Access to this full-text is provided by Springer Nature.
Content available from Malaria Journal
This content is subject to copyright. Terms and conditions apply.
Dellicour et al. Malar J (2015) 14:461
DOI 10.1186/s12936-015-0950-6
RESEARCH
Risks ofmiscarriage andinadvertent
exposure toartemisinin derivatives inthe rst
trimester ofpregnancy: a prospective cohort
study inwestern Kenya
Stephanie Dellicour1*, Meghna Desai2, George Aol3, Martina Oneko3, Peter Ouma3, Godfrey Bigogo3,
Deron C. Burton2, Robert F. Breiman4, Mary J. Hamel2, Laurence Slutsker2, Daniel Feikin2, Simon Kariuki3,
Frank Odhiambo3, Jayesh Pandit5, Kayla F. Laserson2, Greg Calip6, Andy Stergachis7 and Feiko O. ter Kuile1
Abstract
Background: The artemisinin anti-malarials are widely deployed as artemisinin-based combination therapy (ACT).
However, they are not recommended for uncomplicated malaria during the first trimester because safety data from
humans are scarce.
Methods: This was a prospective cohort study of women of child-bearing age carried out in 2011–2013, evaluating
the relationship between inadvertent ACT exposure during first trimester and miscarriage. Community-based surveil-
lance was used to identify 1134 early pregnancies. Cox proportional hazard models with left truncation were used.
Results: The risk of miscarriage among pregnancies exposed to ACT (confirmed + unconfirmed) in the first trimester,
or during the embryo-sensitive period (≥6 to <13 weeks gestation) was higher than among pregnancies unexposed
to anti-malarials in the first trimester: hazard ratio (HR) = 1.70, 95 % CI (1.08–2.68) and HR = 1.61 (0.96–2.70). For
confirmed ACT-exposures (primary analysis) the corresponding values were: HR = 1.24 (0.56–2.74) and HR = 0.73
(0.19–2.82) relative to unexposed women, and HR = 0.99 (0.12–8.33) and HR = 0.32 (0.03–3.61) relative to quinine
exposure, but the numbers of quinine exposures were very small.
Conclusion: ACT exposure in early pregnancy was more common than quinine exposure. Confirmed inadvertent
artemisinin exposure during the potential embryo-sensitive period was not associated with increased risk of miscar-
riage. Confirmatory studies are needed to rule out a smaller than three-fold increase in risk.
Keywords: Anti-malarials, Pharmacovigilance, Drug safety in pregnancy, Teratogenicity, Miscarriage
© 2015 Dellicour et al. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License
(http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium,
provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license,
and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/
publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Background
Artemisinin-based combination therapy (ACT) anti-
malarials have been adopted as first‐line treatment for
falciparum malaria in almost all endemic countries,
providing life‐saving benefits to children, adults and
pregnant women globally [1]. However, their safety is
uncertain when used in early pregnancy. Ascertainment
of risk from exposure to anti-malarials in the first trimes-
ter is difficult in resource-poor settings and data avail-
able for assessing risk are limited [2, 3]. Artemisinins
are embryo-toxic in several animal species, including
non-human primate models [4, 5]. Teratogenic effects
observed in mice and rabbits included death of the foe-
tus, malformations of the heart, great vessels, and limb
defects. Primate models exposed to prolonged courses
of ACT had high rates of foetal loss [6]. Animal mod-
els suggested that artemisinin embryo-toxicity targets
primitive erythroblasts, which are the primary form of
red blood cells in circulation between weeks 4 and 10
Open Access
*Correspondence: Stephanie.Dellicour@lstmed.ac.uk
1 Liverpool School of Tropical Medicine, Pembroke Place, Liverpool L3
5QA, UK
Full list of author information is available at the end of the article
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 2 of 9
Dellicour et al. Malar J (2015) 14:461
post-conception in humans. erefore the embryo-sen-
sitive period to artemisinin, if any, is thought to occur at
6–12 (inclusive) weeks’ gestation from the first day of the
last menstrual period (LMP) in humans [4, 5, 7, 8].
ere are limited data available to assess whether ACT
is embryo-toxic or teratogenic in humans; fewer than
700 exposures in the first trimester have been well doc-
umented [9–15]. After reviewing all existing evidence in
2003 and then in 2006, the World Health Organization
(WHO) recommended that artemisinins could be used
during the second or third trimesters of pregnancy and
that, due to insufficient safety data, treatment in the first
trimester was not recommended unless the life of the
mother is at risk, or oral quinine is not available [5, 16].
e recommended treatment for first trimester malaria
infections is seven days’ oral quinine alone or combined
with clindamycin [17]. However, as women may not be
aware of their pregnancy or do not declare an early preg-
nancy, and because clinic staff do not often assess for
pregnancy in women of child-bearing age (WOCBA), the
risk of exposure to drugs not recommended in pregnancy,
including to potential teratogens, is possible during this
period [18]. As ACT is increasingly available, a growing
number of women will be inadvertently exposed to an
artemisinin compound in early pregnancy, including dur-
ing the period when foetal organs and tissues are formed.
Malaria can have severe consequences to the health
of the pregnant woman and her unborn baby including
maternal anaemia, foetal loss, preterm birth, low birth
weight and perinatal mortality, and in some cases mater-
nal death. e impact of malaria infection in early preg-
nancy has been identified as a major knowledge gap for
estimating the burden of malaria in pregnancy. Recent
studies have provided insight into the potential adverse
consequences of malaria infections early in pregnancy,
showing a major impact on birth weight and maternal
anaemia [19, 20]. Findings from a retrospective analysis
from 25 years of data from the ai-Myanmar border,
where artemisinin deployment has been necessary for
many years because of multi-drug resistance, showed
that malaria infection in the first trimester (both sympto-
matic and asymptomatic) was a significant risk factor for
miscarriage. No association between first trimester arte-
misinin exposure and miscarriage was found. However
more data from a wider range of malaria-endemic coun-
tries are required to provide an increased level of reas-
surance that first trimester artemisinin exposure does
not significantly increase the risk of miscarriage or other
adverse pregnancy outcomes. e findings from a pro-
spective cohort study of WOCBA designed to examine
whether ACT exposure in the first trimester was associ-
ated with miscarriage are reported here.
Methods
Overview ofstudy design
is was a prospective cohort study conducted among
WOCBA (15–49 years of age) residing in a highly
malarious area in western Kenya with a population
under continuous health and demographic surveillance
system (HDSS) monitoring as part of the collabora-
tion between the Kenya Medical Research Institute
(KEMRI) and Centers for Disease Control and Pre-
vention (CDC) [21]. Participants received treatment
through the usual channels, including health facilities
and drug outlets.
Procedures
Recruitment ofwomen ofchild‑bearing age
andpregnancy detection
Between 15 February, 2011 and 15 February, 2013,
6010 WOCBA participating in an ongoing population-
based, infectious disease, surveillance project (PBIDS)
in rural Bondo District, western Kenya [22, 23] (Addi-
tional file1) were invited to participate in the ‘Evalua-
tion of Medications used in Early Pregnancy’ (EMEP)
prospective cohort study. EMEP staff visited all homes
in the PBIDS and enrolled consenting WOCBA who
met eligibility criteria for EMEP. WOCBA were eligi-
ble for EMEP if they were between 15 and 49years of
age and active participants of PBIDS. Exclusion cri-
teria included: inability to give informed consent or
provide an accurate medical history. WOCBA who
consented to participate were asked if they could be
pregnant and offered a pregnancy test at the time of
enrolment and again approximately every 3 months
thereafter. Any participant with a detected pregnancy
was referred to the antenatal clinic at Lwak Hospital
where trained EMEP nurses confirmed the pregnancy
(either by ultrasound if the women presented before
24weeks, or by palpation and by auscultation of the
foetal heart later in pregnancy) and offered free ante-
natal care (ANC). Additionally, all pregnant patients
presenting at the ANC clinic of Lwak Hospital were
enrolled if all criteria were met. EMEP nurses were not
involved in treatment of study participants.
Gestational age assessment
Gestational age was determined using the most accu-
rate measurement available for each participant in the
following order: ultrasound scan taken before 24weeks’
gestation performed by trained study nurses (Sonosite
180 plus portable ultrasound system), Ballard estimates
measured within 96h of birth, LMP or reported gesta-
tion at time of pregnancy loss, and, lastly gestational age
derived from fundal height assessment (Additional file1).
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 3 of 9
Dellicour et al. Malar J (2015) 14:461
Pregnancy outcome
Pregnancy outcomes were assessed using a combina-
tion of health facility- and home-based follow-up vis-
its. e latter is particularly relevant for miscarriages,
because the vast majority of these events occur in the
community, not in health facilities. Village-based staff
received monthly lists of participants with estimated
delivery dates and after visiting the participants’ homes
they informed study nurses of pregnancy outcomes. Fol-
low-ups by study staff were then arranged to administer
structured questionnaires about the delivery, outcome,
any illnesses and medication used during pregnancy.
Pregnancy outcomes captured included: pregnancy loss
(miscarriages, induced abortions and stillbirths), live
births and major congenital malformations detectable
at birth by surface examination. is analysis focuses on
miscarriage defined as spontaneous pregnancy loss at or
before 28 completed weeks’ gestation (2–28weeks inclu-
sive), which is considered the gestational age of viability
in resource-constrained settings.
Anti‑malarial drug exposure ascertainment
Drug exposure data were captured using three
approaches (Table 1): (a) interviews with pregnant
women visiting the antenatal clinic in Lwak Hospital and
at the time of pregnancy outcome follow-up (henceforth
referred to as EMEP data); (b) record linkage to data on
drugs prescribed to WOCBA at the outpatient depart-
ment in Lwak Hospital (henceforth referred to as Lwak-
OPD data); and, (c) weekly to twice monthly home visits
by fieldworkers as part of PBIDS.
Other covariates
Obstetric history and ANC laboratory information col-
lected routinely at antenatal booking (haemoglobin
level, HIV and syphilis testing, and malaria microscopy)
were extracted from the ANC records at Lwak hospital
or antenatal cards by study nurses. Demographic char-
acteristics and medical history, including illnesses (e.g.,
malaria) and drugs used during the current pregnancy
were collected at each EMEP study visit at ANC and dur-
ing pregnancy outcome follow-up visits. Household level
wealth quintiles were obtained from the HDSS [24].
Data analysis
Exposure denition
A trend of increase in risk of miscarriage with ACT expo-
sure during this artemisinin-specific, embryo-sensitive
period would corroborate the biological mechanism
observed in animal models and suggest a causal associa-
tion with ACT exposures. e analysis focused on two
exposure definitions: anti-malarial drug reported/pre-
scribed (1) ‘anytime’ in the first trimester, i.e., gestational
week 2 and 0days (day 14 since LMP) to week 13 and
6days (day 97 since LMP) post-LMP, and (2) between
weeks 6 day 0 (day 42 since LMP) to week 12 day 6
post-LMP (day 90 since LMP) (potential artemisinin
embryo-sensitive period as suggested by animal repro-
toxicology [8]). Unexposed was defined as no evidence
of anti-malarial or malaria exposure in any of the three
data sources. Confirmed exposures were defined as expo-
sures identified by at least two of the three data sources.
Confirmed + unconfirmed exposures were defined as
Table 1 Description ofdrug information sources used todetermine anti-malarial andmalaria exposure status
ANC antenatal care, EMEP evaluation of medications used in early pregnancy study, Lwak OPD Lwak hospital out-patient department, PBIDS population-based
infectious disease surveillance
Approach Format Drug information available
EMEP self-report Retrospective self-report of illness and medication used since
the beginning of the pregnancy collected at every ANC visit
and at pregnancy outcome follow-up visit. A general open
question about any drug use as well as a directed question
for specific anti-malarials were included as using medication/
indication-specific questions have been shown to improve
accuracy. Photographs of all anti-malarial drugs found in the
study area were used to facilitate recognition of drug names. A
calendar marking public holidays and school closures was also
used to enhance recall of dates
Drug name
Drug start date
Duration
Number of tablets per day
Indication and indication diagnosis
Drug source
Lwak-OPD records Prospective documentation by health facility clinic staff of
diagnosis and treatment prescribed at outpatient department
(OPD) whenever a PBIDS participant sought care at Lwak
Hospital for an infectious syndrome
Date of visit
Diagnosis
Prescribed treatment
PBIDS weekly and twice-monthly
home visits Self-report of symptoms, health-seeking behaviour and medica-
tion. This information was collected continuously on a weekly
(from 5 January, 2010 to 26 May, 2011) and then twice-
monthly basis (27 May, 2011 onwards). The same visual aids as
described above were used for recall of drug intake
Date of visit
Symptoms in previous week/2 weeks
Treatment taken for the symptoms including
drug name
If and where care was sought
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 4 of 9
Dellicour et al. Malar J (2015) 14:461
anti-malarial identified by at least one of the three data
sources.
Cox regression model
Analyses were performed using Stata v12.1 (StataCorp
LP, College Station, TX, USA). Cox proportional haz-
ard regression models with left truncation were fitted
to estimate the effect on miscarriage of ACT exposure
during the first trimester and during the artemisinin
embryo-sensitive period. Exposure was treated as a time-
dependent variable (Additional file 1). Known risk fac-
tors for miscarriage were considered and to determine
which variables remained in the final model, assessment
of confounding was based on the impact a variable had
on the hazard ratio, followed by the consideration of its
precision. If the HR changed by ≥10% the variable was
retained in the model [25, 26].
e primary analysis compared the hazard of miscar-
riage among pregnancies with confirmed ACT expo-
sures, either anytime during the first trimester or six to
12weeks post-LMP, with the hazard among women not
exposed to any anti-malarials anytime during the first tri-
mester or among women exposed to quinine anytime in
the first trimester or 6–12weeks post-LMP.
Secondary analyses consisted of similar models but
using (a) less restrictive exposure definitions, including
both confirmed and unconfirmed exposures, and, (b)
more restrictive exposure definitions where only ACT
exposures within estimated gestational age confidence
margins were included (Additional file2).
Ethical review andconsent
e EMEP study was approved by the ethics commit-
tees and institutional review boards of CDC (No. 5889),
KEMRI (No. 1752) and the Liverpool School of Tropi-
cal Medicine (No. 09.70). Written informed consent
or assent was obtained from each participant includ-
ing consent for record linkage with PBIDS and HDSS
databases.
Results
Participant characteristics
Out of 5911 eligible WOCBA, 5536 (94 %) consented
to participate and among them, 1453 pregnancies were
detected, and 1134 (78%) were included in the data anal-
ysis (Fig.1). e mean andmedian gestational age at time
of pregnancy detection was 13.3 (standard deviation 6.9)
and 12.1 (range 0–27.9) weeks (Table2). Overall, 62% of
deliveries took place at a health facility, and 25% of the
miscarriages. Overall, 67% of pregnancy outcomes were
captured within a week of the event; however, for miscar-
riages this was only 20%.
Prevalence ofrst trimester ACT andquinine exposure
Overall, 299 (26.4 %) of the 1134 pregnancies had evi-
dence of possible ACT exposure anytime in the first tri-
mester (confirmed + unconfirmed). For 77 (25.8 % of
exposures and 6.8% of all pregnancies) this could be con-
firmed by at least two of the three sources; 56 of these
confirmed exposures (18.7, 5.3 % of pregnancies) were
within the estimated gestational age confidence mar-
gins. For 212 out of 299 first trimester exposures (70.9%,
18.7% of pregnancies), the exposure occurred between
6 and 12weeks’ gestation; 47 of them were confirmed
exposures (Fig.2). Only 13 pregnancies were exposed to
quinine-alone anytime in the first trimester, and 11 dur-
ing the 6–12weeks’ gestational period.
Association betweenrst trimester ACT‑exposure
andmiscarriage
Conrmed exposure (primary analysis)
Compared to pregnancies without anti-malarial expo-
sure/malaria in the first trimester (793), the hazard
for miscarriage was non-significantly higher among
women with confirmed ACT exposures anytime in the
first trimester (77) [hazard ratio (HR) =1.24, 95% CI
(0.56–2.73)], and this was HR=1.72 (0.66–4.45) in mul-
tivariate analysis (Fig. 2). e corresponding values for
ACT exposure during the embryo-sensitive period (47)
were HR=0.73 (0.19–2.82) and HR=0.81 (0.21–3.03)
(Fig.2).
e values when compared against quinine (13) were:
HR=0.99 (0.12–8.33) and HR=0. 32 (0.03–3.61) (crude
analysis) for exposure anytime and six to 12weeks post-
LMP (Fig.2).
More restrictive definitions to define exposure within
the redefined margins for gestational age resulted in simi-
lar or lower effect estimates, but numbers of exposures
and events were limited (Additional file2). e method
used for missing value did not alter the conclusions
(Additional file3).
Conrmed+unconrmed exposure (secondary analysis)
When using a less restrictive definition of exposure by
including unconfirmed exposures as well, the risk of
miscarriages was significantly higher among the ACT-
exposed pregnancies relative to unexposed pregnan-
cies: adjusted HR=1.66 (1.04–2.67). is was HR=1.
61 (0.96–2.70) for the embryo-sensitive period. e HRs
when compared to quinine were HR=0.64 (0.08–4.91)
and HR=0.46 (0.05–4.44), respectively (Fig.2).
Discussion
Pregnancies exposed to ACT in the first trimester were
at increased risk of miscarriage compared to pregnancies
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 5 of 9
Dellicour et al. Malar J (2015) 14:461
not exposed to anti-malarials in the same gestation
period. is was only statistically significant at the 5%
level in the group with the less restrictive definition for
exposure (confirmed and unconfirmed) which had higher
number of events (29) and exposures (299) [adjusted
HR=1.66 95% CI (1.04–2.67)]. A similar effect measure
Analysis
Assessed for eligibility
N=6010
Excluded (n=474)
♦Not meeting inclusion criteria (n=99)
♦Declined to participate (n=375)
Consented (n=5536)
Pregnancies(n=1453)
Number of pregnancies per participants:
♦Single pregnancy (n=1266)
♦Two pregnancies(n=92)
♦Three pregnancies (n=1)
Follow-Up
Enrolment
Loss to follow-up(n=85)
♦Migrated (n=67)
♦Withdrawal or refused follow-up(n=13)
♦Maternal death (n=5)
Pregnancy Outcomes (n=1368)
Pregnancies included
(n=1134)
Excluded from analysis (n=319)
♦Detected at outcome(n=33)
♦Entered after 28 weeks (n=219)
♦No GA information (n=21)
♦Pregnancy end date error (n=5)
♦No follow-up (n=41)
Fig. 1 Study participant flow diagram from screening to inclusion in data analysis
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 6 of 9
Dellicour et al. Malar J (2015) 14:461
[HR=1.72 95% CI (0.66–4.45)] was obtained when the
analysis was restricted to those exposures that could be
confirmed by the OPD database or the ongoing house-
hold surveillance, which was the primary analysis. How-
ever the available exposures (77) and events (six) were
reduced markedly with this more restrictive analysis and
the difference was not statistically significant. When the
analysis was further restricted to exposures in the poten-
tial embryo-sensitive period in humans for the arte-
misinins, the effect estimates were again similar [HR=1.
61 95% CI (0.96, 2.70)] for confirmed + unconfirmed
exposures, but much lower for confirmed exposures
[HR = 0.73 95 % CI (0.19, 2.82)]. However this latter
analysis, which was also part of the primary analysis,
included only two events and 47 pregnancies exposed to
ACT. ere was no evidence for an increase in the risk
of miscarriage among women treated with ACT versus
women treated with oral quinine, but again the number
exposed to quinine alone was limited to 13 with only one
miscarriage.
It was expected that the risk of miscarriage would
be higher among women who received anti-malarials
than among women without anti-malarial exposure
early in pregnancy. is is related to the potential for
Table 2 Characteristics of1134 pregnancies byACT exposure status [n (%) otherwise stated]
ACT artemisinin combination therapy, SD standard deviation
*P values refer to Pearson Chi square test for categorical variables and ANOVA test for continuous variables
a Gestational age lowest estimate include 0 which reects inaccuracy in the gestational age measurements
b HIV status information was not available for 12% (129) of pregnancies that did not attend antenatal care or have the antenatal card for review. HIV status
information was complemented by HDSS and data which oered home-based HIV testing and counselling to PBIDS participants. Test results were linked to the
study participants using unique ID and missing data were updated if the test was performed before the pregnancy detection for HIV positive test results and for HIV
negative results if the test was performed maximum 3months before or after pregnancy detection. An additional 30 HIV status were ascertained while 8% (99) still
had no HIV status data
Overall
(N=1134) No ACT exposure
inthe rst trimester
(N=835)
Unconrmed ACT
exposure inthe rst
trimester (N=222)
Conrmed ACT
exposure inthe rst
trimester (N=77)
P values*
Age in years [mean (SD;
range)] 26.1 (6.8; 15–47) 26.1 (6.7; 15–45) 26.7 (7.2; 15–47) 25.2 (6.5; 16–41) 0.225
Gravidity Missing n = 16 Missing n = 14 Missing n = 1 Missing n = 1 0.065
Primigravidae 219 (19.6) 151 (18.4) 47 (21.3) 21 (27.6)
1–3 pregnancies 525 (47.0) 405 (49.3) 90 (40.7) 30 (39.5)
4+ pregnancies 374 (33.5) 265 (32.3) 84 (38.0) 25 (32.9)
Previous pregnancy loss 160 (14.3), Missing n = 17 118 (14.4), Missing n = 15 30 (13.6), Missing n = 1 12 (15.8), Missing n = 1 0.888
Gestational age at detec-
tion in weeks [mean (SD;
range)]a
13.3 (6.9; 0–27.9) 13.3 (7.0; 0–27.9) 13.0 (6.7; 0.3–27) 13.6 (7.1; 2.4–27.4) 0.770
Occupation Missing n = 31 Missing n = 28 Missing n = 1 Missing n = 2 0.191
Not working 379 (34.4) 281 (34.8) 68 (30.8) 30 (40.0)
Farming 369 (33.5) 268 (33.2) 80 (36.2) 21 (28.0)
Small business/Skilled
Labour 335 (30.4) 246 (30.5) 65 (29.4) 24 (32.0)
Other 20 (1.8) 12 (1.5) 8 (3.6) 0
Antenatal care summary
Number of ANC visit Missing n = 39 Missing n = 31 Missing n = 5 Missing n = 3 0.125
None 89 (8.1) 64 (8.0) 21 (9.7) 4 (5.4)
1 90 (8.2) 61 (7.6) 24 (11.1) 5 (6.8)
2 155 (14.2) 121 (15.1) 25 (11.5) 9 (12.2)
3 244 (22.3) 193 (24.0) 38 (17.5) 13 (17.6)
4+517 (47.2) 365 (45.4) 109 (50.2) 43 (58.1)
Gestational age at first
ANC visit in weeks
[mean (SD)]*
20.8 (7.8) range: 1.7–41.0 21.24 (7.8) range: 2.7–41.0 19.7 (7.6) range: 1.7–41.0 19.4 (7.7) range: 3.4-37.0 0.020
HIV positivebMissing n = 101 Missing n = 79 Missing n = 18 Missing n = 4 0.354
Negative 771 (74.4) 562 (74.3) 149 (73.0) 60 (82.2)
Positive 262 (25.4) 194 (25.7) 55 (27.0) 13 (17.8)
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 7 of 9
Dellicour et al. Malar J (2015) 14:461
confounding by indication, i.e., women treated with ACT
or quinine sought treatment because of their malaria
or other febrile illness, whereas women who did not
require anti-malarials did not. e comparison with
untreated women is therefore difficult to interpret as it
does not allow for the differentiation between the effects
of malaria and the drug treating it. Malaria itself, even
if it remains asymptomatic, is a known cause of miscar-
riage. A recent meta-analysis of five trials with malaria
chemoprophylaxis or intermittent preventive therapy in
2876 paucigravidae in sub-Saharan Africa showed that
women in the control arms were at a 1.54 95% CI (0.98–
2.44) higher risk of miscarrying than women protected
by chemoprevention [27]. Prospective studies in low
malaria-transmission areas in ailand also found that
asymptomatic malaria in the first trimester increased the
odds of miscarriage nearly three-fold and symptomatic
infections four-fold [13]. e 1.4- to 1.7-fold increased
risk for miscarriage among women exposed to ACT or
quinine relative to pregnancies not requiring treatment
observed in this study is thus within the expected range
of malaria-associated risk of miscarriage.
is study is underpowered to confidently detect or
exclude effects smaller than a three-fold increased risk of
miscarriage associated with ACT. Nevertheless no indi-
cation for such a potential association was found. First,
there was no indication that the effect size associated
with ACT exposure relative to unexposed women was
greater among women treated during the embryo-sensi-
tive period than at anytime during the first trimester. If
ACT was causing miscarriage through this mechanism,
the effect size would be expected to be highest for expo-
sures restricted to that embryo-sensitive period. No such
trend was observed. Secondly, the rates of miscarriage
in the quinine-only and ACT-exposed pregnancies were
similar. Although the comparison with quinine needs to
be interpreted with caution due to the small numbers of
quinine-only exposed women, these results are consist-
ent with observations from the ai-Burmese border by
McGready etal. ey also found no difference in the pro-
portions of pregnancies ending in miscarriages between
women treated with chloroquine (26%), quinine (27%)
or artesunate (31 %) [13]. A recent prospective study
from Tanzania reported higher risk of pregnancy loss
(miscarriage and stillbirth combined) in women exposed
to quinine compared to those exposed to ACT [14]. A
prospective study in Zambia found higher occurrence of
miscarriage in first trimester ACT-exposed pregnancies
(5%) compared to none in those exposed to sulfadox-
ine-pyrimethamine or quinine but the number exposed
to quinine (six) were too small to allow for a meaningful
comparison [12].
e small number of quinine exposures in the first
trimester in this study was surprising as this is the
Fig. 2 Miscarriage rate, unadjusted and adjusted hazard rates for the association between different anti-malarial exposure categories and miscar-
riage
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 8 of 9
Dellicour et al. Malar J (2015) 14:461
recommended first-line malaria treatment in the first tri-
mester. However these observations are consistent with
a recent study on malaria in pregnancy-prescribing prac-
tice carried out in the same area of western Kenya (Riley
etal., unpublished) and a study from Uganda [28]. ese
studies draw attention to the need to assess reasons for
poor adherence to quinine and malaria treatment guide-
lines. Poor tolerability and poor compliance to its seven-
day regimen is a known problem for treatment of malaria
with oral quinine [29, 30].
is study had several limitations that should be con-
sidered. First, the small number of quinine exposures
limited the ability to compare ACT-exposed pregnancies
to the purported ‘control’ drug (as quinine is not known
to cause miscarriages) [3]. Second, it was not possible to
control for confounding by indication (i.e., the disease
itself) because laboratory confirmation of malaria was
not available for most women. Controlling for malaria
and its severity is important, as malaria itself has been
suggested to reduce the potential risk of embryo-tox-
icity from artemisinin as was found in rat models [31].
ird, since induced abortions are illegal in Kenya, this
could have resulted in induced abortions being reported
as miscarriages. However since neither ACT nor qui-
nine exposures are perceived as indications for induced
abortion in this population, it is thus unlikely that such
misclassification would differ according to exposure sta-
tus. Fourth, it was not possible to account for exposure
misclassification due to lack of adherence to prescribed
medication (drug intake was not observed) or from coun-
terfeit anti-malarials [32], which could bias the estimate
towards the null. Fifth, the ability to confirm exposure
was limited because there was limited overlap in the
exposures ascertained in the three data sources. e
group at highest risk for bias are the unconfirmed expo-
sure cases as 32 first-trimester, ACT-exposures were only
reported after pregnancy outcome. Recall bias following
adverse pregnancy outcome has been well documented,
hence the focus in this study was to confirm ACT expo-
sures using prospective drug ascertainment approaches
through record linkage to minimize such bias [33–35].
Another potential source of exposure misclassification is
gestational age measurement errors. e study could not
assess any dose–response effect of exposure.
Conclusion
e results presented here are consistent with two pre-
vious observational studies showing an increased risk of
miscarriage among women treated for malaria with ACT
in the first trimester versus unexposed women, and a
similar [13] or lower risk compared to oral quinine [14].
ese results also suggest that ACT use in the first tri-
mester is much more common than quinine. e risk
associated with malaria in early pregnancy, the compara-
ble observed risk between ACT and quinine exposures,
and the limited compliance to treatment with quinine
suggests a trial comparing ACT versus quinine for the
treatment of uncomplicated malaria in the first trimes-
ter may be merited. Before such a trial is considered,
further safety data on the association between ACT and
congenital malformations, that is forthcoming from stud-
ies conducted by the Malaria in Pregnancy Consortium
and WHO, should be reviewed and all available evidence
pooled to evaluate the evidence of the risk and benefits of
artemisinin use in early pregnancy.
Abbreviations
ACT: artemisinin-based combination therapy; ANC: antenatal care; CDC: US
Centers for Disease Control and Prevention; EMEP: evaluation of medications
used in early pregnancy study; HDSS: health and demographic surveillance
system; HR: hazard ratio; KEMRI: Kenya Medical Research Institute; LMP: last
menstrual period; OPD: outpatient department; PBIDS: population-based
infectious disease surveillance project; WHO: World Health Organization;
WOCBA: women of childbearing age.
Authors’ contributions
SD, FtK, AS, LS, and MJH conceived and designed the experiments. SD, GA, PO,
MO, and GB conducted field work. SD and GC analysed the data. GB, DF, RFB,
SK, DCB, FO, and FtK contributed data/analysis tools. SD, DCB, RFB, MJH, LS, DF,
SK, KL, AS, MD, and FtK interpreted the data. JP acted as Government liaison
and the Kenyan regulator. SD, FtK and MD rote the first draft of the manu-
script. All authors read and approved the final manuscript.
Author details
1 Liverpool School of Tropical Medicine, Pembroke Place, Liverpool L3 5QA,
UK. 2 Centers for Disease Control and Prevention, Atlanta, GA, USA. 3 Kenya
Medical Research Institute Centre for Global Health Research, Kisumu, Kenya.
4 Global Health Institute, Emory University, Atlanta, GA, USA. 5 Bayer Health-
care, Nairobi, Kenya. 6 Pharmacy Systems, Outcomes and Policy Department,
University of Illinois at Chicago, Chicago, USA. 7 Departments of Pharmacy
and Global Health, Schools of Pharmacy and Public Health, University of Wash-
ington, Seattle, USA.
Acknowledgements
The work presented in this paper was performed under the KEMRI and CDC
Collaboration in western Kenya. We are very grateful to all participants for
taking part in the study. We wish to thank the EMEP study team for their
perseverance and hard work. We are grateful to the International Emerging
Infection Program (IEIP) team for their help and collaboration. Furthermore
we thank the Asembo District health and medical team and the Lwak Mission
Hospital Board for their support. We also thank John Williamson and Jane
Bruce for the statistical support and advice. KEMRI/CDC HDSS is a member of
the INDEPTH Network. The findings and conclusions in this paper are those of
the authors and do not necessarily represent the views of the US Centers for
Disease Control and Prevention. This paper is published with the permission
of KEMRI Director. This work was partly supported by the Malaria in Pregnancy
(MiP) Consortium, which is funded through a grant from the Bill and Melinda
Gates Foundation to the Liverpool School of Tropical Medicine, UK and partly
Additional les
Additional le 1. Detailed description of study site and methodology.
Additional le 2. Description of sensitivity analysis looking at potential
gestational age measurement error.
Additional le 3. Description of sensitivity analysis using multiple impu-
tation for missing data.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Page 9 of 9
Dellicour et al. Malar J (2015) 14:461
by the US Centers for Disease Control and Prevention (CDC), Division of
Parasitic Diseases and Malaria through a cooperative agreement with Kenya
Medical Research Institute (KEMRI), Center for Global Health Research (CGHR),
Kisumu, Kenya. The funders had no role in study design, data collection and
analysis, decision to publish, or preparation of the manuscript.
Competing interests
The authors declare that they have no competing interests.
Received: 11 August 2015 Accepted: 21 October 2015
References
1. WHO. World Malaria Report: 2013. Geneva: World Health Organization
2013.
2. Phillips-Howard PA, Steffen R, Kerr L, Vanhauwere B, Schildknecht J, Fuchs
E, et al. Safety of mefloquine and other antimalarial agents in the first
trimester of pregnancy. J Travel Med. 1998;5:121–6.
3. Phillips-Howard PA, Wood D. The safety of antimalarial drugs in preg-
nancy. Drug Saf. 1996;14:131–45.
4. Clark RL. Embryotoxicity of the artemisinin antimalarials and potential
consequences for use in women in the first trimester. Reprod Toxicol.
2009;28:285–96.
5. WHO, TDR. Assessment of the safety of artemisinin compounds in
pregnancy: report of two joint informal consultations convened in 2006.
Geneva: World Health Organization 2007.
6. Clark R, Kumemura M, Makori N, Nakata Y, Bernard F, Harrell A, et al.
Artesunate: developmental toxicity in monkeys. Bir Birth Defects
Research (Part A). 2006;76:329.
7. Kelemen E, Calvo W, Fliedner T. Atlas of Human Hemopoietic Develop-
ment. Berlin: 1979.
8. White TE, Clark RL. Sensitive periods for developmental toxicity of orally
administered artesunate in the rat. Birth Defects Res B Dev Reprod Toxi-
col. 2008;83:407–17.
9. Adam I, Elhassan EM, Omer EM, Abdulla MA, Mahgoub HM, Adam GK.
Safety of artemisinins during early pregnancy, assessed in 62 Sudanese
women. Ann Trop Med Parasitol. 2009;103:205–10.
10. Deen JL, von Seidlein L, Pinder M, Walraven GE, Greenwood BM. The
safety of the combination artesunate and pyrimethamine-sulfadoxine
given during pregnancy. Trans R Soc Trop Med Hyg. 2001;95:424–8.
11. Dellicour S, Hall S, Chandramohan D, Greenwood B. The safety of arte-
misinins during pregnancy: a pressing question. Malar J. 2007;6:15.
12. Manyando C, Mkandawire R, Puma L, Sinkala M, Mpabalwani E, Njunju E,
et al. Safety of artemether-lumefantrine in pregnant women with malaria:
results of a prospective cohort study in Zambia. Malar J. 2010;9:249.
13. McGready R, Lee SJ, Wiladphaingern J, Ashley EA, Rijken MJ, Boel M, et al.
Adverse effects of falciparum and vivax malaria and the safety of anti-
malarial treatment in early pregnancy: a population-based study. Lancet
Infect Dis. 2012;12:388–96.
14. Mosha D, Mazuguni F, Mrema S, Sevene E, Abdulla S, Genton B. Safety
of artemether-lumefantrine exposure in first trimester of pregnancy: an
observational cohort. Malar J. 2014;13:197.
15. Rulisa S, Kaligirwa N, Agaba S, Karema C, Mens PF, de Vries PJ. Pharma-
covigilance of artemether-lumefantrine in pregnant women followed
until delivery in Rwanda. Malar J. 2012;11:225.
16. WHO. Assessment of the safety of artemisinin compounds in pregnancy.
Report of two informal consultations convened by WHO in 2002. Geneva:
World Health Organization; 2003.
17. WHO. Guidelines for the treatment of malaria. Third edition. Geneva:
World Health Organization; 2015.
18. Dellicour S, ter Kuile FO, Stergachis A. Pregnancy exposure registries for
assessing antimalarial drug safety in pregnancy in malaria-endemic coun-
tries. PLoS Med. 2008;5:e187.
19. Huynh BT, Fievet N, Gbaguidi G, Dechavanne S, Borgella S, Guezo-Mevo
B, et al. Influence of the timing of malaria infection during pregnancy
on birth weight and on maternal anemia in Benin. Am J Trop Med Hyg.
2011;85:214–20.
20. Cottrell G, Mary JY, Barro D, Cot M. The importance of the period of
malarial infection during pregnancy on birth weight in tropical Africa. Am
J Trop Med Hyg. 2007;76:849–54.
21. Odhiambo FO, Laserson KF, Sewe M, Hamel MJ, Feikin DR, Adazu K, et al.
Profile: the KEMRI/CDC Health and Demographic Surveillance System-
Western Kenya. Int J Epidemiol. 2012;41:977–87.
22. Bigogo G, Audi A, Aura B, Aol G, Breiman RF, Feikin DR. Health-seeking
patterns among participants of population-based morbidity surveillance
in rural western Kenya: implications for calculating disease rates. Int J
Infect Dis. 2010;14:e967–73.
23. Feikin DR, Audi A, Olack B, Bigogo GM, Polyak C, Burke H, et al. Evalua-
tion of the optimal recall period for disease symptoms in home-based
morbidity surveillance in rural and urban Kenya. Int J Epidemiol.
2010;39:450–8.
24. McKenzie D. Measuring inequality with asset indicators. Journal of Popu-
lation Economics. 2005;18:229.
25. Sonis J. A closer look at confounding. Fam Med. 1998;30:584–8.
26. Kleinbaum DG, Klein M. Chapter 6: Modeling Strategy Guidelines. Logistic
Regression: A Self-Learning Text. 3rd edition ed.: Springer; 2010.
27. Radeva-Petrova D, Kayentao K, ter Kuile FO, Sinclair D, Garner P. Drugs
for preventing malaria in pregnant women in endemic areas: any drug
regimen versus placebo or no treatment. Cochrane Database Syst Rev.
2014;10:CD000169.
28. Sangaré L, Weiss N, Brentlinger P, Richardson B, Staedke S, Kiwuwa M,
et al. Patterns of antimalarial drug treatment among pregnant women in
Uganda. Malar J. 2011;10:152.
29. Achan J, Talisuna AO, Erhart A, Yeka A, Tibenderana JK, Baliraine FN, et al.
Quinine, an old anti-malarial drug in a modern world: role in the treat-
ment of malaria. Malar J. 2011;10:144.
30. Piola P, Nabasumba C, Turyakira E, Dhorda M, Lindegardh N, Nyehangane
D, et al. Efficacy and safety of artemether-lumefantrine compared with
quinine in pregnant women with uncomplicated Plasmodium falciparum
malaria: an open-label, randomised, non-inferiority trial. Lancet Infect Dis.
2010;10:762–9.
31. Clark RL. Effects of artemisinins on reticulocyte count and relationship to
possible embryotoxicity in confirmed and unconfirmed malarial patients.
Birth Defects Res A Clin Mol Teratol. 2012;94:61–75.
32. Newton PN, Green MD, Mildenhall DC, Plancon A, Nettey H, Nyadong
L, et al. Poor quality vital anti-malarials in Africa—an urgent neglected
public health priority. Malar J. 2011;10:352.
33. Jong De, van den Berg LT, Feenstra N, Sorensen HT, Cornel MC. Improve-
ment of drug exposure data in a registration of congenital anomalies.
Pilot-study: pharmacist and mother as sources for drug exposure data
during pregnancy. EuroMAP Group. Europen Medicine and Pregnancy
Group. Teratology. 1999;60:33–6.
34. Rockenbauer M, Olsen J, Czeizel AE, Pedersen L, Sorensen HT, Euro MAPG.
Recall bias in a case-control surveillance system on the use of medicine
during pregnancy. Epidemiology. 2001;12:461–6.
35. de Jong-van den Berg LT, Waardenburg CM, Haaijer-Ruskamp FM, Dukes
MN, Wesseling H. Drug use in pregnancy: a comparative appraisal of data
collecting methods. Eur J Clin Pharmacol. 1993;45:9–14.
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com
Content uploaded by Feiko O Ter Kuile
Author content
All content in this area was uploaded by Feiko O Ter Kuile on Nov 22, 2015
Content may be subject to copyright.