RESEARCH REPORT

Antipsychotic augmentation vs. monotherapy in schizophrenia:

systematic review, meta-analysis and meta-regression analysis
Britta Galling1-3, Alexandra Roldan4, Katsuhiko Hagi2,5, Liz Rietschel6, Frozan Walyzada2, Wei Zheng7, Xiao-Lan Cao8, Yu-Tao Xiang9,
Mathias Zink10, John M. Kane2,3,11,12, Jimmi Nielsen13,14, Stefan Leucht15, Christoph U. Correll2,3,11,12
1
Department of Child and Adolescent Psychiatry, Psychosomatic Medicine and Psychotherapy, Charite-Universit€atsmedizin Berlin, Berlin, Germany; 2Zucker Hillside Hospital,
Psychiatry Research, Northwell Health, Glen Oaks, NY, USA; 3Hofstra Northwell School of Medicine, Hempstead, NY, USA; 4Department of Psychiatry, Institut d’Investigaci o
onoma de Barcelona, Barcelona, Spain; 5Sumitomo Dainippon Pharma Co., Tokyo, Japan; 6University
Biomèdica Sant Pau, Hospital de la Santa Creu i Sant Pau, Universitat Aut
Hospital of Child and Adolescent Psychiatry and Psychotherapy, University of Bern, Bern, Switzerland; 7Guangzhou Brain Hospital, Affiliated Hospital of Guangzhou Medical
University, Guangzhou, China; 8Department of Psychiatry, Chinese University of Hong Kong, Hong Kong SAR, China; 9Unit of Psychiatry, Faculty of Health Sciences, University
of Macao, Taipa, Macao, SAR, China; 10Central Institute of Mental Health, Medical Faculty Mannheim, Heidelberg University, Mannheim, Germany; 11Feinstein Institute for Medi-
cal Research, Manhasset, NY, USA; 12Albert Einstein College of Medicine, Bronx, NY, USA; 13Department of Psychiatry, Aalborg University Hospital, Aalborg, Denmark;
14
Department of Clinical Medicine, Aalborg University, Aalborg, Denmark; 15Klinik und Poliklinik f€ur Psychiatrie und Psychotherapie, Technische Universit€at M€unchen, Munich,
Germany

Antipsychotic polypharmacy in schizophrenia is much debated, since it is common and costly with unclear evidence for its efficacy and safe-
ty. We conducted a systematic literature search and a random effects meta-analysis of randomized trials comparing augmentation with a
second antipsychotic vs. continued antipsychotic monotherapy in schizophrenia. Co-primary outcomes were total symptom reduction and
study-defined response. Antipsychotic augmentation was superior to monotherapy regarding total symptom reduction (16 studies, N5694,
standardized mean difference, SMD5–0.53, 95% CI: 20.87 to 20.19, p50.002). However, superiority was only apparent in open-label and
low-quality trials (both p<0.001), but not in double-blind and high-quality ones (p50.120 and 0.226, respectively). Study-defined response
was similar between antipsychotic augmentation and monotherapy (14 studies, N5938, risk ratio 5 1.19, 95% CI: 0.99 to 1.42, p50.061),
being clearly non-significant in double-blind and high-quality studies (both p50.990). Findings were replicated in clozapine and non-
clozapine augmentation studies. No differences emerged regarding all-cause/specific-cause discontinuation, global clinical impression, as
well as positive, general and depressive symptoms. Negative symptoms improved more with augmentation treatment (18 studies, N5931,
SMD5–0.38, 95% CI: 20.63 to 20.13, p<0.003), but only in studies augmenting with aripiprazole (8 studies, N5532, SMD5–0.41, 95% CI:
20.79 to 20.03, p50.036). Few adverse effect differences emerged: D2 antagonist augmentation was associated with less insomnia
(p50.028), but more prolactin elevation (p50.015), while aripiprazole augmentation was associated with reduced prolactin levels
(p<0.001) and body weight (p50.030). These data suggest that the common practice of antipsychotic augmentation in schizophrenia lacks
double-blind/high-quality evidence for efficacy, except for negative symptom reduction with aripiprazole augmentation.

Key words: Antipsychotics, polypharmacy, augmentation, monotherapy, schizophrenia, clozapine, aripiprazole

(World Psychiatry 2017;16:77–89)

Management options for patients with schizophrenia remain or rejection6-10. Additionally, concerns about antipsychotic poly-
suboptimal, as indicated by insufficient symptom control in a pharmacy include the potential for drug-drug interactions, de-
sizable subgroup of patients and low response rates, frequently creased adherence due to complex drug regimes, higher cost23-25,
leading to functional impairment1-5. Recommendations after in- and increased adverse effects10,22,26-29.
adequate antipsychotic response include waiting for a delayed Meta-analyses aggregate the information of conceptually sim-
response, dose adjustment, switching to another antipsychotic, ilar studies and consolidate their quantitative outcomes using
and – in case of treatment resistance to at least two adequate statistics. The derived pooled estimates of treatment efficacy and
antipsychotic trials – clozapine treatment6-11. safety are more robust compared to primary study results. More-
Another adopted strategy is antipsychotic polypharmacy12. over, meta-analyses enable researchers to contrast results from
Limited data on clinicians’ reasoning suggest various motivations multiple studies and to identify patterns of common effects
for this strategy, including attempts to increase/speed up efficacy, across studies, or reasons for outcome variability. However, to
treat residual positive symptoms, or reduce adverse effects allow- facilitate informative results and meaningful subgroup and
ing dose reduction of the first antipsychotic13. Antipsychotic poly- meta-regression analyses, the study methodology should be as
pharmacy has been reported as a common clinical practice12,14,15, homogeneous as possible; study quality should be taken into
sometimes implemented by clinicians before or instead of trying account; and the total population studied should be sufficiently
clozapine13,16. Although the frequency of antipsychotic polyphar- large (1000 subjects)30.
macy varies according to patient, illness, setting and provider Although four meta-analyses examined the efficacy of anti-
variables17, rates in schizophrenia commonly range between 10 psychotic polypharmacy, either irrespective of the antipsy-
and 30%12,14,17-19. chotics used20 or restricted to clozapine-treated patients21,31,32,
Despite common use, the evidence for the efficacy and tolera- their results remained somewhat inconclusive, possibly influ-
bility of antipsychotic polypharmacy is weak20-22. In fact, guide- enced by: a) mixing together antipsychotic augmentation (add-
lines reserve augmentation with a second antipsychotic as a ing a second antipsychotic after non-response to the first)
last-stage treatment option after clozapine failure, intolerability and co-initiation (combination of two antipsychotics from the

World Psychiatry 16:1 - February 2017 77

beginning) strategies20; b) lack of separating lower from higher continuation of existing antipsychotic monotherapy; and in which
quality studies20, and c) the relatively low number of available meta-analyzable data were reported, including symptomatic/
studies and patients treated in an augmentation paradigm20,21,31,32. functional or adverse effect outcomes. We excluded studies com-
In one of those meta-analyses, polypharmacy was associated paring antipsychotic monotherapy versus two antipsychotics
with significantly greater response than monotherapy, with a started concurrently, as well as those comparing antipsychotic
number-needed-to-treat of 720. However, the improved response augmentation with antipsychotic switch instead of continuation
was moderated by studies lasting at least ten weeks, conducted of the original antipsychotic monotherapy.
in China, examining co-initiation and involving clozapine. Fur- Co-primary outcomes were total symptom reduction, as
ther, that meta-analysis only included six studies of antipsychotic assessed by the Positive and Negative Syndrome Scale (PANSS)37
augmentation (N5197), and did not assess symptom reduction or the Brief Psychiatric Rating Scale (BPRS)38, and study-defined
due to lacking data. treatment response. Secondary outcomes were all-cause and
The three remaining meta-analyses focused on combina- specific-cause discontinuation (inefficacy, intolerability); reduc-
tion treatments with clozapine, either mixing co-initiation and tion of positive symptoms (as assessed by the PANSS positive,
augmentation studies together21, or focusing on augmentation the BPRS positive, or the Scale for the Assessment of Positive
studies but analyzing only individual drug combinations32, or Symptoms, SAPS39), of negative symptoms (as assessed by the
focusing only on symptom reduction and not response rates31. PANSS negative, the BPRS negative, or the Scale for the Assess-
One meta-analysis found clozapine co-treatment to be superior ment of Negative Symptoms, SANS40), and of general symptoms
to clozapine monotherapy, but this finding was only apparent (as assessed by the PANSS general); reduction of global illness
in open-label studies21. In one other meta-analysis, augmenta- severity (as assessed by the Clinical Global Impression Scale -
tion of clozapine with a second antipsychotic was associated Improvement, CGI-I41); reduction of depressive symptoms (as
with a small benefit (effect size 5 0.239, p50.028), but only 14 assessed by the PANSS/BPRS anxiety/depression, the Hamilton
trials and 714 patients provided data, and higher versus lower Scale for Depression, HAM-D42, or the Calgary Depression
quality studies were not analyzed separately31. Scale for Schizophrenia, CDSS43); improvement of functioning
Due to the limitations of those prior meta-analyses, the fre- (as evaluated by the Global Assessment of Functioning Scale,
quent use of antipsychotic polypharmacy in ordinary practice, GAF44); and frequency and severity of adverse effects.
and the recent publication of many additional studies, we con- Data of each study were independently identified, checked
ducted a new systematic review and meta-analysis comparing and extracted by at least two authors, including information
the efficacy and adverse effects of antipsychotic augmentation relevant for the Cochrane risk-of-bias tool45. Inconsistencies
vs. monotherapy. Based on the prior literature20,21,31-34, we were resolved by consensus/involvement of a third reviewer.
hypothesized that antipsychotic augmentation would not be We conducted a random effects46 meta-analysis of outcomes
superior to monotherapy regarding efficacy (measured as total using Comprehensive Meta-Analysis V3 (www.meta-analysis.
and specific symptom reduction as well as response/remission/ com). Study heterogeneity was explored using I2 statistics and
relapse) when focusing on augmentation trials and those with chi-square test of homogeneity, with I2>50% and p<0.05 indi-
higher quality, but that antipsychotic augmentation might con- cating significant heterogeneity. All analyses were two-tailed
fer a higher risk of adverse effects (except for reduction of spe- with alpha50.05, without adjustments for multiple comparisons.
cific adverse effects when adding a partial D2 agonist to D2 For “total” and “specific” psychopathology (except depres-
antagonists). sion and negative symptoms) and for inefficacy-related discon-
tinuation, all studies except those focusing on the amelioration
of adverse effects were analyzed. The reason for using this
METHODS restricted data set was that studies focusing on the ameliora-
tion of adverse effects could have included treatment respond-
The systematic review was conducted in accordance with the ers, leaving little or no room for improvement. In contrast, for
Preferred Reporting Items for Systematic Reviews and Meta- depression and negative symptoms and for individual adverse
Analyses (PRISMA) standard35,36. At least two independent effects, all-cause discontinuation and intolerability-related dis-
authors searched PubMed/MEDLINE, PsycINFO, Chinese Jour- continuation, all available data were analyzed, including stud-
nal Net, Wangfan, and China Biology Medicine databases from ies focusing on the reduction of adverse effects.
inception until May 25, 2015, without language restrictions, Group differences in continuous outcomes were analyzed as
supplemented by a manual review of reference lists from eligi- the pooled standardized mean difference (SMD) in either change
ble publications and relevant reviews. Authors were contacted from baseline to endpoint (preferred) or endpoint scores (only
for additional information if needed. preferred if change score results were skewed, i.e., SD > twice the
We included randomized controlled trials with samples con- mean). Additionally, weighted mean difference (WMD) was cal-
sisting of at least 20 adults with a diagnosis of schizophrenia or culated for weight change in kilograms. Dichotomous data were
schizoaffective disorder; in which patients were assigned to aug- analyzed calculating the pooled risk ratio (RR). Intention-to-treat
mentation of the current antipsychotic with a different antipsy- (ITT) data were always preferred, but observed cases (OC) data
chotic versus augmentation with placebo (in blinded studies) or were also allowed.

78 World Psychiatry 16:1 - February 2017

Figure 1 PRISMA diagram of the literature search

All outcomes were analyzed for the pooled sample and for yielding 31 studies that were included in the meta-analysis
high-quality studies separately. The latter were defined a priori (Figure 1).
as double-blind studies using ITT/last-observation-carried-
forward (LOCF) analyses, as opposed to open-label studies
Efficacy of antipsychotic monotherapy vs. augmentation
and those using OC data. In two studies with more than one
(efficacy data set)
active augmentation arm47,48, the number of patients in the
monotherapy group was divided by the number of active study
Details on the 22 meta-analyzed studies with efficacy as the
arms to avoid double-counting of control subjects.
primary outcome (N51,342) are provided in Table 1. They
For meta-regression analyses, the baseline BPRS total scores
included 13 double-blind and ITT/LOCF “high-quality” stud-
were converted to PANSS total scores using equipercentile link-
ies and 9 open-label and/or OC “low quality” ones.
ing49. Exploratory subgroup and meta-regression analyses were
Antipsychotic augmentation was superior to monotherapy
added post-hoc for negative symptom change (the only overall
regarding total symptom reduction (16 studies, N5694, SMD
significant outcome in both low- and high-quality studies) in
5–0.53, 95% CI: 20.87 to 20.19, p50.002), but only in open-
studies using partial D2 agonists.
label (n56, N5285, SMD5–0.81, 95% CI: 21.18 to 20.43,
We inspected funnel plots, used Egger’s regression test50
p<0.001) and low-quality (n57, N5316, SMD5–0.83, 95% CI:
and the Duval and Tweedie’s trim and fill method51 to quantify
21.16 to 20.50, p<0.001) studies, not in double-blind (n510,
whether publication bias could have influenced the results.
N5409, SMD5–0.37, 95% CI: 20.83 to 0.10, p50.120) and
high-quality (n59, N5378, SMD5–0.30, 95% CI: 20.78 to 0.19,
p50.226) ones (Figures 2 and 3). The funnel plots and Egger’s
RESULTS test did not indicate publication bias (p50.320).
In subgroup analyses, antipsychotic augmentation was superi-
The initial search resulted in 17,653 hits. Altogether, 17,427 or in certain settings (only inpatients: n56, N5316, SMD5–0.82,
studies were excluded at the title/abstract level. Of the remain- 95% CI: 21.22 to 20.43, p<0.001; only outpatients: n55, N5247,
ing 226 references, 195 were excluded after full text review, SMD5–0.76, 95% CI: 21.49 to 20.03, p50.042) and regions

World Psychiatry 16:1 - February 2017 79

 Table 1 Study, patient and treatment characteristics

80
Trial
Risk of Primary duration Monotherapy dose, Augmentation group
Study Agents No. patients bias* Blinding outcome Analysis (weeks) Setting mg/d: mean (range) dose, mg/d: mean (range)

Clozapine 1 first-generation antipsychotic

Liu et al52 (China) CLZ 1 FLU T: 60 1 OL Efficacy ITT 24 Inpatients CLZ: CLZ: FLU:
M: 30 NR (375-500) NR (375-500) NR (25-50)
A: 30

Friedman et al53 (US) CLZ 1 PIM T: 53 3 DB Efficacy ITT 12 Inpatients CLZ: CLZ: PIM:
M: 28 (64.2%) and 478.1 (NR) 518.8 (NR) 6.48 (2.0-8.9)
A: 25 outpatients
(35.8%)

Gunduz-Bruce CLZ 1 PIM T: 28 4 DB Efficacy ITT 12 Outpatients CLZ: CLZ: PIM:

et al54 (US) M: 14 NR (NR) NR (NR) 4 (fixed)
A: 14

Clozapine 1 second-generation antipsychotic

Chang et al55 (Korea) CLZ 1 ARI T: 61 5 DB Efficacy ITT 8 Inpatients and CLZ: CLZ: ARI:
M: 32 outpatients 290.6 (NR) 304.3 (NR) 15.5 (5-30)
A: 29 (% NR)

Fan et al56 (US) CLZ 1 ARI T: 38 2 DB Adverse OC 8 Outpatients CLZ: CLZ: ARI:
M: 18 effects 400 (NR) 397 (NR) 15 (fixed)
A: 20

Fleischhacker et al57 CLZ 1 ARI T: 207 4 DB Adverse ITT 16 Outpatients CLZ: CLZ: ARI:
(Europe, M: 99 effects 363 (163-900) 384 (200-900) 11.1 (5-15)
South Africa) A: 108

Guan58 (China) CLZ 1 ARI T: 60 1 OL Efficacy ITT 16 Inpatients CLZ: CLZ: ARI:
M: 30 NR (300-500) NR (200-300) NR (20-30)
A: 30

Muscatello CLZ 1 ARI T: 40 5 DB Efficacy OC 24 Outpatients CLZ: CLZ: ARI:

et al59 (Italy) M: 20 341.2 (200-450) 310.7 (200-450) 12.5 (10-15)
A: 20

Sun et al60 (China) CLZ 1 ARI T: 62 1 OL Efficacy ITT 6 Inpatients CLZ: CLZ: ARI:
M: 30 368.2 (200-450) 168 (75-300) 21.6 (10-30)
A: 32

Lin et al61 (China) CLZ 1 PAL T: 70 3 DB Efficacy ITT 12 Inpatients CLZ: CLZ: PAL:
M: 35 217.9 (NR) 231.7 (NR) 8.2 (6-12)
A: 35

Freudenreich CLZ 1 RIS T: 24 2 DB Efficacy ITT 6 Outpatients CLZ: CLZ: RIS:

et al62 (US) M: 11 456 (200-700) 456 (200-700) 4 (NR)
A: 13

Anil Yagcioglu CLZ 1 RIS T: 30 6 DB Efficacy ITT 6 Inpatients CLZ: CLZ: RIS:
et al63 (Turkey) M: 14 (20.0%) and 414.3 (300-900) 515.6 (300-900) 5.1 (NR)
A: 16 outpatients
(80.0%)

World Psychiatry 16:1 - February 2017

 Table 1 Study, patient and treatment characteristics (continued)
Trial
Risk of Primary duration Monotherapy dose, Augmentation group
Study Agents No. patients bias* Blinding outcome Analysis (weeks) Setting mg/d: mean (range) dose, mg/d: mean (range)

Honer et al64 CLZ 1 RIS T: 68 6 DB Efficacy ITT 8 Inpatients CLZ: CLZ: RIS:
(International) M: 34 (38.2%) and 487 (NR) 494 (NR) 3 (NR)
A: 34 outpatients
(61.8%)

World Psychiatry 16:1 - February 2017

Josiassen et al65 (US) CLZ 1 RIS T: 40 3 DB Efficacy ITT 12 Inpatients CLZ: CLZ: RIS:
M: 20 (26.1%) and 403 (NR) 529 (NR) 4.4 (NR)
A: 20 outpatients
(73.9%)

Hu66 (China) CLZ 1 RIS T: 60 1 OL Efficacy ITT 12 Inpatients CLZ: CLZ: RIS:
M: 30 253.6 (NR) 126.3 (NR) 2.9 (2-6)
A: 30

Weiner et al67 (US) CLZ 1 RIS T: 69 2 DB Efficacy ITT 16 Inpatients CLZ: CLZ: RIS:
M: 36 (26.1%) and NR (NR) NR (NR) 4 (fixed)
A: 33 outpatients
(73.9%)

Nielsen et al68 CLZ 1 SER T: 50 6 DB Efficacy ITT 12 Outpatients CLZ: CLZ: SER:
(Denmark) M: 25 435 (NR) 394 (NR) 16 (fixed)
A: 25

Shiloh et al69 (Israel) CLZ 1 SUL T: 28 5 DB Efficacy ITT 10 Inpatients CLZ: CLZ: SUL:
M: 12 446 (NR) 403 (NR) NR (100-600)
A: 16

Jiang et al70 (China) CLZ 1 ZIP T: 24 2 OL Efficacy ITT 12 NR CLZ: CLZ: ZIP:
M: 12 597.2 (75-600) 489.7 (75-600) NR (20-160)
A: 12

Muscatello et al71 CLZ 1 ZIP T: 40 6 DB Efficacy ITT 16 Outpatients CLZ: CLZ: ZIP:
(Italy) M: 20 462.5 (350-600) 428.7 (350-600) 80.0 (fixed)
A: 20

First-generation 1 second-generation antipsychotic

Shim et al72 HAL 1 ARI T: 54 2 DB Adverse ITT 8 NR HAL: HAL: ARI:

(US, Korea) M: 28 effects 24.8 (NR) 20.7 (NR) 22.5 (15-30)
A: 26

Second-generation 1 second-generation antipsychotic

Chen et al47 (China) RIS 1 ARI T: 119 3 DB Adverse ITT 8 Inpatients and RIS: RIS: ARI:
M: 30 effects outpatients 4.93 (NR) 4.63 (NR) 5 (fixed)
A: 89 (% NR)
RIS: ARI:
4.79 (NR) 10 (fixed)

81
82
Table 1 Study, patient and treatment characteristics (continued)
Trial
Risk of Primary duration Monotherapy dose, Augmentation group
Study Agents No. patients bias* Blinding outcome Analysis (weeks) Setting mg/d: mean (range) dose, mg/d: mean (range)

RIS: ARI:
5.07 (NR) 20 (fixed)

Kane et al73 (US) QTP/RIS 1 ARI T: 323 1 DB Efficacy ITT 16 Outpatients QTP/RIS: QTP/RIS: ARI:
M: 155 516/4.8 513/4.6 10.3 (2-15)
A: 158 (400-800/4-8) (400-800/4-8)

Lee et al74 (Korea) RIS 1 ARI T: 35 2 DB Adverse ITT 12 Inpatients RIS: 3 RIS: ARI:
M: 18 effects (NR) 3 (NR) 10 (fixed)
A: 17
Liu et al48 (China) RIS 1 ARI T: 86 1 DB Adverse ITT 4 Inpatients RIS: RIS: ARI:
M: 27 effects NR (>4) NR (>4) 5 (fixed)
A: 59
ARI:
10 (fixed)

Ou et al75 (China) OLZ 1 ARI T: 70 1 OL Efficacy OC 8 Inpatients OLZ: OLZ: ARI:

M: 35 18.2 (NR) 17.8 (NR) 10 (fixed)
A: 35

Yasui-Furukori et al76 RIS/OLZ 1 ARI T: 36 1 DB Adverse OC 12 Outpatients RIS/OLZ: RIS/OLZ: ARI:

(Japan) M: 18 effects 5.0/12.5 (3-8/5-20) 5.9/12.1 15.2 (6-30)
A: 18 (2-12/2.5-20

Zhao77 (China) RIS 1 ARI T: 56 1 OL Adverse NR 12 Inpatients and RIS: RIS: ARI:
M: 28 effects outpatients NR (3-8) NR (3-8) 10 (fixed)
A: 28 (% NR)

Zhou et al78 (China) RIS 1 ARI T: 100 0 OL Adverse NR 24 Inpatients RIS: RIS: ARI:
M: 50 effects NR (4-6) NR (4-6) 5 (fixed)
A: 50

Liang & Liu79 (China) ARI 1 CLZ T: 65 1 OL Efficacy ITT 8 NR ARI: ARI: CLZ:
M: 33 NR (20-30) NR (20-30) NR (25-100)
A: 32

Kotler et al80 (Israel) OLZ 1 SUL T: 17 2 OL Efficacy ITT 8 Inpatients OLZ: OLZ: SUL:
M: 8 22.5 (20-30) 22.2 (20-30) 600 (fixed)
A: 9

*number of low risk judgements, T – total, M – monotherapy, A – augmentation, OL – open label, DB – double blind, ITT – intent to treat, OC – observed cases, CLZ – clozapine, FLU – fluphenazine, PIM –
pimozide, ARI – aripiprazole, PAL – paliperidone, RIS – risperidone, SER – sertindole, SUL – sulpiride, ZIP – ziprasidone, HAL – haloperidol, QTP – quetiapine, OLZ – olanzapine, NR – not reported

World Psychiatry 16:1 - February 2017

World Psychiatry 16:1 - February 2017
Figure 2 Forest plots of overall symptom reduction and study-defined response. SMD – standardized mean difference, RR – risk ratio

83
84
Figure 3 Primary outcomes, subgroup analyses and meta-regression in studies with efficacy as primary outcome. SMD – standardized mean difference, RR – risk ratio, CLZ – cloza-
pine, SGA – second generation antipsychotic, FGA – first generation antipsychotic, AP – antipsychotic

World Psychiatry 16:1 - February 2017

(China: n55, N5269, SMD520.86, 95% CI: 21.27 to 20.45, 0.16, p50.230), general symptoms (n54, N5144, SMD5–0.73,
p<0.001; non-North American/European countries: n58, N5374, 95% CI: 21.91 to 0.46, p50.229), and functioning (n52, N580,
SMD5–0.71, 95% CI: 21.05 to 20.37, p<0.001). However, superi- SMD5–0.36, 95% CI: 21.19 to 0.47, p50.389).
ority in these subgroups was not apparent in high-quality stud-
ies (Figure 3).
Efficacy and tolerability of antipsychotic monotherapy
Findings regarding symptom reduction were replicated in aug-
vs. augmentation (complete data set)
mentation studies of clozapine with a second generation antipsy-
chotic (SGA) or a first generation antipsychotic (FGA) (n514,
The complete data set (efficacy-focused plus adverse effect-
N5612, SMD5–0.52, 95% CI: 20.90 to 20.14, p50.007), clozapine
focused studies) included 31 trials (N52,073) (see Table 1).
with a SGA (n512, N5528, SMD5–0.52, 95% CI: 20.93 to 20.11,
The mean PANSS/converted BPRS score was higher in ef-
p50.012), and non-clozapine SGA with a SGA (n52, N582,
ficacy-focused studies (total sample 5 79.7 6 10.8, clozapine
SMD5–0.71, 95% CI: 21.16 to 20.26, p50.002); studies augment-
studies 5 79.3 6 9.6, non-clozapine studies 5 81.7 6 15.9) than
ing with a partial D2 agonist (n54, N5214, SMD5–0.57, 95% CI:
in adverse-effect focused ones (total sample 5 67.4 6 9.2, clo-
21.10 to 20.03, p50.039), and those augmenting with D2 antago-
zapine studies 5 71.5, non-clozapine studies 5 66.6 6 9.9).
nists (n512, N5480, SMD5–0.52, 95% CI: 20.95 to 20.08,
All-cause discontinuation (n522, N51,482, RR51.13, 95% CI:
p50.019). Results persisted independent of the non-response def-
0.90 to 1.42, p50.284), and intolerability-related discontinuation
inition (strict, 2 adequate trial failures vs. lenient, 1 adequate
(n511, N5949, RR50.87, 95% CI: 0.50 to 1.50, p50.611) did not
trial failure): respectively, n513, N5542, SMD5–0.41, 95% CI:
differ between antipsychotic augmentation and monotherapy.
20.79 to 0.03, p50.035; and n52, N586, SMD5–1.15, 95% CI:
Negative symptoms improved with antipsychotic augmenta-
21.61 to 20.70, p<0.001). However, again, differences were non-
tion (n518, N5931, SMD5–0.38, 95% CI: 20.63 to 20.13,
significant when analyzing only high-quality studies (Figure 3).
p50.003), but in subgroup analyses this effect was only signifi-
In meta-regression analyses, a higher augmentation-to-
cant in studies augmenting D2 antagonists with a partial D2 ago-
monotherapy ratio of chlorpromazine equivalent dose (p50.019)
nist (n58, N5532, SMD5–0.41, 95% CI: 20.79 to 20.03,
and higher baseline PANSS/converted BPRS scores (p50.011)
p50.036), not when combining two D2 antagonists (n510,
were associated with less symptom improvement, while studies
N5399, SMD5–0.36, 95% CI: 20.72 to 0.01, p50.055). These
with high risk of bias near-significantly moderated greater im-
findings were replicated in high-quality studies (4 trials augment-
provement with antipsychotic augmentation (p50.050). The influ-
ing D2 antagonists with a partial D2 agonist, N5355, SMD5–
ence of the PANSS/converted BPRS scores was replicated in
0.28, 95% CI: 20.55 to 20.009, p50.043). In exploratory subgroup
high-quality studies (p50.033), whereas the other factors were
and meta-regression analyses, no relevant moderator of negative
non-significant.
symptom improvement with a partial D2 agonist emerged.
Response, as defined by 20% PANSS/BPRS reduction
Antipsychotic augmentation and monotherapy did not dif-
(n510), 25% PANSS reduction (n53), and 20% PANSS re-
fer regarding depressive symptoms (n510, N5351, SMD5
duction or CGI-I of 1 or 2 (n51), did not differ between anti-
20.69, 95% CI: 21.42 to 0.05, p50.066).
psychotic augmentation and monotherapy (n514, N5938,
Few differences in adverse effects emerged. D2 antagonist
RR51.19, 95% CI: 0.99 to 1.42, p50.061). In subgroup analyses,
augmentation was associated with less insomnia (n53, N5169,
antipsychotic augmentation was superior in open-label/
RR50.26, 95% CI: 0.08 to 0.86, p50.028), but more prolactin
low-quality studies (n54, N5245, RR51.30, 95% CI: 1.04 to 1.64,
elevation (n52, each representing augmentation with risperi-
p50.024), but not in double-blind/high-quality ones (n510,
done, N574, SMD52.20, 95% CI: 0.43 to 3.96, p50.015), while
N5693, RR51.00, 95% CI: 0.72 to 1.39, p50.990) (Figure 2). The
aripiprazole augmentation of D2 antagonists was associated
funnel plots and Egger’s test did not indicate publication bias
with reduced prolactin levels (n59, N5450, SMD5–1.60, 95%
(p50.508).
CI: 22.19 to 21.01, p<0.001) and body weight (n56, N5260,
Antipsychotic augmentation was again superior in inpatient
WMD5–0.93, 95% CI5 21.77 to 20.09, p50.030).
only studies (n54, N5207, RR51.67, 95% CI: 1.00 to 2.77,
p50.049), Chinese studies (n54, N5245, RR51.30, 95% CI: 1.04
to 1.64, p50.024) and non-North American/European studies
(n55, N5273, RR51.35, 95% CI: 1.06 to 1.73, p50.017) (Figure 3). DISCUSSION
In these subgroups, the number of high-quality studies was 1,
not allowing for separate analyses. There was no advantage of While some prior meta-analyses have examined the efficacy
any specific antipsychotic combination, or depending on non- of combination or “polypharmacy” strategies in schizophre-
response definition. No significant moderator of treatment nia20,21,31,32, this is the first meta-analysis of randomized con-
response emerged. No between-group differences were observed trolled trials focusing exclusively on augmentation strategies
regarding inefficacy-related discontinuation (n56, N5596, (i.e., the addition of a second antipsychotic after non-response
RR51.08, 95% CI: 0.44 to 2.67, p50.870), global clinical impres- to the first) versus continued treatment with one antipsychotic
sion (n58, N5403, SMD5–0.01, 95% CI: 20.32 to 0.30, p50.947), (with addition of placebo in the blinded studies), irrespective of
positive symptoms (n514, N5604, SMD5–0.25, 95% CI: 20.66 to the baseline antipsychotic.

World Psychiatry 16:1 - February 2017 85

 Although our prior meta-analysis, published in 200920, in- challenge in schizophrenia81,82, these findings, based on eight
cluded 19 studies and 1,229 patients, merely 6 studies with studies (including four high-quality trials) clearly require fur-
only 197 patients were “augmentation” studies, which are the ther investigation, especially comparing augmentation with a
ones that are truly relevant, as they clinically reflect the man- partial D2 agonist versus switching to a partial D2 agonist.
agement of refractory/non-responsive patients. In the current Since two new partial D2 agonists, brexpiprazole83 and cari-
study, we increased the meta-analyzed data from 6 to 31 stud- prazine84, were recently approved for schizophrenia, it will be
ies and 197 to 2,073 patients. This greater number of studies of interest to see if the potential benefits for negative symp-
allowed for an evaluation of various symptom domains beyond toms extend to these other agents.
study-defined response, plus the assessment of adverse effects Different from the generally held notion that antipsychotic
and subgroup and meta-regression analyses, including exami- polypharmacy carries a greater risk of adverse effects22, this
nation of the effect of open versus blinded trials. was only found regarding greater prolactin elevation when
In contrast to that prior meta-analysis20, in which response combining two D2 antagonists. Rather, combination of two D2
rates had been significantly greater in the antipsychotic poly- antagonists was associated with less insomnia, whereas aug-
pharmacy group that mixed co-initiation and augmentation mentation with the partial D2 agonist aripiprazole resulted in
trials (number-needed-to-treat 5 7), the current meta-analysis lower prolactin levels and reduced body weight.
did not provide any evidence for enhanced efficacy of antipsy- The lack of superior efficacy of antipsychotic augmentation
chotic augmentation in high-quality, blinded studies for either in high-quality studies is in contrast to common clinical belief
antipsychotic response or symptom reduction. This finding and practice, where antipsychotic co-treatment is often imple-
suggests that expectation and salience biases, also present in mented for non-response to antipsychotic monotherapy12.
clinical care, may underlie observed improvements and deci- However, the clinical evaluation of improvement with antipsy-
sion making when augmenting one antipsychotic with a sec- chotic augmentation mirrors the findings from open-label
ond one. studies, suggesting that in clinical settings the expectations of
Although in efficacy-focused studies total symptoms decreased patients and clinicians may translate into perceived favorable
significantly more in the augmentation group, this effect was outcomes. Large pragmatic randomized controlled trials of
driven by open-label studies and those using OC analyses. antipsychotic augmentation strategies conducted in generaliz-
Notably, the non-significance regarding total symptom reduc- able settings and samples are needed to confirm the lack of
tion in high-quality studies was not driven by fewer studies and efficacy advantages of antipsychotic augmentation, as we can-
widening of the confidence intervals. Rather, more high-quality not fully rule out a selection bias of less severely ill patients
than low-quality studies were included (nine vs. seven), and the agreeing to participate in blinded trials. However, this possibil-
confidence intervals remained almost identical, whereas the ity seems relatively low, since mean baseline PANSS/converted
between-group effect size was much smaller in high-quality BPRS total symptom severity was around 80 in these pre-
studies. Furthermore, in efficacy-focused studies, no difference treated individuals, and PANSS/converted BPRS total symp-
between antipsychotic augmentation and monotherapy was tom severity did not significantly moderate the results.
found regarding response rate, but, again, in the subgroup of In meta-regression analyses, less symptom improvement was
low-quality studies superiority of the augmentation arm was associated with a higher chlorpromazine equivalent dose in the
observed. augmentation versus monotherapy arms, and a greater baseline
Evidence regarding symptom improvement and treatment symptom severity, with the latter relationship remaining signifi-
response was lacking for augmentation of either clozapine or cant in high-quality studies. These findings suggest that antipsy-
non-clozapine antipsychotics (with the latter studies being sur- chotic augmentation is even less effective in the sicker patients
prisingly uncommon). The previously identified benefit regard- and those requiring higher antipsychotic doses. Alternatively, the
ing augmentation of clozapine with a second antipsychotic31 higher total antipsychotic doses in the combination groups may
could not be confirmed in blinded trials and those using be a reflection of lack of initial improvement, prompting dose
ITT data. Prior meta-analyses that focused on antipsychotic escalation. This relationship might also be due to greater dopa-
co-treatment strategies involving clozapine21,31,32 had much mine blockade resulting in less improvement due to secondary
fewer studies (augmentation studies: 5-14 vs. 20 in our meta- negative symptoms or other unfavorable effects.
analysis; patients: 187-734 vs. 1,112 in our meta-analysis) and Although the moderation of less efficacy by higher baseline
in one instance21 combined antipsychotic augmentation and symptom severity contradicts a recent meta-analysis85, those
co-initiation trials. results pertained to acutely exacerbated patients in whom greater
Despite the overall unfavorable results in high-quality stud- baseline symptom severity created more room for improvement.
ies for total, positive and general symptoms, global clinical Conversely, in our meta-analysis, a substantial number of patients
impression, depression, treatment response and study discon- had likely benefitted to some degree from antipsychotic mono-
tinuation, augmentation of D2 antagonists with a partial D2 therapy in the past, so that higher residual symptom severity is
agonist was associated with significantly reduced negative probably a marker of less treatment responsiveness.
symptoms, a finding that was confirmed in high-quality stud- The results of this study need to be interpreted within some
ies. Since the treatment of negative symptoms remains a big limitations. These include: a) the still relatively small number

86 World Psychiatry 16:1 - February 2017

of double-blind studies comparing antipsychotic augmenta- across relevant outcome domains, including patients with care-
tion with monotherapy in schizophrenia, particularly for aug- fully defined insufficient response to antipsychotic monotherapy.
mentation of non-clozapine antipsychotics and for specific
antipsychotic co-treatment pairs; b) the heterogeneous study
origin, design, definition and degree of insufficient response to ACKNOWLEDGEMENT
monotherapy, measurements and outcomes; c) the limited This study was supported in part by the Zucker Hillside Hospital National
number of studies reporting negative and depressive symp- Institute of Mental Health (NIMH) Advanced Center for Intervention and Serv-
ices Research for the Study of Schizophrenia (grant no. P30MH090590).
toms as well as adverse effects, which were often not compre-
hensively assessed or reported; d) the potential influence of
cultural or ethnic differences (although we addressed regional
effects in pre-planned subgroup analyses, yet the effect of REFERENCES
studies from China overlapped almost 100% with an efficacy 1. Hegarty JD, Baldessarini RJ, Tohen M et al. One hundred years of schizo-
signal only detected in open-label/low-quality studies); e) the phrenia: a meta-analysis of the outcome literature. Am J Psychiatry 1994;
151:1409-16.
exclusion of studies focusing on adverse effects from the effi-
2. Jones PB, Barnes TRE, Davies L et al. Randomized controlled trial of the
cacy analyses to reduce heterogeneity (supported by a >10- effect on Quality of Life of second- vs first-generation antipsychotic drugs
point lower mean PANSS/converted BPRS symptom severity in schizophrenia: Cost Utility of the Latest Antipsychotic Drugs in Schizo-
phrenia Study (CUtLASS 1). Arch Gen Psychiatry 2006;63:1079-87.
in side effect-focused versus efficacy-focused studies); f) the
3. Lieberman JA, Stroup TS, McEvoy JP et al. Effectiveness of antipsychotic
combination of studies augmenting clozapine and non-clozapine drugs in patients with chronic schizophrenia. N Engl J Med 2005;353:1209-
antipsychotics, potentially representing different patient sub- 23.
4. Robinson DG, Woerner MG, McMeniman M et al. Symptomatic and func-
groups (although mean baseline PANSS/converted BPRS total
tional recovery from a first episode of schizophrenia or schizoaffective dis-
scores were comparable, and subgroup analyses replicated order. Am J Psychiatry 2004;161:473-9.
results in both clozapine and non-clozapine studies); g) the 5. €a
Ja €skela
€inen E, Juola P, Hirvonen N et al. A systematic review and meta-
analysis of recovery in schizophrenia. Schizophr Bull 2013;39:1296-306.
restriction of the distinction between high-quality and low-
6. Falkai P, Wobrock T, Lieberman J et al. World Federation of Societies of
quality to blinding and data analysis (although risk-of-bias tool Biological Psychiatry (WFSBP) guidelines for biological treatment of
clearly confirmed quality differences without having a signifi- schizophrenia, Part 1: Acute treatment of schizophrenia. World J Biol Psy-
chiatry 2005;6:132-91.
cant influence in the meta-regression analysis, suggesting that
7. Moore TA, Buchanan RW, Buckley PF et al. The Texas Medication Algo-
major influencing biases were captured through blinding/ITT rithm Project antipsychotic algorithm for schizophrenia: 2006 update.
categorization); h) the lack of adjustment for multiple compari- J Clin Psychiatry 2007;68:1751-62.
8. Buchanan RW, Kreyenbuhl J, Kelly DL et al. The 2009 schizophrenia PORT
sons (yet, adjustment for multiple comparisons would only have
psychopharmacological treatment recommendations and summary state-
increased the level of non-significance of differences between ments. Schizophr Bull 2010;36:71-93.
groups); i) the potential effect of non-adherence, and j) the lack 9. Lehman AF, Lieberman JA, Dixon LB et al. Practice guideline for the treat-
ment of patients with schizophrenia, second edition. Am J Psychiatry
of detailed data to determine whether the effect of partial agonist
2004;161(Suppl. 2):1-56.
augmentation was mainly on primary or secondary negative 10. Goodwin G, Fleischhacker W, Arango C et al. Advantages and disadvan-
symptoms. tages of combination treatment with antipsychotics ECNP Consensus
Meeting, March 2008, Nice. Eur Neuropsychopharmacol 2009;19:520-32.
In summary, data from this study suggest that high-quality
11. Chinese Medical Association. Guideline for the prevention and treatment
evidence is lacking for antipsychotic augmentation in patients of psychiatric disorders in China. Beijing: Chinese Medical Association,
with schizophrenia, which applies also to patients with inade- 2003.
12. Gallego JA, Bonetti J, Zhang J et al. Prevalence and correlates of antipsy-
quate response to clozapine. The clinical relevance of the neg-
chotic polypharmacy: a systematic review and meta-regression of global
ative symptom advantage of adjunctive partial D2 agonist and regional trends from the 1970s to 2009. Schizophr Res 2012;138:18-28.
treatment needs to be further assessed. Additionally, effects of 13. Correll CU, Shaikh L, Gallego JA et al. Antipsychotic polypharmacy: a sur-
vey study of prescriber attitudes, knowledge and behavior. Schizophr Res
augmentation with a partial D2 agonist versus a switch to a
2011;131:58-62.
partial D2 agonist on negative symptoms need to be compared 14. Ganguly R, Kotzan JA, Miller LS et al. Prevalence, trends, and factors asso-
before these results can be considered for clinical care. Anti- ciated with antipsychotic polypharmacy among Medicaid-eligible schizo-
phrenia patients, 1998-2000. J Clin Psychiatry 2004;65:1377-88.
psychotic augmentation treatment should also be compared
15. Clark RE, Bartels SJ, Mellman T et al. Recent trends in antipsychotic com-
with high-dose antipsychotic monotherapy or augmentation bination therapy of schizophrenia and schizoaffective disorder: implica-
with psychosocial interventions. Furthermore, non-clozapine tions for state mental health policy. Schizophr Bull 2002;28:75-84.
16. Nielsen J, Dahm M, Lublin H et al. Psychiatrists’ attitude towards and
antipsychotic augmentation strategies should be compared
knowledge of clozapine treatment. J Psychopharmacol 2010;24:965-71.
against a switch to clozapine or to improving adherence, in- 17. Correll CU, Gallego JA. Antipsychotic polypharmacy. Psychiatr Clin North
cluding monotherapy with a long-acting injectable antipsy- Am 2012;35:661-81.
18. Faries D, Ascher-Svanum H, Zhu B et al. Antipsychotic monotherapy and
chotic, which are each more rational choices for addressing
polypharmacy in the naturalistic treatment of schizophrenia with atypical
antipsychotic non-response. Another gap is the systematic antipsychotics. BMC Psychiatry 2005;5:26.
assessment of adverse effects of antipsychotic augmentation, 19. Tapp A, Wood AE, Secrest L et al. Combination antipsychotic therapy in
clinical practice. Psychiatr Serv 2003;54:55-9.
extending also to cognition, functioning and subjective well-
20. Correll CU, Rummel-Kluge C, Corves C et al. Antipsychotic combinations
being. Finally, more high-quality trials are needed that examine vs monotherapy in schizophrenia: a meta-analysis of randomized con-
antipsychotic augmentation in non-clozapine-treated patients trolled trials. Schizophr Bull 2009;35:443-57.

World Psychiatry 16:1 - February 2017 87

21. Barbui C, Signoretti A, Mulè S et al. Does the addition of a second antipsy- 50. Egger M, Davey Smith G, Schneider M et al. Bias in meta-analysis detected
chotic drug improve clozapine treatment? Schizophr Bull 2009;35:458-68. by a simple, graphical test. BMJ 1997;315:629-34.
22. Gallego JA, Nielsen J, De Hert M et al. Safety and tolerability of antipsy- 51. Duval S, Tweedie R. A nonparametric “trim and fill” method of accounting
chotic polypharmacy. Expert Opin Drug Saf 2012;11:527-42. for publication bias in meta-analysis. J Am Stat Assoc 2000;95:89-98.
23. Rupnow MFT, Greenspan A, Gharabawi GM et al. Incidence and costs of 52. Liu YD, Huang WF, Huang SJ. A control study of clozapine combined with
polypharmacy: data from a randomized, double-blind, placebo-controlled fluphenazine in refractory schizophrenia. Med J Chin People’s Health
study of risperidone and quetiapine in patients with schizophrenia or 2008;20:1274.
schizoaffective disorder. Curr Med Res Opin 2007;23:2815-22. 53. Friedman JI, Lindenmayer J-P, Alcantara F et al. Pimozide augmentation of
24. Zhu B, Ascher-Svanum H, Faries DE et al. Cost of antipsychotic polyphar- clozapine inpatients with schizophrenia and schizoaffective disorder unre-
macy in the treatment of schizophrenia. BMC Psychiatry 2008;8:19. sponsive to clozapine monotherapy. Neuropsychopharmacology 2011;36:
25. Baandrup L, Sørensen J, Lublin H et al. Association of antipsychotic poly- 1289-95.
pharmacy with health service cost: a register-based cost analysis. Eur J 54. Gunduz-Bruce H, Oliver S, Gueorguieva R et al. Efficacy of pimozide aug-
Health Econ 2012;13:355-63. mentation for clozapine partial responders with schizophrenia. Schizophr
26. De Hert M, Detraux J, van Winkel R et al. Metabolic and cardiovascular Res 2013;143:344-7.
adverse effects associated with antipsychotic drugs. Nat Rev Endocrinol 55. Chang JS, Ahn YM, Park HJ et al. Aripiprazole augmentation in clozapine-
2012;8:114-26. treated patients with refractory schizophrenia: an 8-week, randomized,
27. De Hert M, Dobbelaere M, Sheridan EM et al. Metabolic and endocrine double-blind, placebo-controlled trial. J Clin Psychiatry 2008;69:720-31.
adverse effects of second-generation antipsychotics in children and ado- 56. Fan X, Borba CPC, Copeland P et al. Metabolic effects of adjunctive aripi-
lescents: a systematic review of randomized, placebo controlled trials and prazole in clozapine-treated patients with schizophrenia. Acta Psychiatr
guidelines for clinical practice. Eur Psychiatry 2011;26:144-58. Scand 2013;127:217-26.
28. Correll CU. Balancing efficacy and safety in treatment with antipsychotics. 57. Fleischhacker WW, Heikkinen ME, Oli e J-P et al. Effects of adjunctive treat-
CNS Spectr 2007;12(Suppl. 17):12-20. ment with aripiprazole on body weight and clinical efficacy in schizophre-
29. Misawa F, Shimizu K, Fujii Y et al. Is antipsychotic polypharmacy associat- nia patients treated with clozapine: a randomized, double-blind, placebo-
ed with metabolic syndrome even after adjustment for lifestyle effects? A controlled trial. Int J Neuropsychopharmacol 2010;13:1115-25.
cross-sectional study. BMC Psychiatry 2011;11:118. 58. Guan Y. The efficacy and safety of aripiprazole combined with clozapine
30. Trikalinos TA, Churchill R, Ferri M et al. Effect sizes in cumulative meta- in the treatment of refractory schizophrenia. J Bengbu Med Coll 2014;39:
analyses of mental health randomized trials evolved over time. J Clin Epi- 1084-6.
demiol 2004;57:1124-30. 59. Muscatello MRA, Bruno A, Pandolfo G et al. Effect of aripiprazole augmen-
31. Taylor DM, Smith L, Gee SH et al. Augmentation of clozapine with a sec- tation of clozapine in schizophrenia: a double-blind, placebo-controlled
ond antipsychotic – a meta-analysis. Acta Psychiatr Scand 2012;125:15-24. study. Schizophr Res 2011;127:93-9.
32. Sommer IE, Begemann MJH, Temmerman A et al. Pharmacological aug- 60. Sun WH, Chen Y, Wang KY. The study of combining clozapine and aripi-
mentation strategies for schizophrenia patients with insufficient response to prazole for the treatment of negative symptoms of schizophrenia. Med J
clozapine: a quantitative literature review. Schizophr Bull 2012;38:1003-11. Chin People’s Health 2012;24:2592-4.
33. Freudenreich O, Goff DC. Antipsychotic combination therapy in schizo- 61. Lin P, Cheng J, Mao G. Paliperidone ER combined with clozapine in the
phrenia. A review of efficacy and risks of current combinations. Acta Psy- treatment of patients with refractory schizophrenia. Hebei Med 2014;20:
chiatr Scand 2002;106:323-30. 1668-71.
34. Stahl SM. Antipsychotic polypharmacy: evidence based or eminence 62. Freudenreich O, Henderson DC, Walsh JP et al. Risperidone augmentation
based? Acta Psychiatr Scand 2002;106:321-2. for schizophrenia partially responsive to clozapine: a double-blind, place-
35. Stroup DF, Berlin JA, Morton SC et al. Meta-analysis of observational stud- bo-controlled trial. Schizophr Res 2007;92:90-4.
ies in epidemiology: a proposal for reporting. Meta-analysis of Observa- ciog
63. Anil Yag lu AE, Kivircik Akdede BB, Turgut TI et al. A double-blind con-
tional Studies in Epidemiology (MOOSE) group. JAMA 2000;283:2008-12. trolled study of adjunctive treatment with risperidone in schizophrenic
36. Moher D, Liberati A, Tetzlaff J et al. Preferred reporting items for systemat- patients partially responsive to clozapine: efficacy and safety. J Clin Psy-
ic reviews and meta-analyses: the PRISMA statement. PLoS Med 2009;6: chiatry 2005;66:63-72.
e1000097. 64. Honer WG, Thornton AE, Chen EYH et al. Clozapine alone versus cloza-
37. Kay SR, Fiszbein A, Opler LA. The positive and negative syndrome scale pine and risperidone with refractory schizophrenia. N Engl J Med 2006;
(PANSS) for schizophrenia. Schizophr Bull 1987;13:261-76. 354:472-82.
38. Ventura J, Lukoff KH, Nuechterlein KH et al. Manual for the Expanded 65. Josiassen RC, Joseph A, Kohegyi E et al. Clozapine augmented with risperi-
Brief Psychiatric Rating Scale. Int J Methods Psychiatr Res 1993;3:227-44. done in the treatment of schizophrenia: a randomized, double-blind, pla-
39. Andreasen NC. The Scale for the Assessment of Positive Symptoms cebo-controlled trial. Am J Psychiatry 2005;162:130-6.
(SAPS). Iowa City: University of Iowa, 1984. 66. Hu X. Clozapine combined with risperidone in replacement of single cloza-
40. Andreasen NC. The Scale for the Assessment of Negative Symptoms pine in schizophrenia treatment. Med J Chin People’s Health 2014;26:21-2.
(SANS). Iowa City: University of Iowa, 1984. 67. Weiner E, Conley RR, Ball MP et al. Adjunctive risperidone for partially
41. Guy W. ECDEU assessment manual for psychopharmacology, revised. responsive people with schizophrenia treated with clozapine. Neuropsy-
Rockville: US Department of Health, Education, and Welfare, 1976. chopharmacology 2010;35:2274-83.
42. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry 68. Nielsen J, Emborg C, Gydesen S et al. Augmenting clozapine with sertin-
1960;23:56-62. dole: a double-blind, randomized, placebo-controlled study. J Clin Psy-
43. Addington D, Addington J, Schissel B. A depression rating scale for schizo- chopharmacol 2012;32:173-8.
phrenics. Schizophr Res 1990;3:247-51. 69. Shiloh R, Zemishlany Z, Aizenberg D et al. Sulpiride augmentation in peo-
44. Hall RCW. Global Assessment of Functioning: a modified scale. Psychoso- ple with schizophrenia partially responsive to clozapine. A double-blind,
matics 1995;36:267-75. placebo-controlled study. Br J Psychiatry 1997;171:569-73.
45. Higgins JPT, Altman DG, Gøtzsche PC et al. The Cochrane Collaboration’s 70. Jiang T, Zhang YF, Chi Y et al. The treatment efficacy of clozapine com-
tool for assessing risk of bias in randomised trials. BMJ 2011;343:d5928. bined with ziprasidone in refractory schizophrenia. Clin J Postgrad Med
46. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials 2011;34:62-4.
1986;7:177-88. 71. Muscatello MRA, Pandolfo G, Mic o U et al. Augmentation of clozapine
47. Chen J-X, Su Y-A, Bian Q-T et al. Adjunctive aripiprazole in the treatment of with ziprasidone in refractory schizophrenia: a double-blind, placebo-
risperidone-induced hyperprolactinemia: a randomized, double-blind, pla- controlled study. J Clin Psychopharmacol 2014;34:129-33.
cebo-controlled, dose-response study. Psychoneuroendocrinology 2015; 72. Shim JC, Shin JG, Kelly DL et al. Adjunctive treatment with a dopamine
58:130-40. partial agonist, aripiprazole, for antipsychotic-induced hyperprolactine-
48. Liu L, Qi SG, Dong XH et al. A placebo-controlled trail of adjunctive treat- mia: a placebo-controlled trial. Am J Psychiatry 2007;164:1404-10.
ment with aripiprazole for risperidone-induced hyperprolactinemia. Chin 73. Kane JM, Correll CU, Goff DC et al. A multicenter, randomized, double-
J Health Psychol 2011;19:1288-90. blind, placebo-controlled, 16-week study of adjunctive aripiprazole for
49. Leucht S, Rothe P, Davis JM et al. Equipercentile linking of the BPRS and schizophrenia or schizoaffective disorder inadequately treated with que-
the PANSS. Eur Neuropsychopharmacol 2013;23:956-9. tiapine or risperidone monotherapy. J Clin Psychiatry 2009;70:1348-57.

88 World Psychiatry 16:1 - February 2017

74. Lee BJ, Lee SJ, Kim MK et al. Effect of aripiprazole on cognitive function 81. Millan MJ, Fone K, Steckler T et al. Negative symptoms of schizophrenia:
and hyperprolactinemia in patients with schizophrenia treated with ris- clinical characteristics, pathophysiological substrates, experimental mod-
peridone. Clin Psychopharmacol Neurosci 2013;11:60-6. els and prospects for improved treatment. Eur Neuropsychopharmacol
75. Ou Y, Shi J, Chen F. The effect of aripiprazole on serum levels of Insulin- 2014;24:645-92.
like Growth factor-1 in schizophrenia treatment with olanzapine. Chin J 82. Rabinowitz J, Berardo CG, Bugarski-Kirola D et al. Association of promi-
Health Psychol 2014;22:161-3. nent positive and prominent negative symptoms and functional health,
76. Yasui-Furukori N, Kaneda A, Sugawara N et al. Effect of adjunctive treat- well-being, healthcare-related quality of life and family burden: a CATIE
ment with aripiprazole to atypical antipsychotics on cognitive function in analysis. Schizophr Res 2013;150:339-42.
schizophrenia patients. J Psychopharmacol 2012;26:806-12. 83. Correll CU, Skuban A, Ouyang J et al. Efficacy and safety of brexpipra-
77. Zhao YQ. A control study on aripiprazole reduces risperidone-induced zole for the treatment of acute schizophrenia: a 6-week randomized,
weight gain and increase of prolactin in patients with schizophrenia. double-blind, placebo-controlled trial. Am J Psychiatry 2015;172:870-
Sichuan Ment Health 2013;26:122-3. 80.
78. Zhou P, Liu LQ, Hao JF et al. The study of aripiprazole on preventing the 84. Kane JM, Zukin S, Wang Y et al. Efficacy and safety of cariprazine in acute
hyperprolactinemia induced by antipsychotics on female. J Int Psychiatry exacerbation of schizophrenia: results from an international, phase III
2014;41:69-72. clinical trial. J Clin Psychopharmacol 2015;35:367-73.
79. Liang Y, Liu MD. A control study of aripiprazole combined with low-dose 85. Furukawa TA, Levine SZ, Tanaka S et al. Initial severity of schizophrenia
clozapine in female schizophrenia. J Clin Psychosom Dis 2013;10:3. and efficacy of antipsychotics: participant-level meta-analysis of 6
80. Kotler M, Strous RD, Reznik I et al. Sulpiride augmentation of olanzapine in placebo-controlled studies. JAMA Psychiatry 2015;72:14-21.
the management of treatment-resistant chronic schizophrenia: evidence
for improvement of mood symptomatology. Int Clin Psychopharmacol 2004;
19:23-6. DOI:10.1002/wps.20387

World Psychiatry 16:1 - February 2017 89