The British Journal of Psychiatry (2015)

206, 93–100. doi: 10.1192/bjp.bp.114.148551

Review article

Pharmacotherapy for post-traumatic stress

disorder: systematic review and meta-analysis
Mathew Hoskins, Jennifer Pearce, Andrew Bethell, Liliya Dankova, Corrado Barbui, Wietse A. Tol,
Mark van Ommeren, Joop de Jong, Soraya Seedat, Hanhui Chen and Jonathan I. Bisson

Background
Pharmacological treatment is widely used for post-traumatic difference 70.23, 95% CI 70.33 to 70.12). For individual
stress disorder (PTSD) despite questions over its efficacy. pharmacological agents compared with placebo in two or
more trials, we found small statistically significant evidence
Aims of efficacy for fluoxetine, paroxetine and venlafaxine.
To determine the efficacy of all types of pharmacotherapy,
as monotherapy, in reducing symptoms of PTSD, and to Conclusions
assess acceptability. Some drugs have a small positive impact on PTSD symptoms
Method and are acceptable. Fluoxetine, paroxetine and venlafaxine
A systematic review and meta-analysis of randomised may be considered as potential treatments for the disorder.
controlled trials was undertaken; 51 studies were included. For most drugs there is inadequate evidence regarding
efficacy for PTSD, pointing to the need for more research in
Results this area.
Selective serotonin reuptake inhibitors were found to be
statistically superior to placebo in reduction of PTSD Declaration of interest
symptoms but the effect size was small (standardised mean None.

Post-traumatic stress disorder (PTSD) is a common mental assessed the efficacy of pharmacological treatment compared with
disorder with an estimated prevalence of 15.4% in the most placebo control groups at reducing traumatic stress symptoms in
robust epidemiological studies (those using diagnostic interviews individuals experiencing PTSD.
and random samples) of conflict-affected populations,1 and a
12-month prevalence across the world of 3–4%.2 The disorder
may occur in people of any age who have been exposed to one or
more exceptionally threatening or horrifying events. Characteristic Method
symptoms include re-experiencing, avoidance and hyperarousal.3,4
The disorder is associated with substantial comorbidity, such as All double-blind, randomised, placebo-controlled and comparative
depression, anxiety and substance misuse,5 and significant trials of the pharmacological treatment of PTSD completed from
economic burden.6 Previous meta-analyses of pharmacological October 2005 (to ensure all eligible trials not published at the time
treatment of PTSD have been inconsistent. The UK’s National of the NICE, Cochrane and ACPMH searches would be included)
Institute for Health and Care Excellence (NICE) guidelines found were considered in our primary and additional searches, covering
that only paroxetine, mirtazapine, amitriptyline and phenelzine 13 separate databases. Trials completed before October 2005 that
were significantly superior to placebo.7 Owing to the relatively were included in the NICE, Cochrane and ACPMH reviews were
small effect sizes and sample sizes, none of these drugs was also considered. Published and unpublished abstracts and reports
included as a first-line treatment for PTSD; all were recommended were sought out in any language. Studies were not excluded on the
as second-line treatment after the initiation of trauma-focused basis of differences between them such as sample size and
psychological treatment. The guidelines of the Australian Centre duration. Pharmacotherapy trials in which there was ongoing or
for Posttraumatic Mental Health (ACPMH), consistent with newly initiated trauma-focused psychotherapy or where the
NICE, recommended that pharmacological interventions should experimental medication served as an augmentation agent to
not be used in preference to trauma-focused psychological ongoing pharmacotherapy were excluded. Pharmacotherapy trials
treatment.8 Other reviews have been more positive about in which there was ongoing supportive counselling were allowed
pharmacological treatment, grouping selective serotonin reuptake provided it was not initiated during the course of the treatment,
inhibitors (SSRIs) together and rating them as equivalent to on the basis that this is common in trials and the limited evidence
trauma-focused psychological treatments.9,10 A Cochrane review for supportive counselling.14 Open label trials were not considered.
reported strong benefits,11 but the Institute of Medicine found Our review followed the Preferred Reporting Items for Systemic
inadequate evidence to determine the efficacy of pharmacological Reviews and Meta-analyses (PRISMA) checklist and reporting
treatment for PTSD.12 There are, however, major differences guidance.15
between the methodological quality of these reviews, making
direct comparison problematic.13 Given the inconsistent findings of
previous meta-analyses and the increasing number of randomised
controlled trials (RCTs) of pharmacological treatments, the World Participants
Health Organization (WHO) commissioned an update of the All studies of participants with PTSD according to ICD or DSM
information obtained by the most methodologically robust criteria were eligible.3,4 There was no restriction on the basis of
systematic reviews published to date: those by NICE, ACPMH onset, duration or severity of PTSD symptoms, or on the presence
and the Cochrane Collaboration.7,8,11 We reviewed RCTs that of comorbid disorders, trauma type, age or gender of participants.

93
Hoskins et al

Interventions studies included in previous systematic reviews were double-

Pharmacological treatments for adults with PTSD, in which the extracted by two independent reviewers, cross-checked for accuracy
comparator was a placebo or other medication, were eligible for and any discrepancies were discussed with a third reviewer.
inclusion.
Risk of bias
Outcome measures Two reviewers, using the domain-based evaluation method
The primary outcomes of interest were clinician-administered recommended by the Cochrane Collaboration,16 independently
continuous measures of symptom severity such as the Clinician assessed risk of bias in individual studies for each trial. This
Administered PTSD Scale. Self-rated PTSD symptom scales were method considers the following domains: random sequence
considered separately. The total number of participants who left generation; allocation concealment; masking of participants and
the trial early for any reason was used as a proxy measure of personnel; masking of outcome assessment; incomplete outcome
treatment acceptability. data; selective reporting; and other sources of bias. Any
discrepancies between reviewers’ decisions were resolved by
Search strategy discussion with a third reviewer. Risk of bias across studies was
assessed by considering publication bias through the visual
We conducted a primary bibliographic database search of
examination of funnel plots.
Medline, Medline In Process, EMBASE, the Health Management
Information Consortium (HMIC) database, PsycINFO, ASSIA
and CINAHL using the Ovid interface on 22 February 2013. In Statistical analysis
order to avoid language bias, we separately searched (using the
Review Manager 5 software was used to synthesise data using
same search strategy, and in consultation with regional experts)
meta-analysis and to provide forest plots for dichotomous and
the Japanese, Chinese and WHO regional databases (Spanish,
continuous data. Confidence intervals of 95% were used for all
Russian and Portuguese languages). This initial broad search
analyses. Categorical outcome measures such as leaving the study
was intended to identify not only the RCTs of interest but also
early were analysed using relative risk (RR) calculations. For
reviews of pharmacotherapy for PTSD. The comprehensive search
continuous data, standardised mean differences (SMD) were used.
term used (see online Appendix DS1) was created by amalgamating
The degree of heterogeneity between studies was calculated using
the previous search strategies from the NICE, Cochrane and
the I2 statistic. Where this was less than 30%, indicating a mild
ACPMH guideline reviews with an updated list of medications.
degree of heterogeneity, a fixed effects model was used; a random
Additional searches for published and unpublished studies were
effects model was used when I2 was greater than 30%. Data were
made in the National Center for PTSD PILOTS database, the
analysed from the intention to treat (ITT) sample in the ‘once
Cochrane Library, the Controlled Trials Register, Web of
randomised always analysed’ fashion where possible, to avoid
Knowledge, OpenSIGLE and Google Scholar. Reference lists of
effects of bias from completers-only analyses. Analyses were
all selected studies and reviews were further scrutinised for any
performed for individual drugs and, in order to maximise the
additional RCTs. An expert group was consulted to identify any
information available from data synthesis, whenever possible for
additional studies they were aware of. Authors of identified studies
classes of drugs (e.g. tricyclic antidepressants, SSRIs) and trauma
were contacted to request data if outcome information was
type (e.g. combat trauma, sexual trauma, interpersonal violence
missing.
trauma, all non-combat trauma).

Study selection
One reviewer transferred the initial search hits and studies Results
included in the earlier systematic reviews into EndNoteX4
software for Windows. Duplicates were removed. Two reviewers The database search yielded 13 634 records and an additional 74
then independently screened the titles and abstracts. Studies that were identified from other sources. From these, 4613 duplicates
were clearly irrelevant were excluded; potentially relevant ones were removed leaving 9095 records. These were screened and
were assessed for inclusion as full texts. Any discrepancies between 8064 irrelevant studies were removed. This left 1031 abstracts from
reviewers’ decisions were resolved by discussion with a third which 896 were removed as ineligible. This left 135 full-text
reviewer. articles which were reviewed to find 84 not meeting the inclusion
criteria. A total of 51 studies were included in this systematic
review (Fig. 1). Sixteen new studies were found from our database
Data extraction and risk of bias assessment search.17–32 The remaining 35 were found in previous reviews. An
All data from newly identified studies were double-extracted by additional 28 studies were identified from the NICE guidelines, of
two independent reviewers into a standardised table and any which five did not meet our inclusion criteria.33–37 This left 23
discrepancies were discussed with a third reviewer. Data for studies included from the NICE guidelines (references 38–54 plus
change from baseline to end-point were extracted when available, six unpublished studies A–F in Appendix 1). Thirty-five studies
otherwise end-point data were used. Continuous data were were identified in the Cochrane review, of which 19 were already
extracted for clinician-administered PTSD symptom severity included in the NICE review and removed as duplicates. A further
using the Clinician Administered PTSD Scale as the gold standard; four used completer-only data and were not included;55–58
for self-rated PTSD, the Davidson Trauma Scale was used as the another allowed concomitant psychotropic medications and was
gold standard. If these scales were not used, data from alternative excluded.59 This left an additional 11 studies from the Cochrane
scales were extracted. Dichotomous data were extracted for review,60–70 which were included. There were 12 additional
number of people withdrawing from the trial using total number pharmacotherapy studies identified from the ACPMH guidelines,
of participants in the group as the denominator. One reviewer of which seven were already present in the above reviews and were
entered the outcome data into Review Manager 5 software for removed as duplicates. A further four did not meet our inclusion
Windows, which was then checked by another reviewer. Data from criteria,71–74 leaving one additional study which was included.75

94
 Pharmacotherapy for PTSD

sufficient information to judge the masking of participants,

Records identified through Additional records identified
database searching through other sources
personnel and outcome assessors as being of low risk. Incomplete
(n = 13 634) (n = 740 outcome data were addressed adequately in 15 studies. Thirty-
eight of the studies were deemed to be free from selective reporting.

6 6 Efficacy of pharmacotherapy
Records after removing Data from 21 studies (n = 3932) were available for inclusion in a
duplicates
(n = 9095)
meta-analysis of reduction in severity of PTSD symptoms for
any SSRI v. placebo (Fig. 2). This found a small positive effect
6 for SSRIs when grouped together. The results of meta-analyses for
Full abstracts screened individual drugs, which had been tested against placebo in at least
(n = 1031) two RCTs, are shown in Table 1. Three drugs were significantly
Records excluded superior to placebo on either clinician- and self-rated PTSD
7 (n = 896)
6 symptom severity combined (paroxetine) or clinician-rated PTSD
Full-text articles assessed symptom severity alone (fluoxetine and venlafaxine). We found
for eligibility
(n = 135)
insufficient evidence to support the preferential use of individual
Full-text articles agents in either combat-related or non-combat-related trauma
7 excluded with reasons (Table 2).
6
(n = 84)
Studies included in
qualitative synthesis
(n = 51)
Discussion

6 This systematic review found statistically significant evidence on

Studies included in meta-analysis for three pharmacological agents v. placebo in the
quantitative synthesis treatment of PTSD (fluoxetine, paroxetine and venlafaxine) but
(meta-analysis) no evidence for brofaromine, olanzapine, sertraline or topiramate.
(n = 41 for efficacy)
(n = 35 for acceptability) Four drugs – amitriptyline, GR205171 (a neurokinin-1 antagonist),
mirtazapine and phenelzine – showed superiority over placebo in
single RCTs, whereas eleven did not: alprazolam, citalopram,
Fig. 1 Study search profile. desipramine, escitalopram, imipramine, lamotrigine, nefazadone,
risperidone, tiagabine and valproate semisodium. When meta-
analyses were undertaken by class of drug rather than individual
drug, SSRIs were found to perform better than placebo. The
Description of studies
absence of sufficient data precluded meta-analyses of other drug
The characteristics of the 51 RCTs included in this review are classes. The absence of difference in numbers of individuals
summarised in online Table DS1. Only three of the studies did leaving the study early for any reason suggests that the drugs
not employ a placebo comparator arm.28,65,67 There were 31 SSRI included were well tolerated overall.
trials, 14 of which assessed sertraline,21,26,38,42,43,45,54,61,65,67,68,A,C,D The effect sizes for pharmacological treatments for PTSD
nine assessed fluoxetine,23,40,47,51,52,69,70,75,B seven assessed par- compared with placebo are low and inferior to those reported
oxetine,22,32,50,53,64,E,F one citalopram,68 one escitalopram,27 and one for psychological treatments with a trauma focus over waiting-list
fluvoxamine.28 There were three tricyclic antidepressant trials,42,49,66 or treatment as usual controls.14 They are, however, similar to
and three monoamine oxidase inhibitor (MAOI) trials.48,49,60 those found for antidepressants for depression compared with
Another six trials assessed other antidepressants,18,44,46,65,67,A placebo.76 Unfortunately, the absence of a common control
and a further ten trials assessed other agents.17,18,20,24,25,29,30,39,62,63 condition and head-to-head pharmacological v. psychological
The average sample size was 130 (range 13–538), the average age of treatment trials makes comparison of the relative efficacies of
participants was 41 years (range 18–82) and 54% of participants these treatment approaches difficult. This is compounded by the
were women. The average duration of the trials was 12.4 weeks fact that a significant number of participants in psychological
(range 4–36). Thirty-four studies were based in the USA, five were treatment trials were continuing pre-existing pharmacological
international, four were from Israel and there was one each from treatment at the same time.
Brazil, China, Iran, South Africa and Turkey. The location was It is well accepted that a well-masked, placebo-controlled trial
unclear in three studies. The predominant trauma (450%) was is a tougher test for an experimental intervention than a trial with
combat in 13 studies, physical/sexual assault in eight, physical a waiting-list control. The UK’s NICE guidelines development
assault alone in four, sexual assault alone in four, natural group attempted to address this by determining a higher effect size
disaster in two and road traffic accidents in two. For 18 studies threshold for psychological treatments than pharmacological ones
participants had been exposed to various traumas with no one (0.8 v. 0.5),7 but it is unclear whether such arbitrary cut-offs may
type accounting for over 50%. have introduced bias against either form of treatment. What is
clear is the marked effect of the placebo in several of the trials
reported in this review. For example, in the two venlafaxine
Risk of bias assessments studies the mean reduction in PTSD symptoms of those on
Risk of bias assessments are included in online Table DS1. placebo was greater than 40%.18,47 There is a need for RCTs of
Fourteen of the studies included enough information on adequate specific psychological interventions compared with an adequate
random sequence generation to be judged as having a low risk of non-specific psychological intervention. It is only through such
bias; the remainder were unclear. Only eight studies included an trials and direct head-to-head comparisons that the relative
adequate description of allocation concealment. All 51 studies effectiveness of differing treatment approaches could be
described themselves as ‘double blind’ but only eight provided established. The current meta-analyses are a later iteration of the

95
Hoskins et al

Table 1 Efficacy and tolerability of individual agents v. placebo

Active drug PTSD symptoms (clinician-rated) PTSD symptoms (self-reported) Leaving study early

SSRIs
Paroxetine Four studies,50,53,64,E n = 1134 Three studies,50,53,E n = 1251 Five studies,22,50,53,64,E n = 1307 RR = 0.98
SMD = 70.43 (95% CI 70.53 to 70.29)* SMD = 70.38 (95% CI 70.49 to 70.26)* (95% CI 0.71 to 1.34)
I 2 = 17%a I 2 = 0% I 2 = 52%
Sertraline Nine studies,21,38,42,54,61,68,A,C,D n = 1441 Six studies,21,38,42,A,C,D n = 1257 Eight studies,21,38,42,54,68,A,C,D n = 1387
SMD = -0.13 (95% CI 70.27 to 0.01) SMD = 70.15 (95% CI 70.35 to 0.05) RR = 1.14 (95% CI 0.95 to 1.37)
I 2 = 37% I 2 = 68% I 2 = 15%
Fluoxetine Three studies,23,52,70 n = 771 Three studies,40,47,52 n = 363 Six studies,23,40,47,52,69,70 n = 989
SMD = 70.24 (95% CI 70.41 to 70.08)* SMD = 70.41 (95% CI 70.98 to 0.15) SMD = 1.02 (95% CI 0.72 to 1.45)
I 2 = 0% I 2 = 64% I 2 = 26%
MAOIs
Brofaromine Two studies,48,60 n = 159 NA Two studies,48,60 n = 180
SMD = 70.24 (95% CI 70.81 to 0.33) RR = 1.56 (95% CI 0.87 to 2.79)
I 2 = 63% I 2 = 0%
Other antidepressants
Venlafaxine Two studies,18,A n = 687 Two studies,18,A n = 687 One study,18 n = 329
SMD = 70.20 (95% CI 70.35 to 70.05)* SMD = 70.19 (95% CI 70.4 to 0.01) RR = 0.91 (95% CI 0.67 to 1.25)
I 2 = 0% I 2 = 0%
Antipsychotics
Olanzapine Two studies,31,39 n = 39 Two studies,31,39 n = 39 Two studies,31,39 n = 49
SMD = 70.61 (95% CI 71.27 to 0.05) SMD = 70.50 (95% CI 71.16 to 0.15) RR = 1.15 (95% CI 0.36 to 3.71)
I 2 = 53% I 2 = 0%
Other drugs
Topiramate Two studies,29,30 n = 69 One study,29 n = 38 Two studies,29,30 n = 73
SMD = 70.46 (95% CI 70.94 to 0.02) SMD = 70.6 (95% CI 71.26 to 0.05) RR = 0.92 (95% CI 0.3 to 2.84)
I 2 = 0% I 2 = 38%

MAOI, monoamine oxidase inhibitor; NA, not applicable; PTSD, post-traumatic stress disorder; RR, relative risk; SMD, standardised mean difference; SSRI, selective serotonin
reuptake inhibitor.
a. The I 2 statistic is noted only when 41 study as not applicable otherwise.
*P50.05.

evidence used to determine what to recommend for recent WHO recently updated Australian guidelines also concluded that
guidelines.77 A WHO guidelines development group examined pharmacological interventions should not be preferentially used
this evidence and recommended antidepressants as a second as a routine first treatment of PTSD over trauma-focused
line of treatment of PTSD when psychological interventions psychological treatments.78 There are various other reasons for
with known efficacy did not work or were not available. The considering the prescription of antidepressants, including

Experimental Control
Study or subgroup Mean s.d. Total Mean s.d. Total Weight SMD IV, random, 95% CI SMD IV, random, 95% CI
Brady (2000)38 733 28.1 94 723.2 28.7 93 5.8% 70.34 (70.63, 70.05)
Brady (2005)61 32.56 15.69 49 32.7 28.75 45 4.1% 70.01 (70.41, 0.40)
Connor (1999)40 10.1 9.8 25 20.5 12.6 22 2.4% 70.91 (71.52, 70.31) 8
DavidsonA 739.4 27.12 173 734.17 28.42 179 7.3% 70.19 (70.40, 0.02)
Davidson (2001)42 733 23.76 100 726.2 23.9 108 6.1% 70.28 (70.56, 70.01)
Eli LillyB 710.42 7.5 323 710.59 10.21 88 6.8% 0.02 (70.21, 0.26)
Friedman (2007)21 713.1 27.5 86 715.4 28.07 83 5.6% 0.08 (70.22, 0.38)
Hertzberg (2000)47 47 8 6 42 11 6 0.8% 0.48 (70.68, 1.64) 7

Marshall (2001)50 738.75 27.2 375 725.3 25.8 188 8.0% 70.50 (70.68, 70.32)
Marshall (2004)64 55.6 33.4 25 62.8 40.8 27 2.8% 70.19 (70.73, 0.36)
Martenyi (2002)51 734.6 28.1 226 26.8 26.1 75 6.3% 70.28 (70.54, 70.02)
Martenyi (2007)23 742.85 25.5 323 736.6 25.7 88 6.8% 70.24 (70.48, 70.01)
Panahi (2011)26 722.7 7.3 35 717.5 7.5 35 3.3% 70.69 (71.18, 70.21)
Pfizer 588C 727.4 27.12 94 727.9 28.42 94 5.9% 0.02 (70.27, 0.30)
Pfizer 589D 713.1 27.12 84 715.4 28.42 82 5.6% 0.08 (70.22, 0.39)
Shaleve (2011)27 731.12 29.63 23 727.8 20.13 23 2.5% 70.13 (70.71, 0.45)
SKB627E 736.5 26.1 109 730.8 25.37 103 6.2% 70.22 (70.49, 0.05)
Tucker (2001)53 735.5 24.58 151 724.7 24.98 156 7.0% 70.43 (70.66, 70.21)
Tucker (2003)68 741.82 29.09 23 738.7 29.07 10 1.7% 70.10 (70.85, 0.64)
Van der Kolk (2007)70 733.23 22.11 30 730.95 22.6 29 3.0% 70.10 (70.61, 0.41)
8
Zohar (2002)54 718.7 6.7 23 713.5 6.6 19 2.2% 70.77 (71.40, 70.14)
Total (95% CI)
2377 1553 100.0% 70.23 (70.33, 70.12)
Heterogeneity: t2 = 0.03; w2 = 42.90, d.f. = 20 (P = 0.002); I 2 = 53%
71 70.5 0 0.5 1
Test for overall effect: Z = 4.23 (P50.0001)
Favours experimental Favours control

Fig. 2 Meta-analysis of selective serotonin reuptake inhibitors v. placebo (SMD, standardised mean difference).

96
 Pharmacotherapy for PTSD

Table 2 Trauma type subanalysis of individual agents v. placebo

Natural
disaster/ All non-combat trauma
Active drug Combat Sexual Interpersonal violence accident combined

SSRIs
Paroxetine NA NA One study,64 n = 52 NA One study,64 n = 52
SMD = 70.19 SMD = 70.19
(95% CI 70.73 to 0.36) (95% CI 70.73 to 0.36)
Sertraline Four studies,21,26,54,D n = 447 One study,38 n = 183 Two studies,61,A n = 446 NA Five studies,42,61,A,C,D n = 1008
SMD = 70.26 SMD = 70.30 SMD = 70.15 SMD = 70.11
(95% CI 70.68 to 0.16) (95% CI 70.59 to 70.01)* (95% CI 70.34 to 0.04) (95% CI 70.23 to 0.02)
I 2 = 77%a I 2 = 68% I 2 = 14%
Fluoxetine Two studies,47,52 n = 313 Two studies,23,40 n = 458 One study,70 n = 59 NA Three studies,29,40,47 n = 517
SMD = 70.12 SMD = 70.52 SMD = 70.10 SMD = 70.36
(95% CI 70.73 to 0.50) (95% CI 71.16 to 0.13) (95% CI 70.61 to 0.41) (95% CI 70.75 to 0.02)
I 2 = 37% I 2 = 76% I 2 = 58%
MAOIs
Brofaromine One study,60 n = 114 NA One study,48 n = 55 NA NA
SMD = 0.01 SMD = 70.58
(95% CI 70.36 to 0.38) (95% CI 71.58 to 0.02)
Other antidepressants
Venlafaxine NA NA Two studies,18,A n = 687 NA NA
SMD = 70.20
(95% CI 70.35 to 70.05)*
I 2 = 0%
Antipsychotics
Olanzapine NA One study,39 n = 11 One study,31 n = 28 NA Two studies,31,39 n = 39
SMD = 0.16 SMD = 70.93 SMD = 70.61
(95% CI 71.07 to 1.39) (95% CI 71.71 to 70.14)* (95% CI 71.27 to 0.05)
I 2 = 53%
Other drugs
Topiramate NA One study,29 n = 38 One study,30 n = 32 NA Two studies,29,30 n = 69
SMD = 70.39 SMD = 72.38 SMD = 70.46
(95% CI 71.03 to 0.25) (95% CI 73.33 to 1.43)* (95% CI 70.94 to 0.02)
I 2 = 0%

MAOI, monoamine oxidase inhibitor; NA, not applicable; SMD, standardised mean difference; SSRI, selective serotonin reuptake inhibitor.
a. The I 2 statistic is noted only when >1 study as not applicable otherwise.
*P<0.05.

concurrent depression, lack of availability of psychological for three of the four drugs recommended in the NICE guidelines
treatments, failure to respond to or tolerate psychological (mirtazapine, amitriptyline and phenelzine). Of particular note is
treatments and the personal preference of the individual with that new evidence resulted in drugs not recommended by NICE
PTSD. (fluoxetine and venlafaxine) now meeting the requirements laid
The variation in efficacy between different agents is striking down by NICE for recommendation as a second-line treatment.
and could be explained by a number of factors. First, there may This review used particular rigour in assessing studies. Most of
be real differences between drugs, even those in the same family. the other reviews that have concluded more favourably about
For example, paroxetine’s superior efficacy over sertraline may pharmacotherapy grouped drugs from the same class together,10,11
possibly be explained by its increased dopamine receptor activity. did not include unpublished studies,79,80 and adopted different
Differences in the pharmacology of phenelzine and brofaromine inclusion criteria.8,10,79,80 This probably accounts for differences
are also marked, and grouping all drugs in one class together in conclusions between different meta-analyses despite apparently
seems less desirable than considering each one individually. This having the same raw data available. For example, sertraline
confirms previous observation that class membership does not performed significantly better than placebo in two recent meta-
confer the same level of efficacy in the treatment of PTSD.11 This analyses,79,80 which did not include the two unpublished studies
meta-analysis supports this assertion, with the SSRIs paroxetine included in our review.C,D
and fluoxetine having better evidence of efficacy than sertraline.
It is interesting to note, however, that despite failing to show
superiority over placebo, sertraline showed equivalence to Study limitations
venlafaxine in one of the studies included.A This may be a result There were, however, some important methodological flaws in the
of the strength of placebo effect varying across studies. studies identified in this review as illustrated by the risk of bias
The differences found between paroxetine and sertraline could assessment undertaken. Clinical and statistical heterogeneity was
be due to chance or lack of power, but it is noteworthy that over apparent in the meta-analyses; this makes interpretation and
1000 individuals were included in the meta-analyses of both generalisation across different trauma populations difficult. A
paroxetine and sertraline, which suggests that power was not the major issue was that, despite having systematically tried to access
issue. The same cannot be said for the majority of drugs reviewed, unpublished data on the studies without ITT efficacy data,
the evaluation of which was based on single RCTs with fewer than sufficient data for analysis were only available from only 41 of
100 participants. This review added 15 new RCTs to the evidence the 51 studies included. This raises the risk of outcome reporting
previously reported in systematic reviews but added no new trials bias.81 Incorrect definition of the ITT sample population was a

97
Hoskins et al

problem. This is an important principle where data from every C. Pfizer588. Unpublished data. Appendix 14 of 2nd consultation draft for
person who is randomised should be included in the analysis of the NICE (2005) PTSD guidelines.
efficacy. Participants might leave a study because of adverse D. Pfizer589. Unpublished data. Appendix 14 of 2nd consultation draft for
effects, failed efficacy or greater symptom severity at baseline; the NICE (2005) PTSD guidelines.
not including their data in the analysis leads to bias, as a treatment E. SKB627 Bryson H, Lawrinson S, Edwards GJ, Grotzinger KM. A 12 week,
might appear to be more effective than it is. Another important double-blind, placebo-controlled, parallel group study to assess the
issue to consider is that this review considers drug monotherapy efficacy and tolerability of paroxetine in patients suffering from post-
as opposed to drug or placebo in combination with psychological traumatic stress disorder.
treatment. This evidence cannot, therefore, be used to consider F. SKB650 Bryson H, Dillingham KE, Jeffery PJ. A study of the maintained
whether or not pharmacotherapy augments psychological efficacy and safety of paroxetine versus placebo in the long-term
treatment for PTSD; many people with the disorder do receive both. treatment of posttraumatic stress disorder.
Several of the new studies identified by our search defined
their ITT population as participants who received at least one dose References
of the study medication and received at least one post-baseline
assessment. This could exclude some participants from analysis 1 Steel Z, Chey T, Silove D, Marnane C, Bryant RA, van Ommeren M.
who had received treatment but not been assessed after the Association of torture and other potentially traumatic events with mental
health outcomes among populations exposed to mass conflict and
baseline. The most extreme case was a study of fluoxetine that
displacement: a systematic review and meta-analysis. JAMA 2009; 302:
defined ITT in this way,23 but measured PTSD symptom severity 537–49.
only at baseline and end-point, so was actually a completers 2 Karam EG, Andrews G, Bromet E, Petukhova M, Ruscio AM, Salamoun M,
sample. We decided to include these studies in our analysis but et al. The role of Criterion A2 in the DSM-IV diagnosis of post-traumatic
acknowledge that this affects the reliability of the results. stress disorder. Biol Psychiatry 2009; 68: 465–73.
3 World Health Organization. The ICD-10 Classification of Mental and
Implications Behavioural Disorders: Clinical Descriptions and Diagnostic Guidelines. WHO,
1992.
Pharmacological interventions for PTSD can be effective but the
4 American Psychiatric Association. Diagnostic and Statistical Manual of
magnitude of effect unfortunately is small, and the clinical Mental Disorders Fourth Edition, Text Revision (DSM-IV-TR). APA, 2000.
relevance of this small effect is unclear. This review supports the 5 Brunello N, Davidson JR, Deahl M, Kessler RC, Mendlewicz J, Racagni G, et al.
use of paroxetine, venlafaxine and fluoxetine as pharmacological Posttraumatic stress disorder: diagnosis and epidemiology, comorbidity and
interventions for PTSD. For most drugs, there remains inadequate social consequences, biology and treatment. Neuropsychobiology 2001; 43:
150–62.
evidence regarding efficacy for PTSD, pointing to the need
for more research in this area to confirm the utility of 6 Ferry F, Bolton D, Bunting B, O’Neill S, Murphy S, Devine B. Economic Impact
of Post Traumatic Stress in Northern Ireland. Northern Ireland Centre for
pharmacological treatments for this disorder. Trauma and Transformation and University of Ulster Psychology Research
Institute, 2010.
Mathew Hoskins, MRCPsych, Jennifer Pearce, BSc, Andrew Bethell, BSc, Liliya 7 National Collaborating Centre for Mental Health. Post-traumatic Stress
Dankova, BSc, Cardiff University, Cardiff, UK; Corrado Barbui, MD, Department of Disorder: The Management of PTSD in Adults and Children in Primary and
Public Health and Community Medicine, Section of Psychiatry, University of Verona, Secondary Care. British Psychological Society/Royal College of Psychiatrists,
Verona, Italy; Wietse A. Tol, PhD, Johns Hopkins Bloomberg School of Public Health,
2005.
Baltimore, USA; Mark van Ommeren, PhD, World Health Organization, Geneva,
Switzerland; Joop de Jong, MD, PhD, Vrije Universiteit, Amsterdam, The Netherlands; 8 Australian Centre for Posttraumatic Mental Health. Australian Guidelines for
Soraya Seedat, MD, Stellenbosch University, Stellenbosch, South Africa; Hanhui the Treatment of Adults with Acute Stress Disorder and Posttraumatic Stress
Chen, MD, Shanghai Mental Health Centre, Shanghai Jiaotong University, Shanghai, Disorder. ACPMH, 2007.
China; Jonathan I. Bisson, DM, Cardiff University, Cardiff, UK
9 American Psychiatric Association. Practice Guideline for the Treatment of
Correspondence: Professor Jonathan I. Bisson, Cardiff University, Institute of Patients with Acute Stress Disorder and Posttraumatic Stress Disorder.
Psychological Medicine and Clinical Neurosciences, Hadyn Ellis Building, Maindy American Psychiatric Press, 2004.
Road, Cathays, Cardiff CF24 4HQ, UK. Email [email protected]
10 Foa EB, Keane TM, Friedman MJ, Cohen J, eds. Effective Treatments for
First received 13 Oct 2013, final revision 8 Jun 2014, accepted 16 Jun 2014 PTSD: Practice Guidelines from the International Society for Traumatic Stress
Studies (2 ed). Guilford, 2008.
11 Stein DJ, Zungu-Dirwayi N, Van der Linden GJ, Seedat S. Pharmacotherapy for
Funding posttraumatic stress disorder. Cochrane Database of Systematic Reviews
2004; issue 2.
The World Health Organization supported this work to inform the development of Mental 12 Institute of Medicine. Treatment of PTSD: An Assessment of the Evidence.
Health Gap Action Programme (mhGAP) guidelines on disorders and problems specifically National Academies Press, 2008.
related to stress. The views expressed in this article are those of the authors solely and do
not necessarily represent the views, policies or decisions of their employers. 13 Forbes D, Creamer M, Bisson JI, Cohen JA, Crow BE, Foa EB, et al. A guide to
guidelines for the treatment of PTSD and related conditions. J Trauma Stress
2010; 23: 537–52.
Acknowledgements
14 Bisson JI, Ehlers A, Matthews R, Pilling S, Richards D, Turner S. Psychological
treatments for chronic post-traumatic stress disorder: systematic review and
We acknowledge Jorge Castro and Keiko Sakurai (World Health Organization interns) for
assistance in multi-language searches. meta-analysis. Br J Psychiatry 2007; 190: 97–104.
15 PRISMA Statement. 2009 (http://prisma-statement.org/statement.htm).
16 Higgins JPT, Green S, eds. Cochrane Handbook for Systematic Reviews of
Appendix 1
Interventions Version 5.1.0. Cochrane Collaboration, 2011.

Unpublished studies 17 Connor KM, Davidson JR, Weisler RH, Zhang W, Abraham K. Tiagabine for
posttraumatic stress disorder: effects of open-label and double-blind
Further details of the studies listed below are available from the authors. discontinuation treatment. Psychopharmacology 2006; 184: 21–5.
18 Davidson J, Baldwin D, Stein DJ, Kuper E, Benattia I, Ahmed S, et al.
A. Davidson J, Lipschitz A, Musgnung JJ. Venlafaxine XR and sertraline in Treatment of posttraumatic stress disorder with venlafaxine extended
posttraumatic stress disorder: a placebo-controlled trial. Appendix release: a 6 month randomized controlled trial. Arch Gen Psychiatry 2006;
14 of 2nd consultation draft for the NICE (2005) PTSD guidelines. 63: 1158–65.
19 Davidson JR, Brady K, Mellman TA, Stein MB, Pollack MH. The efficacy and
B. Eli Lilly. Unpublished data. Appendix 14 of 2nd consultation draft for the tolerability of tiagabine in adult patients with post-traumatic stress disorder.
NICE (2005) PTSD guidelines. J Clin Psychopharmacol 2007; 27: 85–8.

98
 Pharmacotherapy for PTSD

20 Davis LL, Davidson JR, Ward LC, Bartolucci A, Bowden CL, Petty F. Divalproex a 28-week double-blind, placebo-controlled study. Am J Psychiatry 2001;
in the treatment of posttraumatic stress disorder. J Clin Psychopharmacol 158: 1974–81.
2008; 281: 84–8.
43 Davidson JR, Rothbaum BO, van der Kolk BA, Sikes CR, Farfel GM.
21 Friedman MJ, Marmar CR, Baker DG, Sikes CR, Farfel GM. Randomized, Multicenter, double-blind comparison of sertraline and placebo in the
double blind comparison of sertraline for posttraumatic stress disorder in a treatment of posttraumatic stress disorder. Arch Gen Psychiatry 2001; 58:
Department of Veterans Affairs setting. J Clin Psychiatry 2007; 68: 711–20. 485–92.
22 Wang H-H, Tan Q-R, Wang Z-Z, Wang H-N, Cheng C-G, Chen Y-C, et al. 44 Davidson JR, Weisler RH, Butterfield MI, Casat CD, Connor KM, Barnett S.
A randomised controlled study on prevention of posttraumatic stress Mirtzapine vs. placebo in posttraumatic stress disorder: a pilot trial. Biol
disorder. J Clin Psychol Med 2009; 1: 1–3. Psychiatry 2003; 53: 188–91.
23 Martenyi F, Brown EB, Caldwell CD. Failed efficacy of fluoxetine in the 45 Davidson JR. Remission in post-traumatic stress disorder (PTSD): effects of
treatment of posttraumatic stress disorder. J Clin Psych 2007; 27: 166–70. sertraline as assessed by the Davidson Trauma Scale, Clinical Global
24 Matthew SJ, Vythilingam M, Murrough JW, Feder A, Luckenbaugh DA, Impressions and the Clinician-Administered PTSD scale. Int Clin
Kinkead B, et al. A selective neurokinin-1 receptor antagonist in chronic Psychopharmacol 2004; 19: 85–7.
PTSD: a randomized, double blind, placebo controlled proof of concept trial. 46 Davis LL, Jewell ME, Ambrose S, Farley J, English B, Bartolucci A. A placebo-
Euro Neuropsychopharmacol 2011; 21: 221–9. controlled study of nefazadone for the treatment of chronic posttraumatic
25 Padala PR, Madison J, Monnahan M, Marcil W, Price P, Ramaswamy S, et al. stress disorder. J Clin Psychopharmacol 2001; 24: 291–7.
Risperidone monotherapy for posttraumatic stress disorder related to sexual 47 Hertzberg MA, Feldman ME, Beckham JC, Kudler HS, Davidson JR. Lack of
assault and domestic abuse in women. Int Clin Psychopharmacol 2006; 21: efficacy for fluoxetine in PTSD: a placebo controlled trial in combat veterans.
275–80. Ann Clin Psychiatry 2000; 12: 101–5.
26 Panahi Y, Moghaddam BR, Sahebkar A, Sahebkar A, Nazari MA, Beiraghdar F,
48 Katz RJ, Lott MH, Arbus P, Crocq L, Herlobsen P, Lingjaerde O.
et al. A randomized, double blind, placebo controlled trial on the efficacy and
Pharmacotherapy of post-traumatic stress disorder with a novel
tolerability of sertraline in Iranian veterans with post-traumatic stress
psychotropic. Anxiety 1994; 1: 169–174.
disorder. Psychol Med 2011; 41: 2159–66.
49 Kosten TR, Frank JB, Dan E, McDougle CJ, Giller EL. Pharmacotherapy for
27 Shalev AY, Ankri Y, Israeli-Shalev Y, Peleg T, Adessky R, Freedman S.
posttraumatic stress disorder using phenelzine or imipramine. J Nerv Ment
Prevention of posttraumatic stress disorder by early treatment. Arch Gen
Dis 1991; 179: 366–70.
Psychiatry 2011; 69: 166–76.
50 Marshall RD, Beebe KL, Oldham M, Zaninelli R. Efficacy and safety of
28 Spivak B, Strous RD, Shaked G, Shabash E, Kotler M, Weizman A.
paroxetine treatment for chronic PTSD: a fixed-dose, placebo-controlled
Reboxetine versus fluvoxamine in treatment of motor vehicle accident
study. Am J Psychiatry 2001; 158: 1982–8.
related posttraumatic stress disorder. J Clin Psychopharmacol 2006; 26:
152–6. 51 Martenyi F, Brown EB, Zhang H, Koke SC, Prakash A. Fluoxetine v. placebo in
29 Tucker P, Trautman RP, Wyatt DB, Thompson J, Wu SC, Capece JA, et al. prevention of relapse in post-traumatic stress disorder. Br J Psychiatry 2002;
Efficacy and safety of topiramate monotherapy in civilian posttraumatic 181: 315–20.
stress disorder: a randomized, double blind, placebo controlled trial. J Clin 52 Martenyi F, Brown EB, Zhang H, Prakash A, Koke SC. Fluoxetine versus
Psychiatry 2007; 68: 201–6. placebo in posttraumatic stress disorder. J Clin Psychiatry 2002; 63: 199–206.
30 Yeh MS, Mari JJ, Costa MC, Andreoli SB, Bressan RA, Mello MF. A double 53 Tucker P, Zaninelli R, Yehuda R, Ruggiero L, Dillingham K, Pitts CD. Paroxetine
blind, randomized controlled trial to study the efficacy of topiramate in a in the treatment of chronic posttraumatic stress disorder: results of a
civilian sample of PTSD. CNS Neurosci Ther 2010; 175: 305–10. placebo-controlled, flexible-dosage trial. J Clin Psychiatry 2001; 62: 860–8.
31 Carey P, Suliman S, Ganesan K, Seedat S, Stein DJ. Olanzapine monotherapy 54 Zohar J, Amital D, Miodownik C, Kotler M, Bleich A, Lane RM. Double-blind
in posttraumatic stress disorder: Efficacy in a randomized, double-blind, placebo-controlled pilot study of sertraline in military veterans with
placebo-controlled study. Hum Psychopharmacol 2012; 27: 386–91. posttraumatic stress disorder. J Clin Psychopharmacol 2002; 22: 190–5.
32 Petrakis IL, Ralevski E, Desai N, Trevisan L, Gueorguieva R, Rounsaville B, 55 Kaplan Z, Amir M, Swartz M, Levine J. Inositol treatment of posttraumatic
et al. Noradrenergic vs serotonergic antidepressant with or without stress disorder. Anxiety 1996; 2: 51–2.
naltrexone for veterans with PTSD and comorbid alcohol dependence.
Neuropsychopharmacology 2012; 37: 996–1004. 56 Chung MY, Min KH, Jun YJ, Kim SS, Kim WC, Jun EM. Efficacy and tolerability
of mirtazapine and sertraline in Korean veterans with posttraumatic stress
33 Cohen H, Kaplan Z, Kotler M, Kouperman I, Moisa R, Grisaru N. Repetitive disorder: a randomized open label trial. Hum Psychopharmacol 2004; 19:
transcranial magnetic stimulation of the right dorsolateral prefrontal cortex 489–94.
in posttraumatic stress disorder: a double-blind, placebo-controlled study.
Am J Psychiatry 2004; 161: 515–24. 57 Shestatzky M, Greenberg D, Lerer B. A controlled trial of phenelzine in
posttraumatic stress disorder. Psychiatr Res 1988; 24: 149–55.
34 Frommberger U, Stieglitz RD, Nyberg E, Richter H, Novelli FU, Angenendt J.
Comparison between paroxetine and behaviour therapy in patients with 58 Smajkic A, Weine S, Djuric-Bijedic Z, Boskailo E, Lewis J, Pavkovic I.
posttraumatic stress disorder (PTSD): a pilot study. Int J Psychiatry Clin Sertraline, paroxetine, and venlafaxine in refugee posttraumatic stress
Practice 2004; 8: 19–23. disorder with depression symptoms. J Trauma Stress 2001; 14: 445–52.
35 Hamner MB, Faldowski RA, Ulmer HG, Frueh BC, Huber MG, Arana GW. 59 Reich DB, Winternitz S, Hennen J, Watts T, Stanculescu C. A preliminary study
Adjunctive risperidone treatment in post-traumatic stress disorder: of risperidone in the treatment of posttraumatic stress disorder related to
a preliminary controlled trial of effects on comorbid psychotic symptoms. childhood abuse in women. J Clin Psychiatry 2004; 65: 1601–5.
Int Clin Psychopharmacol 2003; 18: 1–8.
60 Baker DG, Diamond BI, Gillette G, Hamner M, Katzelnick D, Keller T. A double-
36 Kosten TR. Alexithymia as a predictor of treatment response in post- blind randomized placebo-controlled multi-center study of brofaromine in the
traumatic stress disorder. J Trauma Stress 1992; 5: 563–73. treatment of posttraumatic stress disorder. Psychopharmacology 1995; 122:
37 Stein MB, Kline NA, Matloff JL. Adjunctive olanzapine for SSRI resistant 386–9.
combat-related PTSD: a double-blind, placebo-controlled study. Am J 61 Brady KT, Sonne S, Anton RF, Randall CL, Back SE, Simpson K. Sertraline in
Psychiatry 2002; 159: 1777–9. the treatment of co-occurring alcohol dependence and posttraumatic stress
38 Brady K, Pearlstein T, Asnis GM, Baker D, Rothbaum B, Sikes CR. Efficacy and disorder. Alcohol Clin Exp Res 2005; 29: 395–401.
safety of sertraline treatment of posttraumatic stress disorder. Arch Gen 62 Braun P, Greenberg D, Dasberg H, Lerer B. Core symptoms of posttraumatic
Psychiatry 2000; 283: 1837–44. stress disorder unimproved by alprazolam treatment. J Clin Psychiatry 1991;
39 Butterfield MI, Becker ME, Connor KM, Sutherland S, Churchill LE, Davidson 51: 236–8.
JR. Olanzapine in the treatment of post-traumatic stress disorder: a pilot 63 Hertzberg MA, Butterfield MI, Feldman ME, Beckham JC, Sutherland SM,
study. Int Clin Psychopharmacol 2001; 16: 197–203. Connor KM. A preliminary study of lamotrigine for the treatment of
40 Connor KM, Sutherland SM, Tupler LA, Malik ML, Davidson JR. Fluoxetine posttraumatic stress disorder. Biol Psychiatry 1999; 45: 1226–9.
in post-traumatic stress disorder: randomised, double-blind study. 64 Marshall RD, Lewis-Fernandez R, Blanco C, Simpson HB, Lin S-H, Vermes D,
Br J Psychiatry 1999; 175: 17–22. et al. A controlled trial of paroxetine for chronic PTSD, dissociation and
41 Davidson JR, Kudler H, Smith R, Mahorney SL, Lipper S, Hammett E. interpersonal problems in mostly minority adults. Depress Anxiety 2007;
Treatment of posttraumatic stress disorder with amitriptyline and placebo. 24: 77–84.
Arch Gen Psychiatry 1990; 47: 259–66.
65 McRae AL, Brady KT, Mellman TA, Sonn SC, Killeen TK, Timmerman MA.
42 Davidson J, Pearlstein T, Londborg P, Brady KT, Rothbaum B, Bell J. Efficacy Comparison of nefazodone and sertraline for the treatment of posttraumatic
of sertraline in preventing relapse of posttraumatic stress disorder: results of stress disorder. Depress Anxiety 2004; 19: 190–6.

99
Hoskins et al

66 Reist C, Kauffmann CD, Haier RJ, Sangdahl C, DeMet EM, Chicz-DeMet A. A 74 Stein MB, Kline NA, Matloff JL. Adjunctive olanzapine for SSRI resistant
controlled trial of desipramine in 18 men with posttraumatic stress disorder. combat-related PTSD: a double-blind, placebo-controlled study. Am J
Am J Psychiatry 1989; 146: 513–6. Psychiatry 2003; 159: 1777–9.

67 Saygin MZ, Sungur MZ, Sabol EU, Cetinkaya P. Nefazodone versus sertraline 75 Davidson JR, Connor KM, Hertzberg MA, Weisler RH, Wilson WH, Payne VM.
in treatment of posttraumatic stress disorder. Bull Clin Psychopharmacol Maintenance therapy with fluoxetine in posttraumatic stress disorder: a
2002; 12: 1–5. placebo-controlled discontinuation study. J Clin Psychopharmacol 2005; 25:
166–9.
68 Tucker P, Potter-Kimball R, Wyatt DB, Parker DE, Burgin C, Jones DE. Can
76 Leucht S, Hierl S, Kissling W, Dold M, Davis JM. Putting the efficacy of
physiologic assessment and side effects tease out differences in PTSD trials?
psychiatric and general medicine medication into perspective: review of
A double-blind comparison of citalopram, sertraline, and placebo. Gen
meta-analyses. Br J Psychiatry 2012; 200: 97–106.
Psychopharmacol 2003; 37: 135–49.
77 World Health Organization. Guidelines for the Management of Conditions
69 Van der Kolk BA, Dreyfuss D, Michaels M, Shera D, Berkowitz R, Fisler R.
Specifically Related to Stress. WHO, 2013.
Fluoxetine treatment in posttraumatic stress disorder. J Clin Psychiatry 1994;
55: 517–22. 78 Australian Centre for Posttraumatic Mental Health. Australian Guidelines for
the Treatment of Acute Stress Disorder and Posttraumatic Stress Disorder.
70 Van der Kolk BA, Spinazzola J, Blaustein ME, Hopper JW, Hopper EK, Korn DL, ACPMH, 2013.
et al. A randomized clinical trial of eye movement desensitization and
79 Jonas DE, Cusack K, Forneris CA, Wilkins TM, Sonis J, Middleton JC, et al.
reprocessing (EMDR), fluoxetine and pill placebo in the treatment of
Psychological and Pharmacological Treatments for Adults With Posttraumatic
posttraumatic stress disorder: treatment effects and long-term maintenance.
Stress Disorder (PTSD). US Agency for Healthcare Research and Quality
J Clin Psychiatry 2007; 68: 37–46.
Comparative Effectiveness Review 92, 2013.
71 Mooney P, Oakley J, Ferriter M, Travers R. Sertraline as a treatment for PTSD:
80 Watts BV, Schnurr PP, Mayo L, Young-Xu, Weeks WB, Friedman M. Meta-
a systematic review and meta-analysis. Irish J Psychol Med 2004; 21: 100–3.
analysis of the efficacy of treatments for posttraumatic stress disorder. J Clin
72 Otto MW, Hinton D, Korbly NB, Chea A, Ba P, Gershuny BS, et al. Psychiatry 2013; 74: e541–50.
Treatment of pharmacotherapy-refractory posttraumatic stress disorder 81 Furukawa TA, Watanabe N, Omori IM, Montori VM, Guyatt GH. Association
among Cambodian refugees: a pilot study of combination treatment between unreported outcomes and effect size estimates in Cochrane
with cognitive–behavior therapy vs sertraline alone. Behav Res Ther 2003; meta-analyses. JAMA 2007; 297: 468–70.
41: 1271–6.
82 Davidson JRT, Connor KM, Hertzberg MA, Weisler RH, Wilson WH, Payne VM.
73 Rothbaum BO, Cahill SP, Foa EB, Davidson JR, Compton J, Connor KM, et al. Maintenance therapy with fluoxetine in posttraumatic stress disorder: a
Augmentation of sertraline with prolonged exposure in the treatment of placebo-controlled discontinuation study. J Clin Psychopharmacol 2005; 25:
posttraumatic stress disorder. J Trauma Stress 2006; 19: 625–38. 166–9

100
Pharmacotherapy for post-traumatic stress disorder: systematic
review and meta-analysis
Mathew Hoskins, Jennifer Pearce, Andrew Bethell, Liliya Dankova, Corrado Barbui, Wietse A. Tol, Mark
van Ommeren, Joop de Jong, Soraya Seedat, Hanhui Chen and Jonathan I. Bisson
BJP 2015, 206:93-100.
Access the most recent version at DOI: 10.1192/bjp.bp.114.148551

Supplementary Supplementary material can be found at:

Material http://bjp.rcpsych.org/content/suppl/2015/02/02/206.2.93.DC1.html

References This article cites 68 articles, 4 of which you can access for free at:
http://bjp.rcpsych.org/content/206/2/93#BIBL
Reprints/ To obtain reprints or permission to reproduce material from this paper, please
permissions write to [email protected]

You can respond http://bjp.rcpsych.org/letters/submit/bjprcpsych;206/2/93

to this article at
Downloaded http://bjp.rcpsych.org/ on February 3, 2015
from Published by The Royal College of Psychiatrists

To subscribe to The British Journal of Psychiatry go to:

http://bjp.rcpsych.org/site/subscriptions/