CNS Drugs

https://doi.org/10.1007/s40263-019-00612-8

SYSTEMATIC REVIEW

Treatment Strategies for Clozapine‑Induced Sialorrhea: A Systematic

Review and Meta‑analysis
Shih‑Yu Chen1,2 · Gopi Ravindran1,2 · Qichen Zhang2 · Steve Kisely1,2   · Dan Siskind1,2 

© Springer Nature Switzerland AG 2019

Abstract
Background  Clozapine is the most effective medication for treatment-refractory schizophrenia. However, it has a high burden
of adverse events, including common adverse events such as sialorrhea. Sialorrhea can lead to severe physical complica-
tions such as aspiration pneumonia, as well as psychological complications including embarrassment and low self-esteem.
Compromised adherence and treatment discontinuation can occur due to intolerability. There have been no meta-analyses
examining strategies to mitigate clozapine-induced sialorrhea.
Methods  We systematically searched Chinese and Western research databases for randomised controlled trials examining
agents for clozapine-induced sialorrhea. No limit to language or date were applied to the search. Where sufficient data for
individual agents was available, pairwise meta-analyses were conducted. Results were provided as risk ratios and number
needed to treat. Sensitivity analysis was conducted by study quality. Adverse events were provided as number needed to harm.
Results  19 studies provided data for use in the meta-analysis. Improvement in clozapine-induced sialorrhea was seen in
meta-analyses of propantheline (studies = 6, risk ratio [RR] 2.38, 95% confidence interval [CI] 1.52–3.73; number needed
to treat [NNT] 3, 95% CI 1.9–2.7), diphenhydramine (studies = 5, RR 3.09, 95% CI 2.36–4.03; NNT 2, 95% CI 1.5–2.0),
chlorpheniramine (studies = 2, RR 2.37, 95% CI 1.59–3.55; NNT 3, 95% CI 1.6–3.5), and benzamide derivatives (odds ratio
[OR] 6.93, 95% CI 3.03–15.86). When meta-analyses were limited to high-quality studies, all these results remained sig-
nificant. Single studies of benzhexol, cyproheptadine, doxepin and Kongyan Tang showed promise. Propantheline increased
rates of constipation with a number needed to harm (NNH) of 9 (95% CI 4.2–204.1).
Conclusion  Clozapine-induced sialorrhea is a potentially serious adverse event. Included studies in this meta-analysis were
limited by poor study quality. Diphenhydramine, chlorpheniramine and benzamide derivatives appear to have the best
supporting evidence and lowest reported adverse events. Caution should be exercised when using propantheline given its
increased risk of constipation.

Key Points 

Clozapine-induced sialorrhea is very common and can

lead to serious complications such as aspiration pneumo-
nia.

Electronic supplementary material  The online version of this

We found meta-analytic level evidence to support the use
article (https​://doi.org/10.1007/s4026​3-019-00612​-8) contains of diphenhydramine, propantheline, chlorpheniramine
supplementary material, which is available to authorized users. and benzamide derivatives to reduce the rates of clozap-
ine-induced sialorrhea.
* Dan Siskind
d.siskind@uq.edu.au
1
Metro South Addiction and Mental Health Service, Level 2
Mental Health, Woolloongabba Community Health Centre,
228 Logan Rd, Brisbane, QLD, Australia
2
School of Medicine, University of Queensland, Brisbane,
QLD, Australia

Vol.:(0123456789)

1 Introduction
Schizophrenia affects 0.7% of the population globally [1].
Antipsychotics are the major pharmacological treatment of
schizophrenia, yet up to 30% of patients are treatment refrac-
tory, defined as having an inadequate treatment response
to two or more trials of antipsychotics [2]. Clozapine is the
most effective medication for reducing positive symptoms [3]
and hospitalisations [4] for people with treatment-resistant
schizophrenia.
Although rare, the most serious adverse drug reactions
(ADRs) associated with clozapine are haematological [5] and
cardiac [6]. Sialorrhea is one of the most common clozapine-
associated ADRs with prevalence of up to 92% [7]. While
48% of people on clozapine experience daytime sialorrhea,
nocturnal hypersalivation occurs more frequently (85%) [7].
Clozapine-induced sialorrhea is disabling and negatively
impacts quality of life [7]. It can lead to both physical com-
plications including aspiration pneumonia, painful inflamma-
tion of the salivary glands, insomnia, malodour, skin irritation
and infection [8, 9], as well as psychological complications
including embarrassment, social stigma and low self-esteem
[10, 11]. These adverse reactions can become intolerable and
lead to compromised adherence [12] or treatment discontinu-
ation [13].
In addition to sialorrhea, Clozapine has also been reported
to cause dry mouth with prevalence of 53% [14]. The mecha-
nisms by which clozapine exerts these opposing effects are
complex and poorly understood. As such, treatment rationale
for clozapine-induced sialorrhea has been lacking [15].
Non-pharmacological interventions for clozapine-induced
sialorrhea include a fitted towelling pillowcase at night or
reducing the clozapine dosage. However, towels or pillow-
cases do not address the underlying issues while reducing the
clozapine dose is not always clinically feasible [16].
There have been no recent systematic reviews of pharmaco-
logical interventions for clozapine-induced sialorrhea, while
older reviews were restricted to case reports or were limited
to English language publications [8, 17]. In particular, there
have been no meta-analyses of randomised controlled trials
of interventions to treat clozapine-induced sialorrhea. We
therefore conducted a systematic review and meta-analysis of
randomised controlled trials on treatment of clozapine-induced
sialorrhea.
2 Methods
2.1 Protocol and Study Registration
This study was registered with PROSPERO (registration ID
CRD42017065908). Ethical approval was not required for
S.-Y. Chen et al.
this study, as all included data had been previously published
with ethical approval obtained from relevant regulatory bod-
ies by the original researchers. We followed the Preferred
Reporting Items for Systematic Reviews and Meta-Analyses
(PRISMA) guidelines [18].
2.2 Eligibility Criteria
Studies were included if they met the following criteria:
(1) randomised controlled trials (RCTs) (these could be
either parallel or cross-over designs); (2) published jour-
nal articles; (3) participants experienced clozapine-induced
sialorrhea; (4) participants randomised to pharmacological
treatment or placebo. Exclusion criteria were (1) non-RCTs;
(2) non-pharmacological intervention; (3) non-placebo con-
trolled. There were no restrictions to age, gender, diagnosis
or language.
2.3 Search Strategy and Study Selection
PubMed, Embase, PsycINFO, The Cochrane Library
(Cochrane Database of Systematic Reviews) and Chinese
Knowledge Resource Integrated Database (CNKI) databases
were searched from inception to 6 July 2018 using terms
related to sialorrhea and clozapine (Supplementary Table 1,
see electronic supplementary material [ESM]). No limit to
language was applied to the search. Only fully published
studies were included. The Mandarin translation of these
terms was used for searches in the CNKI database. Refer-
ence lists of systematic reviews were screened for potential
additional studies. All identified studies were screened at the
title and abstract level, and then at full text by two authors
independently; SC and GR screened the studies from non-
Chinese databases while native Mandarin speakers SC and
QZ screened the studies identified in the Chinese database.
2.4 Data Extraction
Data from studies identified in non-CNKI databases were
extracted by GR and checked by SC. Data from studies
identified in the CNKI database were extracted by SC and
checked by QZ.
2.5 Data Items
Data extracted from included studies included the following:
year and setting of study, number, age and gender of partici-
pants, diagnostic tools and diagnoses, dosage and duration
on clozapine, dose and duration of intervention(s), outcome
measures of clozapine-induced sialorrhea and adverse drug
reactions.
Treatment of Clozapine-Induced Sialorrhea: A Meta-analysis
2.6 Outcomes
The primary outcome was change in sialorrhea, measured
using either qualitative methods or objective rating scales.
2.7 Quality Assessment
The quality of the included studies was assessed based on
the Cochrane Collaboration guidelines [19] assessing the
following:
a. adequate generation of allocation sequence;
b. concealment of allocation sequence to participant and/
or assessor;
c. blinding of participants and personnel from knowledge
of which intervention a participant received;
d. blinding of outcome assessors from knowledge of which
intervention a participant received;
e. completeness of outcome data for each main outcome,
including attrition and exclusions from the analysis;
f. possibility of selective outcome reporting;
g. other sources of potential bias including pharmaceutical
company funding.
Studies were considered to be of low quality if they had
three or more factors with high or unknown risk of bias.
2.8 Statistical Analysis
We used Review Manager (RevMan) version 5.3 for Mac
for the meta-analyses. Where dichotomous data was pro-
vided, analyses of risk ratio (RR) were conducted. When
dichotomous and continuous data was available for inter-
ventions of the same class, results were log transformed in
Comprehensive Meta-Analysis version 3 for Windows and
analysed as log risk ratios in RevMan. A number needed to
treat (NNT) was calculated for each intervention.
Where possible, sensitivity analyses were conducted for
study quality and duration. If there were outcomes with ten
or more studies, we tested for publication bias using funnel
plot asymmetry [20].
We used the random effects model for all the analyses as
we could not definitely exclude between-study variation,
even in the absence of statistical heterogeneity, given the
range of medications under review.
Where possible, meta-analyses of rates of ADRs
between intervention and control were conducted where
three or more studies of the same intervention provided
data on a specific ADR. A number needed to harm (NNH)
was calculated for these ADRs.
3 Results
3.1 Study Selection
The selection process is summarised in Fig. 1. We found
2149 and 129 studies in the initial searches of the non-Chi-
nese and Chinese databases, respectively. After removal
of duplicate studies, 1299 studies were screened at title
and abstract level. Of these, 119 studies were assessed at
full-text level. The reasons for exclusion are listed in Fig. 1
and Supplementary Table 2: Excluded Studies (see ESM).
A total of 19 studies were included in the meta-analysis.
3.2 Study Characteristics
Studies were published between 1993 and 2017. Charac-
teristics of the included studies are provided in Table 1.
Studies ranged in duration from 10 days to 6 weeks. Only
one study included clozapine levels, with the mean clo-
zapine level of the intervention group: 1887.6 nmol/L (SD
868.33 nmol/L) and control group: 1758.10 nmol/L (SD
849.49 nmol/L) [12].
Rating scales used include the Nocturnal Hypersaliva-
tion Rating Scale (NHRS), used in Kreinin et al. [21] and
Kreinin et al. [22]. This is a validated five-point scale rang-
ing from 0 for no hypersalivation to 4 for severe hypersali-
vation leading to awakening more than three times during
the night. The Toronto Nocturnal Hypersalivation Scale
(TNHS) was developed by Sockalingam et al. [12] and is a
validated five-point scale ranging from 0 for no hypersali-
vation to 4 for very severe hypersalivating leading to awak-
ening during the night. The Patient Global Impression of
Improvement (PGI-I) [23] is a validated seven-point par-
ticipant self-report scale with ratings of sialorrhea from
very much worse to very much improved. The modified
Simpson Angus [24, 25] is a validated five-point observer-
rated scale of saliva in the mouth. The other studies rated
clozapine-induced sialorrhea by measuring the diameter
of wet pillow area.
All included studies except Kreinin et al. [21] provided
dichotomous data on the number of participants in inter-
vention and control arms meeting response criteria for
reduction in sialorrhea. Rating scales and response criteria
are outlined in Supplementary Table 3 (see ESM).
3.3 Study Quality
Using the Cochrane Collaboration’s assessment-of-bias
tool [19], three out of the four studies from the non-
Chinese databases were of high quality (Supplementary
 S.-Y. Chen et al.

PRISMA 2009 Flow Diagram

Records idenfied through Addional records idenfied

database searching through other sources
• Pubmed, PsycInfo, Embase • Pubmed, PsycInfo, Embase
Idenficaon

and Cochrane n = 2149 and Cochrane n = 0

• CNKI n = 129 • CNKI n = 0

Records aer duplicates removed

• n =1299
Screening

Records excluded
• n = 1180

Full-text arcles assessed for

eligibility
Full-text arcles excluded
Eligibility

• n = 119
n = 100
• Non-RCT n = 55
• Not placebo
controlled n = 24
• Non-pharmacological
intervenon n = 1
• Data published
elsewhere n = 7
Studies included in • No usable data n = 5
Included

quantave synthesis • Review = 2

(meta-analysis) • Study proposal only n
• n = 19 =2
• Unable to locate study
n=4

From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009). Preferred Reporting Items for Systematic Reviews and Meta-
Analyses: The PRISMA Statement. PLoS Med 6(7): e1000097. doi:10.1371/journal.pmed1000097

Fig. 1  PRISMA 2009 flow diagram of study selection. CNKI Chinese Knowledge Resource Integrated Database, RCT​ randomised controlled
trial
Table 4, see ESM). Six out of the fifteen studies from the 3.4 Response
Chinese database were of high quality. Two studies were
open label and three studies did not provide a placebo Definition of response varied between studies. These defi-
intervention to the control groups. nitions are outlined in Supplementary Table 3 (see ESM).
Table 1  Study characteristics
Paper Interventions (Dos-Duration Setting No. of Gender Age (mean) (intervention/ Diagnoses Clozapine duration (interven- Clozapine dose/day
age per day) participants (male %) control) (diagnostic tion/control) (intervention/control)
(interven- (interven- tool)
tion/con- tion/con-
trol) trol)

Anti-muscarinics
Tao et al., 2007 Propantheline 6 weeks Inpatient 33/33 48.5/54.5 38.38 ± 4.13/40.13 ± 6.09 Schizophrenia NI 368.8 ± 111.2/
[29] 30–120 mg (CCMD-3) 389.1 ± 83.2
Qin et al., 1999 Propantheline NI Inpatient 30/30 NI NI Schizophre- NI Max 500 mg
[30] 30–75 mg nia, mania
(NI)
Lin, 1999 [26] Propantheline 10 days Inpatient 16/16 100/100 37 ± 7.2/39 ± 6.8 NI (NI) 4.3 ± 1.6 years/ 217 ± 94 mg/
15 mg BD 4.5 ± 1.4 years 244 ± 98.8 mg
Yang et al., Propantheline 2 weeks Inpatient 31/39 NI NI NI (NI) NI NI
1999 [27] 60–120 mg
Treatment of Clozapine-Induced Sialorrhea: A Meta-analysis

Gong et al., Propantheline 10 days Inpatient 31/39 NI NI Schizophrenia NI NI

1998 [28] 60–120 mg (NI)
Fan et al., 1996 Propantheline NI Inpatient 23/23 100/100 24.75 ± 5.18 (combined) Schizophrenia NI 200–400 mg
[31] 15 mg TDS (CCMD-2)
Cheng and Benzhexol 1 tablet NI Inpatient 30/30 NI NI Schizophrenia NI Max 450 mg
Jiang, 2005 (CCMD-
[32] 2-R)
Man et al., Glycopyrrolate 18 daysa Community 16/16 65.6 (com- 38.9 ± 11.2 (combined) Schizophre- 83.9 ± 63.2 months Median 250 mg (200 mg
2017 [23] 1 mg blind bined) nia (23), IQR)
crossover with schizoaf-
control, then fective (4),
2 mg open label psychosis
NOS (3),
bipolar type
1 (2) (DSM-
IV)
Sockalingam Ipratropium 0.03% 5 or 19 outpa- 10/10 70/70 40.0 ± 12.0/38.8 ± 7.4 Schizophre- 41.4 ± 43.3 months/44.6 ± 46 376.3 ± 70.8 mg/320.0 
et al., 2009 2 sprays at 6 weeks tients: 1 nia, schiz- .4 months ± 107.9 mg
[12] bedtime inpatient oaffective
(DSM-IV)
Anti-histamines
Zhao et al., Diphenhydramine 2 weeks Inpatient 31/31 100/100 42.3 ± 9.9/41.0 ± 8.3 Schizophrenia 29.6 ± 27.8 weeks/30.2 ± 29 306.1 ± 156.0 mg/351.6
2000 [33] (CCMD- .6 weeks  ± 126.7 mg
2-R)
Yang et al., Diphenhydramine 2 weeks Inpatient 32/39 NI NI NI (NI) NI NI
1999 [27] 100–200 mg
Gong et al., Diphenhydramine 10 days Inpatient 32/19b NI NI Schizophrenia NI NI
1998 [28] 100–200 mg (NI)


Table 1  (continued)
Paper Interventions (Dos-Duration Setting No. of Gender Age (mean) (intervention/ Diagnoses Clozapine duration (interven- Clozapine dose/day
age per day) participants (male %) control) (diagnostic tion/control) (intervention/control)
(interven- (interven- tool)
tion/con- tion/con-
trol) trol)
Astemizole 36/19b
10–20 mg
Lu et al., 1998 Diphenhydramine 10 days Inpatient 30/30 0/0 26 ± 11.0/27.0 ± 5.0 Schizophrenia NI 335.2 ± 103.7 mg/352.7 
[34] 50 mg (CCMD- ± 105.2 mg
2-R)
Xu et al., 1997 Diphenhydramine 2 weeks Inpatient 60/30 100/100 18–50 (combined) Schizophrenia NI NI
[24] 25–50 mg (NI)
Li et al., 1993 Astemizole 10 mg 2 weeks Inpatient 11/11 54.5/54.5 39.5 ± 4.8/42.3 ± 7.5 Schizophrenia NI NI
[36] (NI)
Qin et al., 1999 Chlorpheniramine NI Inpatient 30/30 NI NI Schizophre- NI Max 500 mg
[30] 16–24 mg nia, mania
(NI)
Lu et al., 1994 Chlorpheniramine 10 days Inpatient 26/24 100/100 30 ± 10 (combined) Schizophrenia NI NI
[35] 16 mg (NI)
Yang et al., Cyproheptadine 10 days Inpatient 24/23 45.8/47.8 34.2 ± 12.8/33.4 ± 10.2 Schizophrenia NI 325 ± 150 mg
1996 [25] 8–36 mg (CCMD-2)
Anti-depressants
Yang et al., Doxepin 2 weeks Inpatient 30/39 NI NI NI (NI) NI NI
1999 [27] 50–100 mg
Substitute benzamide derivatives
Tao et al., 2007 Sulpiride 6 weeks Inpatient 34/33 55.9/54.5 39.50 ± 6.51/40.13 ± 6.09 Schizophrenia NI 369.6 ± 160.0 mg/389.1 
[29] 200–600 mg (CCMD-3) ± 83.2 mg
Kreinin et al., Amisulpride 3 weeks Inpatient 9/11 50 (com- NI Schizophrenia NI NI
2006 [21] (400 mg/day, bined) (DSM-IV)
up-titrated from
100 mg/day over
1 week)
Kreinin et al., Metoclopramide 3 weeks Inpatient 30/28 90/78.6 41.4 ± 11.1/39.1 ± 12.9 Schizophre- NI 366.7 ± 129.5 mg/306.3 
2016 [22] (10–30 mg/days) nia (48), ± 130.8 mg
schizoaf-
fective (10)
(DSM-IV)
Traditional Chinese medicine
Mao et al., Kongyan Tang 4 weeks Inpatient 32/33 40.6/39.4 30.4 ± 7.4/31.8 ± 7.6 Schizophre- NI 290.6 ± 827 mg/290 ± 
2013 [37] nia, schiz- 71.2 mg
oaffective
(ICD-10)
S.-Y. Chen et al.
 Treatment of Clozapine-Induced Sialorrhea: A Meta-analysis

In our analysis, we have created five groups of interven-

BD twice daily, CCMD Chinese classification of mental disorders, DSM Diagnostic and Statistical Manual of Mental Disorders, ICD International Classification of Diseases, IQR interquartile
310.1 ± 30.4 mg/325.8 
tions: anti-muscarinics, anti-histamines, benzamide deriva-
(intervention/control)
Clozapine dose/day
tives, anti-depressants, and traditional Chinese medicine.

 As two interventions from the same study were compared in the same class, half of the number of participants in the placebo group were used as comparators for each intervention
3.4.1 Anti‑muscarinics

± 4.5 mg 3.4.1.1  Propantheline There were three studies of high

quality [26–28] and three of low quality [29–31] that had
usable data for 164 participants on propantheline and 180
Clozapine duration (interven-

on placebo. People taking propantheline were significantly

Schizophrenia 1.5 ± 7.1 years/1.7 ± 0.8 years

more likely to experience a reduction in sialorrhea (Fig. 2

and Table 2). When only high-quality studies were included
(78 on propantheline and 94 on placebo) [26–28], the results
 6 days, 1-week washout, then 6 days’ crossover, 1-week washout, then 6-day open-label phase. Results from the first 6 days used in our meta-analysis
tion/control)

were similar (RR 1.63, 95% confidence interval [CI] 1.19–

2.25; p = 0.003; I2 29%).

3.4.1.2  Glycopyrrolate One high-quality study [23] had

usable data for 32 participants on glycopyrrolate and 32 on
(diagnostic
Diagnoses

placebo. People taking glycopyrrolate were not significantly

(NI)

more likely to experience a reduction in sialorrhea (Fig. 2

tool)

and Table 2).
Age (mean) (intervention/

3.4.1.3  Benzhexol  One low-quality study [32] had usable

data for 30 participants on benzhexol and 30 on placebo.
People taking benzhexol were significantly more likely to
experience a reduction in sialorrhea (Fig. 2 and Table 2).
control)

3.4.1.4  Ipratropium Bromide  One high-quality study [12]

NI

had usable data for 20 participants on ipratropium bromide

and 20 on placebo. There was no significant difference in
(interven-
tion/con-
(male %)

100/100

the number of responders between the ipratropium group

Gender

and the placebo group (Fig. 2 and Table 2).

trol)

range, NI no information, NOS not otherwise specified, TDS three times daily

When considered as a class, anti-muscarinics were

participants

superior to placebo, with a RR of 2.22 (95% CI 1.58–3.11;

(interven-
tion/con-

p < 0.001; I2 66%).
No. of

trol)
8/8

3.4.2 Anti‑histamines
Inpatient
Setting

3.4.2.1  Diphenhydramine Four high-quality [27, 28, 33,

34] and one low-quality study [24] had usable data for 185
participants on diphenhydramine and 149 on placebo. Peo-
Interventions (Dos-Duration

4 weeks

ple taking diphenhydramine were significantly more likely

to experience a reduction in sialorrhea (Fig. 3 and Table 2).
When only high-quality studies were included (125 on
diphenhydramine and 119 on placebo) [27, 28, 33, 34],
Wu Dan San
age per day)

the results were unchanged (RR 2.48, 95% CI 1.93–3.19;

p < 0.001; I2 74%).
Table 1  (continued)

3.4.2.2  Chlorpheniramine One high-quality [35] and one

Qian and Gao,

low-quality study [30] had usable data for 56 participants on

1996 [38]

chlorpheniramine and 54 on placebo. People taking chlor-

pheniramine were significantly more likely to experience
Paper

a reduction in sialorrhea (Fig.  3 and Table  2). When only

b
a

the high-quality study (26 on chlorpheniramine and 24 on
placebo) [35] was included, the results remained significant
(RR 3.69, 95% CI 1.65–8.28; p = 0.002).
3.4.2.3  Cyproheptadine One low-quality study [25] had
usable data for 24 participants on cyproheptadine and 23
on placebo. People taking cyproheptadine were significantly
more likely to experience a reduction in sialorrhea (Fig. 3
and Table 2).
3.4.2.4  Astemizole One high-quality [28] and one low-
quality study [36] had usable data for 47 participants on
astemizole and 30 on placebo. People taking astemizole
were significantly more likely to experience a reduction in
Fig. 2  Forest plot of risk ratio of anti-muscarinics versus placebo on clozapine-induced sialorrhea. 95% CI 95% confidence interval, Chi2 chi-
squared, df degrees of freedom, I2 I square, P p-value, Tau2 tau-squared, Z z score
S.-Y. Chen et al.
sialorrhea (Fig. 3 and Table 2). When only the high-qual-
ity study (36 on astemizole and 19 on placebo) [28] was
included, the results were not statistically significant (RR
1.06, 95% CI 0.70–1.60; p = 0.80).
When considered as a class, anti-histamines were supe-
rior to placebo, with an RR of 2.76 (95% CI 1.78–4.29;
p < 0.001; I2 80%).
3.4.3 Benzamide Derivatives
One high-quality [22] and two low-quality studies [21, 29]
had usable data for 73 participants on either sulpiride (34),
amisulpride (9) or metoclopramide (30) and 72 on placebo.
Treatment of Clozapine-Induced Sialorrhea: A Meta-analysis

Table 2  Risk ratios and number needed to treat Kongyan Tang (translated to salivation-reducing decoc-
tion) is developed by Mao and colleagues and comprises
Intervention RR (95% CI) NNT (95% CI)
Fuling (the sclerotium of Poriae cocos), Bai Zhu (the
Anti-muscarinics rhizome of Atractylodes macrocephala), Cang Zhu (The
 Propantheline 2.38 (1.52–3.73) 3 (1.8–2.7) rhizome of Atractylodes lancea), Yiyiren (Coix lacryma-
 Glycopyrrolate 3.00 (0.65–13.76) 8 (NS) jobi), white beans (Lablab purpureus), Houpu (Magnolia
 Benzhexol 2.64 (1.64–4.24) 2 (1.3–2.4) officinalis), Chenpi (dried tangerine peel), Banxia (Pinellia
 Ipratropium bromide 1.13 (0.55–2.32) 20 (NS) ternata), Zexie (the rhizome of Alisma orientalis), Guizhi
 Class 2.22 (1.58–3.11) 3 (2.1–3.1) (Ramulus Cinnamomi), ginger and jujube, which all play
Anti-histamines different roles in regulating saliva production according to
 Diphenhydramine 3.09 (2.36–4.03) 2 (1.5–2.0) traditional Chinese medicine theories [37].
 Cyproheptadine 5.03 (2.04–12.42) 2 (1.1–2.0)
 Chlorpheniramine 2.37 (1.59–3.55) 3 (1.6–3.5) 3.4.5.2  Wu Dan San  One low-quality study [38] had usable
 Astemizole 1.53 (1.00–2.34) 4 (2.0–20.6) data for eight participants on Wu Dan San and eight on pla-
 Class 2.76 (1.78–4.29) 2 (1.7–2.2) cebo. There was no significant difference in the number of
Benzamide derivatives 7.33 (2.25–23.85)a b
responders between the Wu Dan San group and the placebo
Anti-depressants group (Supplementary Fig. 2 and Table 2).
 Doxepin 2.82 (1.72–4.60) 2 (1.3–2.7) Wu Dan San comprises Wuyuzi (the seed of Tetradium
Traditional Chinese medicine ruticarpum) and Dannanxing (Rhizoma Arisaematis Cum
 Kongyan Tang 1.55 (1.03–2.32) 4 (2.0–26.3) Bile). The two ingredients are powdered and mixed with rice
 Wu Dan San 1.31 (0.85–2.02) 4 (NS) vinegar to make a paste, which is then applied topically to
the centre of the plantar foot at an acupuncture point named
95% CI 95% confidence interval, NNT number needed to treat, NS not
statistically significant, RR risk ratio Yongquan.
a
 Odds ratio for benzamide derivatives
b
 NNT unable to be calculated 3.5 Adverse Drug Reactions (ADRs)

Eighteen studies commented on adverse drug reactions

Continuous and dichotomous data were combined to gener-
ate an odds ratio (OR). People receiving interventions from
the benzamide derivative class were significantly more
likely to experience a reduction in sialorrhea (OR 6.93,
95% CI 3.03–15.86; p  <  0.001; I2 0%) (Supplementary
Fig. 1, see ESM). When only the high-quality study (30 on
metoclopramide and 28 on placebo) [22] was included, the
results remained significant (OR 7.33, 95% CI 2.25–23.85;
p < 0.01).
3.4.4 Anti‑depressants
3.4.4.1  Doxepin One high-quality study [27] had usable
data for 30 participants on doxepin and 39 on placebo. Peo-
ple taking doxepin were significantly more likely to experi-
ence a reduction in sialorrhea (Table 2).
(Supplementary Table 5); however, only seven studies pro-
vided quantitative data that could be included in a meta-anal-
ysis of comparative rates of ADRs between the intervention
medication and placebo.
Three studies of propantheline [27, 28, 31] provided
usable data on rates of constipation, with rates of constipa-
tion significantly higher for the intervention group compared
with the placebo group (RR 1.86, 95% CI 1.05–3.31), with
an NNH of 9 (95% CI 4.2–204.1). Four studies of diphen-
hydramine [27, 28, 33, 34] provided usable data on rates
of constipation, with no statistically significant difference
in rates between intervention and placebo groups, and an
NNH of 35 (95% CI not statistically significant). No other
interventions had appropriate quantitative data to conduct
meta-analyses of ADRs. All ADRs reported in the included
studies are outlined in Supplementary Table 5 (see ESM).

3.4.5 Traditional Chinese Medicine

4 Discussion
3.4.5.1  Kongyan Tang  One low-quality study [37] had usa-
Improvement in clozapine-induced sialorrhea was dem-
ble data for 32 participants on Kongyan Tang and 33 on pla-
onstrated in our meta-analyses of propantheline, diphen-
cebo. People taking Kongyan Tang were significantly more
hydramine, chlorpheniramine and benzamide deriva-
likely to experience a reduction in sialorrhea (Supplemen-
tives (sulpiride, amisulpride and metoclopramide). When
tary Fig. 2 [see ESM] and Table 2).
meta-analyses were limited to high-quality studies, all
results remained significant. Single studies of benzhexol,
 S.-Y. Chen et al.

Fig. 3  Forest plot of risk ratio of anti-histamines versus placebo on clozapine-induced sialorrhea. 95% CI 95% confidence interval, Chi2 chi-
squared, df degrees of freedom, I2 I square, P p-value, Tau2 tau-squared, Z z score

cyproheptadine, doxepin and Kongyan Tang showed prom-
ise. Of concern, propantheline increased rates of constipa-
tion, which could potentiate clozapine’s constipating adverse
effects [14].
This study provides the first systematic review and meta-
analysis of pharmacological interventions for clozapine-
induced sialorrhea. A major strength of this study was the
inclusion of Chinese studies that were otherwise not acces-
sible through searches of non-Chinese databases. We also
only included randomised, placebo-controlled trials to max-
imise the strength of available evidence. We were able to
include 14 more RCTs than the previous systematic review
from 2011, three of which were published since Bird et al.’s
search was conducted [8]. Bird et al. [8] were not able to
undertake a meta-analysis from their systematic review,
although they did note the positive reported results of the
substituted benzamine class of medications.
The current view of regulatory mechanisms of salivary
secretion is that acinar cells of salivary glands are sup-
plied with muscarinic (M1 and M3) receptors, α1- and
β1-adrenergic receptors and vasoactive intestinal peptide
(VIP) receptors. Muscarinic and α1-adrenergic receptors
are thought to be preferentially responsible for fluid secre-
tion, while protein secretion occurs primarily as a result of
β1-adrenergic and VIP-ergic stimulation. The role of dopa-
mine, serotonin and histamine in the regulation of salivary
secretion is not established. There is no evidence for their
innervations in the salivary glands [15].
Clozapine has been reported to both increase and
reduce the salivary flow rate and the mechanisms by
Treatment of Clozapine-Induced Sialorrhea: A Meta-analysis

which clozapine exerts these opposing effects are complex therefore should be considered with caution when used with
and poorly understood [15]. Clozapine is metabolised to patients on clozapine who may already be sedated. Astemi-
N-desmethylclozapine, an M1 receptor agonist [39]. In an zole is a potent nonsedative ­H1-receptor antagonist [52] that
animal experimental study [13], it was proposed that clo- was not effective in reducing clozapine-induced sialorrhea,
zapine-induced muscarinic M1-receptor agonism interacts and it was voluntarily withdrawn from the market in the US
synergistically with VIP and isoproterenol (a β-adrenergic in 1999 as a result of side effects of QT prolongation and
receptor agonist) to increase salivary volume. Concurrently, ventricular arrhythmia [53].
clozapine exerts an antagonistic effect on muscarinic M3
and α1-adrenergic receptors. In situations where the sali-
4.3 Substituted Benzamide Derivative Class
vary glands are stimulated due to increased demands, such
as during meal times, patients are likely to experience dry
Substituted benzamide derivatives are structurally related
mouth [39].
compounds, including antipsychotics such as amisulpride
Clozapine-induced sialorrhea may also be the result of
and sulpiride, antidepressants such as moclobemide, and
impaired laryngeal peristalsis rather than an increase in
antiemetics such as metoclopramide. While the exact mecha-
saliva production [40, 41]. Finally, clozapine may disrupt
nisms of actions of these medications are largely unclear,
circadian rhythms thereby reversing the normal pattern of
metoclopramide appears to exert its effect from D ­ 2 receptor
the salivary flow being greatest during daytime and falling
antagonism in the central nervous system [22]. Amisulpride
at night [42].
and sulpiride are reported to have high affinity for D­ 2 and D
­ 3
dopamine receptors [51] while having low affinity for sero-
4.1 Anti‑muscarinic Class
tonin, adrenergic, histamine and muscarinic receptors [50].
As a group, these interventions were effective in reduc-
There were improvements in clozapine-associated sialorrhea
ing clozapine-induced sialorrhea. This finding provides
in our meta-analysis of multiple studies for propantheline, as
some support for the hypothesis that dopamine pathways
well as a single study of benzhexol. There were no RCTs for
are involved in clozapine-induced sialorrhea [54, 55]. Ami-
atropine and hyoscine, despite these agents appearing fre-
sulpride/sulpiride augmentation may allow for a reduction
quently in case reports [43–47]. Glycopyrrolate at 1 mg was
in total clozapine dosage, although a recent meta-analysis of
not superior to placebo in a randomised double blind trial;
augmentation of clozapine-refractory schizophrenia did not
however, when the investigators conducted an open-label
find a reduction of psychotic symptoms with amisulpride/
follow-on study of glycopyrrolate at 2 mg, it was superior
sulpiride [56]. Neither of the studies of amisulpride/sulpiride
to placebo in reducing clozapine-induced sialorrhea [23].
included in our meta-analysis adjusted the clozapine dose
The main side effects of anti-muscarinics are due to
[21, 29]. An increase in prolactin was reported in 95% of
peripheral parasympathetic stimulation and include dry
participants on amisulpride [21]. Prolactin levels should
mouth, blurry vision, constipation and urinary retention. Our
therefore be monitored if a trial of amisulpride/sulpiride is
meta-analysis confirmed that constipation was a notable side
undertaken.
effect of propantheline and may potentiate the already con-
Doxepin, a tricyclic antidepressant that has serotonin
stipatory effects of clozapine. Propantheline and glycopyr-
reuptake inhibitor, antimuscarinic and H ­ 1- and H
­ 2-antagonist
rolate do not readily cross the blood–brain barrier [48, 49]
activities, showed promise in a single study [27]. It crosses
and have lower risk of causing central nervous system effects
the blood–brain barrier and is sedating [51].
such as restlessness, confusion and hallucinations than other
anti-muscarinic agents that cross the blood–brain barrier,
4.4 Strength and Limitations
such as benzhexol.
One major strength of our methodology was the inclusion of
4.2 Anti‑histamine Class studies identified in Chinese databases in our searches [57].
This allowed access to a greater pool of studies and may
In our meta-analyses of the multiple studies on diphenhy- reduce the risk of selection bias. A notable limitation of our
dramine, chlorpheniramine and the single study on cypro- findings is that over half of the studies identified in Chinese
heptadine, we noted improvement in clozapine-induced sial- databases were of low quality (compared with a quarter of
orrhea. In addition to their ­H1-receptor antagonism effect, the studies identified in non-Chinese databases). However,
these anti-histamines have various degrees of antagonism when we conducted sensitivity analyses on study quality,
on serotonin, α-adrenergic and anticholinergic effects [50]. all but one intervention (astemizole) yielded similar results.
Diphenhydramine, chlorpheniramine and cyproheptadine In addition, all the included studies were of short duration,
are all considered to be sedating antihistamines [51] and

with the longest being 6 weeks. The short duration limits the
generalisability of the findings.
Another limitation of the published studies is the inherent
challenges in measuring sialorrhea, especially that occurring
overnight. Studies that have used objective measurements of
the flow of saliva are lacking in humans [13]. As a result,
validated objective scales for clozapine-induced sialorrhea
used ratings based on clinicians’ subjective impression of
volume of saliva or diameter of wet pillow as a measure of
saliva volume. There were no validated rating scales used
to objectively measure the volume or flow of saliva. In addi-
tion, no included measures solely rated the patient’s subjec-
tive experience of sialorrhea and its impact on their function.
Sialorrhea rating scales were inconsistently used by studies
included in our meta-analyses.
We were unable to meaningfully analyse the impact
of clozapine dose on sialorrhea from the aggregate data
included in the meta-analysis. As only five studies reported
duration on clozapine, we were unable to do a meaningful
sensitivity analysis of duration on clozapine.
5 Conclusion
While the aetiology of clozapine-induced sialorrhea and the
exact mechanisms of action of many of the pharmacological
interventions remain largely unknown, several interventions
have demonstrated promising results. The most information
on effectiveness is from propantheline, chlorpheniramine,
diphenhydramine and benzamide derivatives as a group
(sulpiride, amisulpride and metoclopramide). Benzhexol,
cyproheptadine and doxepin were shown to be effective
in one study each. Traditional Chinese medicines, such as
Kongyan Tang, are an alternative area for potential research.
Caution should be exercised when using propantheline given
its increased risk of constipation.
Considering patients with clozapine-induced sialorrhea
are likely to remain on clozapine for a long time, future stud-
ies should be of longer duration, with a greater focus on
evaluating tolerability and adverse effects. Validated rating
scales that measure both objective and subjective levels of
clozapine-induced sialorrhea should also be incorporated
into study designs. Finally, non-pharmacological interven-
tions for clozapine-induced sialorrhea should be considered
in conjunction with pharmacological interventions.
Compliance with Ethical Standards  14. Ignjatović Ristić D, Cohen D, Obradović A, Nikić-Đuričić
Funding  Dan Siskind is supported in part by an NHMRC ECF
APP1111136. The authors have no other funding to disclose.
Conflict of interest Shih-Yu Chen, Gopi Ravindran, Qichen Zhang,
Steve Kisely and Dan Siskind declared no potential conflicts of inter-
S.-Y. Chen et al.
est with respect to the research, authorship and/or publication of this
article.
References
1. Saha S, Chant D, Welham J, McGrath J. A systematic review of
the prevalence of schizophrenia. PLoS Med. 2005;2(5):e141.
2. Howes OD, McCutcheon R, Agid O, De Bartolomeis A, Van Bev-
eren NJ, Birnbaum ML, et al. Treatment-resistant schizophrenia:
Treatment Response and Resistance in Psychosis (TRRIP) work-
ing group consensus guidelines on diagnosis and terminology. Am
J Psychiatry. 2016;174(3):216–29.
3. Siskind D, McCartney L, Goldschlager R, Kisely S. Clozapine v.
first-and second-generation antipsychotics in treatment-refractory
schizophrenia: systematic review and meta-analysis. Br J Psychia-
try. 2016;209(5):385–92.
4. Land R, Siskind D, Mcardle P, Kisely S, Winckel K, Hollingworth
SA. The impact of clozapine on hospital use: a systematic review
and meta-analysis. Acta Psychiatr Scand. 2017;135(4):296–309.
5. Myles N, Myles H, Xia S, Large M, Kisely S, Galletly C, et al.
Meta-analysis examining the epidemiology of clozapine-associ-
ated neutropenia. Acta Psychiatr Scand. 2018;138(2):101–9.
6. Rohde C, Polcwiartek C, Kragholm K, Ebdrup B, Siskind D,
Nielsen J. Adverse cardiac events in out-patients initiating clozap-
ine treatment: a nationwide register-based study. Acta Psychiatrica
Scandinavica. 2018;137(1):47–53.
7. Maher S, Cunningham A, O’Callaghan N, Byrne F, Mc Don-
ald C, McInerney S, et al. Clozapine-induced hypersalivation:
an estimate of prevalence, severity and impact on quality of
life. Ther Adv Psychopharmacol. 2016;6(3):178–84. https​://doi.
org/10.1177/20451​25316​64101​9.
8. Bird AM, Smith TL, Walton AE. Current treatment strate-
gies for clozapine-induced sialorrhea. Ann Pharmacother.
2011;45(5):667–75. https​://doi.org/10.1345/aph.1P761​.
9. Citrome L, McEvoy JP, Saklad SR. A guide to the management of
clozapine-related tolerability and safety concerns. Clin Schizophr
Relat Psychoses. 2016. https​://doi.org/10.3371/csrp.saci.07081​6.
10. Liang C-S, Ho P-S, Shen L-J, Lee W-K, Yang F-W, Chiang K-T.
Comparison of the efficacy and impact on cognition of glycopyr-
rolate and biperiden for clozapine-induced sialorrhea in schizo-
phrenic patients: a randomized, double-blind, crossover study.
Schizophr Res. 2010;119(1–3):138–44. https​://doi.org/10.1016/j.
schre​s.2010.02.1060.
11. Sagya R, Weizman A, Katz N. Pharmacological and behavioral
management of some often-overlooked clozapine-induced side
effects. Int Clin Psychopharmacol. 2014;29(6):313–7. https​://doi.
org/10.1097/YIC.00000​00000​00004​4.
12. Sockalingam S, Shammi C, Remington G. Treatment of clozapine-
induced hypersalivation with ipratropium bromide: a randomized,
double-blind, placebo-controlled crossover study. J Clin Psychia-
try. 2009;70(8):1114–9. https​://doi.org/10.4088/JCP.08m04​495.
13. Ekström J, Godoy T, Loy F, Riva A. Parasympathetic vasoactive
intestinal peptide (VIP): a likely contributor to clozapine-induced
sialorrhoea. Oral Dis. 2014;20(3):e90–6. https​://doi.org/10.1111/
odi.12139​.
K, Drašković M, Hinić D. The Glasgow antipsychotic side-
effects scale for clozapine in inpatients and outpatients with
schizophrenia or schizoaffective disorder. Nord J Psychiatry.
2018;72(2):124–9. https:​ //doi.org/10.1080/080394​ 88.2017.14000​
97.
15. Villa A, Wolff A, Narayana N, Dawes C, Aframian D, Lynge Ped-
ersen A, et al. World Workshop on Oral Medicine VI: a systematic
Treatment of Clozapine-Induced Sialorrhea: A Meta-analysis
review of medication-induced salivary gland dysfunction. Oral
Dis. 2016;22(5):365–82. https​://doi.org/10.1111/odi.12402​.
16. Reinstein MJ, Sirotovskaya LA, Chasanov MA, Jones LE, Mohan
S. Comparative efficacy and tolerability of benzatropine and tera-
zosin in the treatment of hypersalivation secondary to clozapine.
Clin Drug Investig. 1999;17(2):97–102.
17. Syed R, Au K, Cahill C, Duggan L, He Y, Udu V, et al. Phar-
macological interventions for clozapine-induced hypersaliva-
tion. Cochrane Database Syst Rev. 2008;3:Cd005579. https​://doi.
org/10.1002/14651​858.cd005​579.pub2.
18. Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, Petticrew
M, et al. Preferred reporting items for systematic review and
meta-analysis protocols (PRISMA-P) 2015 statement. Syst Rev.
2015;4:1. https​://doi.org/10.1186/2046-4053-4-1.
19. Higgins J, Green S. Cochrane handbook for systematic reviews
of interventions. The Cochrane Collaboration; 2011. http://handb​
ook.cochr​ane.org.
20. Barendregt JJ, Doi SA. MetaXL user guide version 5.3. EpiGear
International Pty Ltd; 2016.
21. Kreinin A, Novitski D, Weizman A. Amisulpride treatment of
clozapine-induced hypersalivation in schizophrenia patients: a
randomized, double-blind, placebo-controlled cross-over study.
Int Clin Psychopharmacol. 2006;21(2):99–103.
22. Kreinin A, Miodownik C, Mirkin V, Gaiduk Y, Yankovsky Y,
Bersudsky Y, et al. Double-blind, randomized, placebo-controlled
trial of metoclopramide for hypersalivation associated with clo-
zapine. J Clin Psychopharmacol. 2016. https​://doi.org/10.1097/
jcp.00000​00000​00049​3.
23. Man W, Colen-de KJ, Schulte P, Cahn W, Haelst I, Doodeman
H, et al. The effect of glycopyrrolate on nocturnal sialorrhea in
patients using clozapine: a randomized, crossover, double-blind,
placebo-controlled trial. J Clin Psychopharmacol. 2017. https​://
doi.org/10.1097/jcp.00000​00000​00065​7.
24. Xu H, Dong A, Zhang T, Shu W. Diphenhydramine for clozap-
ine -induced salivation at various dosage levels. New Drugs Clin
Rem. 1997;16(04):14–6.
25. Yang S, Shundi X, Yin Y. A double-blinded placebo-controlled
trial on the use of cyproheptadine in clozapine-induced sialor-
rhoea. Chin J Neuropsychiatr Disord. 1996;08(01):12–3.
26. Lin H. A double-blinded placebo-controlled trial on the use of
propantheline in clozapine-induced sialorrhoea. Chin J Neuropsy-
chiatr Dis. 1999;25(05):316–7.
27. Yang H, Huang S, Feng C, Chen J. A study on the treat-
ments for clozapine-induced sialorrhoea. J Binzhou Med Coll.
1999;22(05):451.
28. Gong G, Yuan L, Qin D, Li H, Zhao L, Hu C, et al. A double-
blinded placebo-controlled randomized trial on the use of diphen-
hydramine, propantheline and astemizole in clozapine-induced
sialorrhoea. Chin J Psychiatry. 1998;31(03):10.
29. Tao J, Yin X, Liu J. A randomized controlled trial on the use of
sulpiride in clozapine-induced sialorrhoea. Neurol Disord Ment
Health. 2007;07(02):141–3.
30. Qin J, Cui C, Ge F. A comparison of treatments for clozapine-
induced sialorrahoea. Chin J Clin Psychiatry. 1999;09(04):225–6.
31. Fan Z, Zheng C, Li S. A placebo-controlled trial on the use of
propantheline in clozapine-induced sialorrhoea. Strait Pharm J.
1996;08(02):33–4.
32. Cheng J, Jiang S. A study on the use of benzhexol in clozapine-
induced sialorrhoea. J Mod Chin Integr Med. 2005;14(16):2135–6.
33. Zhao Z, Chen H, Xie B, Sun S, Wu H, Ma Q, et al. A controlled
trial on the use of Zhiyan capsule in clozapine-induced sialor-
rhoea. Med J Chin Civ Admin. 2000;12(02):34–5.
34. Lu X, Xue S, Yang M. The use of diphehydramine in clozapine-
induced sialorrhoea. Sichuan J Ment Health. 1998;11(01):38–40.
35. Lu L, Liu G, Zhao J, Liu A, Liu S, Liu L, et  al. A double-
blinded placebo-controlled trial on the use of chlorpheniramine
in clozapine-induced sialorrhoea. Chin J Neuropsychiatr Dis.
1994;27(04):252–3.
36. Li Z, Song Z, Cheng P, Li C, Xiang X, Shi H. A double-blinded
controlled trial on the use of astemizole in clozapine-induced sial-
orrhoea. Chin J Clin Psychiatry. 1993;03(01):34–5.
37. Mao Z, Yang Y, Wang Y, Ma J, Cheng J. A controlled trial on
the use of traditional chinese medication Kongyan Decoction in
clozapine-induced sialorrhoea. In: The 12th national academic
conference on integrated traditional Chinese medicine and western
medicine in Psychiatry. Lanzhou, Gansu, China. 2013.
38. Qian Q, Gao Z. The use of topical Wu Dan San on acupuncture
points inclozapine-induced sialorrhoea. Shanghai Arch Psychia-
try. 1996;08(02):103.
39. Ekström J, Godoy T, Riva A. Clozapine: agonistic and antago-
nistic salivary secretory actions. J Dent Res. 2010;89(3):276–80.
https​://doi.org/10.1177/00220​34509​35605​5.
40. Rabinowitz T, Frankenburg FR, Centorrino F, Kando J. The
effect of clozapine on saliva flow rate: a pilot study. Biol Psy-
chiatry. 1996;40(11):1132–4. https​: //doi.org/10.1016/s0006​
-3223(96)89255​-9.
41. Webber MA, Szwast SJ, Steadman TM, Frazer A, Malloy FW,
Lightfoot JD, et al. Guanfacine treatment of clozapine-induced
sialorrhea. J Clin Psychopharmacol. 2004;24(6):675–6.
42. BenAryeh H, Jungerman T, Szargel R, Klein E, Laufer D. Sali-
vary flow-rate and composition in schizophrenic patients on clo-
zapine, subjective reports and laboratory data. Biol Psychiatry.
1996;39(11):946–9. https:​ //doi.org/10.1016/0006-3223(95)00296​
-0.
43. Antonello C, Tessier P. Clozapine and sialorrhea: a new interven-
tion for this bothersome and potentiallydangerous side effect. J
Psychiatry Neurosci. 1999;24(31):250.
44. Sharma A, Ramaswamy S, Dahl E, Dewan V. Intraoral applica-
tion of atropine sulfate ophthalmic solution for clozapine-induced
sialorrhea. Ann Pharmacother. 2004;38(9):1538.
45. Comley C, Galletly C, Ash D. Use of atropine eye drops for
clozapine induced hypersalivation. Aust N Z J Psychiatry.
2000;34(6):1033–4.
46. Gaftanyuk O, Trestman R. Scopolamine patch for clozapine-
induced sialorrhea. Psychiatr Serv. 2004;55(3):318.
47. McKane J, Hall C, Akram G. Hyoscine patches in clozapine-
induced hypersalivation. Psychiatr Bull. 2001;25(7):277.
48. Canaday BR, Stanford RH. Propantheline bromide in the man-
agement of hyperhidrosis associated with spinal cord injury. Ann
Pharmacother. 1995;29(5):489–92. https:​ //doi.org/10.1177/10600​
28095​02900​507.
49. Duggal H. Glycopyrrolate for clozapine-induced sialorrhea. Prog
Neuro Psychopharmacol Biol Psychiatry. 2007;31:1546–7.
50. MIMS. MIMS online. Crows Nest, N.S.W: monthly index of
medical specialities Australia; 1996.
51. Parfitt K, Martindale W. Martindale: the complete drug reference.
London: Pharmaceutical Press; 1999.
52. Matsumoto S, Yamazoe Y. Involvement of multiple human
cytochromes P450 in the liver microsomal metabolism of astemi-
zole and a comparison with terfenadine. Br J Clin Pharmacol.
2001;51(2):133–42.
53. WHO. Astemizole: voluntary withdrawal: Janssen, USA. WHO
Pharmaceuticals Newsletter. 1999.
54. Caldara R, Masci E, Cambielli M, Tittobello A, Piepoli V, Barbieri
C. Effect of sulpiride isomers on gastric acid and gastrin secretion
in healthy man. Eur J Clin Pharmacol. 1983;25(3):319–22.
55. Sockalingam S, Shammi C, Remington G. Clozapine-induced
hypersalivation: a review of treatment strategies. Can J Psychiatry.

2007;52(6):377–84. https​://doi.org/10.1177/07067​43707​05200​
607.
56. Siskind DJ, Lee M, Ravindran A, Zhang Q, Ma E, Motamarri
B, Kisely S. Augmentation strategies for clozapine refractory
schizophrenia: a systematic review and meta-analysis. Aust N Z
J Psychiatry. 2018;52:751–67.
S.-Y. Chen et al.
57. Cohen JF, Korevaar DA, Wang J, Spijker R, Bossuyt PM. Should
we search Chinese biomedical databases when performing sys-
tematic reviews? Syst Rev. 2015;4:23. https​://doi.org/10.1186/
s1364​3-015-0017-3.