, a serotonin and norepinephrine reuptake inhibitor, was approved for the treatment
of MDD (and diabetic peripheral neuropathic pain) in 2004. The latest second-generation
antidepressants approved for the treatment of MDD in adults were desvenlafaxine, the major active
metabolite of venlafaxine XR, and agomelatine, with
Several systematic reviews have assessed the comparative efficacy and safety of second-generation
antidepressants but two recent comparative effectiveness reviews have provided the most
comprehensive assessments to date, notwithstanding conflicting interpretation of results (Cipriani et
Multiple treatments meta-analysis (MTM) is a statistical technique that allows both direct and
indirect comparisons to be undertaken, even when two of the treatments have not been directly
compared (Salanti et al., 2009). MTM can summarise randomised controlled trials (RCTs) of several
different treatments providing point estimates (together with 95% credibility intervals [CIs]) for their
association with a given endpoint, as well as an estimate of inconsistency (that is, a measure of how
well the entire network fits together, with small values suggesting better internal agreement of the
model). MTM has already been used successfully in other fields of medicine (Psaty et al., 2003; Elliott
et al., 2007) and psychiatry (Cipriani et al., 2009; Cipriani et al., 2011). The present review will be
based on our previous MTM on antidepressants (Cipriani et al., 2009), but we will carry out a
different project enlarging the number of antidepressants under investigation, adding new and
clinically informative outcome measures and, most of all, including also placebo controlled trials.
To compare individual first- and second-generation antidepressants in terms of efficacy and
acceptability in the acute treatment of major depression to better inform clinical practice and
Criteria for considering studies for this review
Types of studies
All double-blind RCTs comparing one drug with another within the following group of selected first-
and second-generation antidepressants (namely, agomelatine, amitriptyline, bupropion, citalopram,
clomipramine, desvenlafaxine, duloxetine, escitalopram, fluoxetine, fluvoxamine, hypericum,
milnacipran, mirtazapine, paroxetine, reboxetine, sertraline, trazodone, venlafaxine and vilazodone)
or with placebo, as monotherapy, in the acute phase treatment of depression will be included. RCTs
in which antidepressants were used as an augmentation strategy will be excluded. Quasi-randomized
trials (such as those allocating by using alternate days of the week) will be excluded. For trials which
have a crossover design only results from the first randomisation period will be considered. Cluster
Types of participants
Patients aged 18 or older, of both sexes with a primary diagnosis of major depression. Studies
adopting any standard operationalised criteria to define patients suffering from unipolar major
depression will be included, such as Feighner criteria, Research Diagnostic Criteria, DSM-III, DSM-
III-R, DSM-IV and ICD-10. Studies in which less than 20% of the participants may be suffering from
bipolar depression will be included. A concurrent secondary diagnosis of another psychiatric
disorder will not be considered as exclusion criteria. RCTs in which all participants have a concurrent
primary diagnosis of another Axis I or II disorders will be excluded. Studies in which all participants
have a diagnosis of resistant depression will be excluded. Antidepressant trials in depressive patients
with a serious concomitant medical illness will be excluded. RCTs of women with post-partum
depression will be also excluded, because post-partum depression appears to be clinically different
from major depression (Cooper & Murray, 1998). Trials which allow rescue medications will be
included so long as they are equally provided among the randomised arms.
Primary outcomes: (1) Efficacy (as dichotomous outcome) - Response
Measured by the total number of patients who had a reduction of at least 50% on the total
score between baseline and week 8 (range 4 to 12 weeks) on a standardized rating scale for
depression (Hamilton Depression Rating Scale (HDRS) or another standardised rating
scale, if HDRS was not used). Any version of HDRS will be accepted.
(2) Acceptability of treatment
Treatment discontinuation (acceptability) is defined as the proportion of patients who leave
the study early for any reason during the first 8 weeks of treatment (range 4 to 12 weeks).
Secondary outcomes: (3) Efficacy (as continuous outcome)
Measured by the mean change on the HDRS or Montgomery-Åsberg Depression Rating
Scale (MADRS), if HDRS was not used, after 8 weeks (range 4 to 12 weeks). If none of the
former scales was used, we will consider other standardised rating scales.
(4) Efficacy (as dichotomous outcome) - Remission
Measured by the total number of patients who had a remission of depressive symptoms
between baseline and week 8 (range 4 to 12 weeks) on a standardized rating scale for
depression (HDRS or another standardised rating scale, if HDRS was not used). The
remission will usually be defined as =<7 or 8 on the 17-item HDRS or as =<10 or =<11 on
(5) Tolerability of treatment
The proportion of patients who leave the study early due to adverse events during the first
8 weeks of treatment (range 4 to 12 weeks).
All published and unpublished RCTs that compared the efficacy and acceptability of one
antidepressant with another (see the list of included antidepressants here above) or placebo in the
treatment of major depression will be identified by searching the Cochrane Collaboration CENTRAL
register, AMED, CINAHL, EMBASE, LiLACS, MEDLINE, UK National Research Register,
Trial databases of the following drug-approving agencies (the Food and Drug Administration in the
USA, the Medicines and Healthcare products Regulatory Agency in the UK, the European Medicines
Agency in the EU, the Medicines Evaluation Board in the Netherlands, the MedicalProducts Agency
in Sweden, the Pharmaceuticals and Medical Devices Agency (PMDA) in Japan, the Therapeutic
Goods Administration (TGA) in Australia) and ongoing trial registers (clinicaltrials.gov in the USA,
ISRCTN and National Research Register in the UK, the Netherlands Trial Register, EUDRACT in the
EU, the UMIN-CTR, JapicCTI and JMACCT in Japan, the Australian Clinical Trials Registry and the
WHO International Clinical Trials Registry Platform) will be hand-searched for published,
unpublished and ongoing controlled trials. The Natinal Institue for Clinical Excellence (UK) and the
Institut für Qualität und Wirtschaftlichkeit im Gesundheitswesen (Germany) will also be contacted
for additional information. No language restrictions will be applied. The following phrase will be
used: [depress* or dysthymi* or adjustment disorder* or mood disorder* or affective disorder or affective symptoms] and combined with a list of all included antidepressants. The list will be supplemented by
reference search and personal contacts. All relevant authors will be contacted to supplement the
incomplete report of the original papers.
We are aware that there are many trials carried out in China (Chakrabarti et al., 2007). However, for
many of these studies only incomplete or conflicting information is available and it has been reported
many of them do not use appropriate randomisation procedures (Wu et al., 2006). In an effort to
avoid the potential biases that may be introduced by including these trials without further
information, we will exclude these studies.
Study selection and data extraction
Two persons will independently review references and abstracts retrieved by the search. If both
reviewers agree that the trial doesn’t meet eligibility criteria, we will exclude it. We will obtain the
full text of all remaining articles and use the same eligibility criteria to determine which, if any, to
exclude at this stage. Any disagreements will be solved via discussion with a third member of the
Two reviewers will then independently read each article, evaluate the completeness of the data
abstraction, and confirm the quality rating (see details below). We will design and use a structured
data abstraction form to ensure consistency of appraisal for each study. Information extracted will
include study characteristics (such as lead author, publication year, journal), participant
characteristics (such as diagnostic criteria for depression, age range, setting, diagnosis of bipolar
depression), intervention details (such as dose ranges, mean doses of study drugs) and outcome
measures. A double-entry procedure will be employed by two reviewers.
Length of trial
It is a problem of systematic reviews that usually trials have different durations. Clinically, the
assessment of efficacy after 8 weeks of treatment or after 16 to 24 weeks or more may lead to
differences in terms of treatment outcome. Clinicians need to know whether (and to what extent)
treatments work within a clinically reasonable period of time. Unfortunately, there is no consensus
on what the appropriate duration of an acute phase trial is. In the present review, acute treatment
will be defined as an 8-week treatment in both the efficacy and acceptability analyses (Bauer et al.,
2002). If 8-week data are not available, we will use data ranging between 4 to 12 weeks and the time
point given in the original study as the study endpoint will be given the preference. Longer-term
studies will be excluded if they do not provide data for the 4-12 weeks period.
We will assess risk of bias in the included studies using the tool described in the Cochrane
Collaboration Handbook as a reference guide . . Where inadequate details of
allocation concealment and other characteristics of trials are provided, the trial authors will be
contacted in order to obtain further information. We will not include studies where sequence
generation was at high risk of bias and where allocation was clearly not concealed. The quality
assessment will be done by two independent raters. If the raters disagree, the final rating will be
made by consensus with the involvement (if necessary) of another member of the review group.
Comparability of dosages We will include only studies randomizing patients to drugs within the therapeutic dose (both fixed-
dose and flexible-dose designs will be allowed) (Cipriani et al., 2009). There is the possibility that
some trials compare one agent at the upper limit of its therapeutic range with another agent at the
lower limit of its therapeutic range within the same study. We may look at heterogeneity and then
add a variable (yes/no) that report if dosages are comparable and use this information for sensitivity
Considering that clinical trials of antidepressant drugs are usually small and that data distribution is
difficult to assess for studies with small samples, in this review priority will be given to the use and
analysis of dichotomous variables both for efficacy and acceptability. When dichotomous efficacy
outcomes are not reported but baseline mean and endpoint mean and standard deviation of the
depression rating scales (such as HDRS or MADRS) are provided, we will calculate the number of
responding patients at 8 weeks (range 4 to 12 weeks) employing a validated imputation method
(Furukawa et al., 2005). We are aware that other methods to impute response rate are available and
have been investigated (Anzures-Cabrera et al., 2011). Even though these imputation methods are
valid and may give odds ratios (ORs) with narrower CIs, they only produce logORs and their
variances rather than raw data. As we opt for 2x2 tables to model using the binomial likelihood, the
Fukurawa method will be used in our review. We will use for imputation the endpoint scores for the
following reasons: (i) standardised mean difference should focus on standard deviation of endpoint
scores (standard deviation of change does not represent population variation); (ii) reporting change
may represent outcome reporting bias; (iii) we would need to make up more data to impute standard
deviation of change scores; (iv) observed standard deviation of change is about the same as observed
standard deviation of endpoint. Where outcome data or standard deviations are not recorded,
authors will be asked to supply the data. When only the standard error or t-statistics or p values are
reported, standard deviations will be calculated according to Altman (Altman, 1996). In the absence
of data from the authors, the mean value of known standard deviations will be calculated from the
group of included studies according to Furukawa and colleagues (Furukawa et al., 2006). We will
check that the original standard deviations are properly distributed, so that the imputed standard
deviation represent the average. The continuous efficacy outcome of this review will be the endpoint
score of the HDRS (or MADRS). Dichotomous outcomes will be analysed on an intention-to-treat
(ITT) basis: drop-outs will be assumed to have had negative outcomes and will always be included in
this analysis. When data on drop-outs are carried forward and included in the evaluation (Last
Observation Carried Forward, LOCF), they will be analysed according to the primary studies.
Synthesis of results
We will generate descriptive statistics for trial and study population characteristics across all eligible
trials, describing the types of comparisons and some important variables, either clinical or
methodological (such as year of publication, age, severity of illness, sponsorship, clinical setting). For
each pair-wise comparison between antidepressants, the standardized mean difference (SMD) or
Hedges’s adjusted g will be calculated as the effect size for continuous outcomes and the OR will be
calculated for dichotomous outcomes, both with a 95% CI. We will first perform pair-wise meta-
analyses by synthesizing studies that compare the same interventions using a random effects model
(DerSimonian & Laird, 1986) to incorporate the assumption that the different studies are estimating
different, yet related, treatment effects (Higgins & Green, 2006). Visual inspection of the forest plots
will be used to investigate the possibility of statistical heterogeneity. This will be supplemented
using, primarily, the I2 statistic. This provides an estimate of the percentage of variability due to
heterogeneity rather than a sampling error (Higgins et al., 2003). 95% confidence intervals will be
calculated for I2, and a p value from a standard Q-test for heterogeneity will be used to assess
We will conduct a MTM. MTM is a method of synthesizing information from a network of trials
addressing the same question but involving different interventions. For a given comparison, say A
versus B, direct evidence is provided by studies that compare these two treatments directly.
However, indirect evidence is provided when studies that compare A versus C and B versus C are
analyzed jointly. The combination of the direct and indirect into a single effect size can increase
precision while randomization is respected. The combination of direct and indirect evidence for any
given treatment comparison can be extended when ranking more than three types of treatments
according to their efficacy: every study contributes evidence about a subset of these treatments. We
will perform MTM within a Bayesian framework (Ades et al., 2006). This enables us to estimate the
probability for each intervention of being the best for each positive outcome, given the results of the
MTM. The analysis will be performed using WinBUGS (MRC Biostatistics Unit, Cambridge, U.K.,
the underlying assumptions of the analysis should be investigated carefully. Key among these is that
the network is consistent, meaning that direct and indirect evidence on the same comparisons agree.
Joint analysis of treatments can be misleading if the network is substantially incoherent, i.e., if there is
disagreement between indirect and direct estimates. To evaluate statistically the assumption of
consistency we will employ two approaches. As a first step, we will calculate the difference between
indirect and direct estimates in each closed loop formed in the network and we will subsequently
examine whether there are any material discrepancies (Salanti 2009). To evaluate the assumption in
the network as whole (rather than testing each closed loop) we will employ the design-by-treatment
interaction (White 2011). In case of significant inconsistency, we will investigate possible sources of it.
Inconsistency may result as an uneven distribution of effect modifiers across groups of trials that
compare different treatments. Therefore, we will investigate the distribution of clinical and
methodological variables that we suspect may be potential sources of either heterogeneity or
inconsistency in each comparison-specific group of trials.
Sensitivity analyses and meta-regression analysis
The following effect modifiers will be examined in a meta-regression analysis or in sensitivity
analyses in order to see if they are responsible for heterogeneity and/or inconsistency, if any, and if
they modify the obtained effect estimates (Khan et al., 2004; Posternak & Zimmerman, 2007): baseline
depression severity, number of randomised arms, use of placebo comparator, sponsorship, use of
imputed values. A sensitivity analysis will address whether unbalanced doses affected the results
and we will apply a similar approach as that used in our previous MTM on antidepressants (Cipriani
et al. 2009) to exclude studies with unfair dose comparisons. Further analyses to address dose effects
or other factor possibly related to treatment effect will be performed if necessary. Subgroup analyses
1. Ades AE, Sculpher M, Sutton A, Abrams K, Cooper N, Welton N, Lu G. Bayesian methods for
evidence synthesis in cost-effectiveness analysis. Pharmacoeconomics 2006;24(1):1-19.
2. Altman DG, Bland JM. Detecting skewness from summary information. BMJ 1996;313:1200.
3. Anzures-Cabrera J, Sarpatwari A, Higgins JP
4. Bauer M, Whybrow PC, Angst J, Versiani M, Moller HJ; World Federation of Societies
Biological Psychiatry Task Force on Treatment Guidelines for Unipolar Depressive Disorders.
World Federation of Societies of Biological Psychiatry (WFSBP) Guidelines for Biological
Treatment of Unipolar Depressive Disorders, Part 1: Acute and continuation treatment of
major depressive disorder. World J Biol Psychiatry 2002;3(1):5-43.
5. Chakrabarti A, Adams CE, Rathbone J, Wright J, Xia J, Wong W, Von Reibnitz P, Koenig C,
Baier S, Pfeiffer C, Blatter J, Mantz M, Kloeckner K. Schizophrenia trials in China: a survey.Acta Psychiatr Scand 2007;116(1):6-9.
6. Cipriani A, Furukawa TA, Salanti G, Geddes JR, Higgins J, Churchill R, Watanabe, Nakagawa
AN, Omori IM, McGuire H, Tansella M, Barbui C. Comparative efficacy and acceptability of
12 new generation antidepressants: a multiple treatment meta-analysis. Lancet 2009; 373:746-
Comparative efficacy and acceptability of antimanic drugs in acute
mania: a multiple-treatments meta-analysis.2011; 378:1306-15.
8. Cooper PJ, Murray L. Postnatal depression. BMJ 1998;316(7148):1884-6.
9. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 1986;7(3):177-88
10. Elliott WJ, Meyer PM. Incident diabetes in clinical trials of antihypertensive drugs: a network
meta-analysis. Lancet 2007;369(9557):201-7.
11. Furukawa TA, Cipriani A, Barbui C, Brambilla P, Watanabe N. Imputing response rates from
means and standard deviations in meta-analysis. Int Clin Psychopharm 2005;20(1):49-52.
12. Furukawa TA, Barbui C, Cipriani A, Brambilla P, Watanabe N. Imputing missing standard
deviations in meta-analyses can provide accurate results. J Clin Epidemiol 2006;59(1):7-10.
Benefits and Harms of Second-Generation Antidepressants for Treating Major Depressive
Disorder: An Updated Meta-analy2011;155(11):772-85.
14. Greenberg PE, Kessler RC, Birnbaum HG, Leong SA, Lowe SW, Berglund PA, et al. The
economic burden of depression in the United States: how did it change between 1990 and
2000? J Clin Psychiatry 2003;64(12):1465-75.
15. Salanti G, Marinho V, Higgins JP. A case study of multiple-treatments meta-analysis
demonstrates that covariates should be considered. J Clin Epidemiol 2009; 8:857-64.
16. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses.
17. Higgins JPT, Altman DG, Sterne JAC (editors). Chapter 8: Assessing risk of bias in included
studies. In: Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 (updated March 2011).The Cochrane Collaboration, 2011. Available
18. Khan A, Kolts RL, Thase ME, Krishnan KR & Brown W (2004) Research design features and
patient characteristics associated with the outcome of antidepressant clinical trials. American
19. Kessler RC, Berglund P, Demler O, Jin R, Koretz D, Merikangas KR, et al. The epidemiology of
major depressive disorder: results from the National Comorbidity Survey Replication (NCS-
20. Psaty BM, Lumley T, Furberg CD, Schellenbaum G, Pahor M, Alderman MH, Weiss NS.
Health outcomes associated with various antihypertensive therapies used as first-line agents:
a network meta-analysis. JAMA. 2003 May 21;289(19):2534-44.
21. Posternak MA & Zimmerman M (2007) Therapeutic effect of follow-up assessments on
antidepressant and placebo response rates in antidepressant efficacy trials: meta-analysis.
British Journal of Psychiatry, 190, 287-292
22. White IR. Multivariate random-effects meta-regression. The Stata Journal 2011;11: 255-70
23. World Health Organization (WHO). WHO Model List of Essential Medicines, 17th list - March
24. Wu TX, Li YP, Liu GJ, Bian Z, Li J, Zhang J, Xie L, Ni J. Investigation of authenticity of
'claimed' randomized controlled trials (RCTs) and quality assessment of RCT reports
published in China. Presented at XIV Cochrane Colloquium, Dublin, Ireland, October 23-26
ARBEITSKREIS MEDIZINISCHER ETHIK-KOMMISSIONEN Positionspapier des „Arbeitskreises Medizinischer Ethik- Kommissionen in der Bundesrepublik Deutschland e.V. „Notwendigkeit einer Pflicht zur Studienregistrierung“ Die selektive Publikation von Studienergebnissen schadet Patienten durch eine fehlgeleitete Entschei- dungsfindung B) in der vergleichenden Nutzenbewertung, D) bei
UNIVERSIDAD MARÍA AUXILIADORA - UMA SILABO DE FARMACOLOGÍA Y TERAPÉUTICA DATOS GENERALES: La asignatura de Farmacología y Terapéutica corresponde al área de formación básica y es de naturaleza teórico-práctica obligatoria. Tiene como propósito brindar conocimientos para explicar la acción principal y colateral de fármacos, sus interacciones a nivel de receptores y su relac