Menu
Menu
Menu
Menu
Background Pregnancy increases the risk of harmful effects from cholera for both mothers and their fetuses. A killed oral cholera vaccine, Shanchol (Shantha Biotechnics, Hydrabad, India), can protect against the disease for up to 5 years. However, cholera vaccination campaigns have often excluded pregnant women because of insufficient safety data for use during pregnancy. We did an observational cohort study to assess the safety of Shanchol during pregnancy. Methods This observational cohort study was done in two adjacent districts (Nsanje and Chikwawa) in Malawi. Individuals older than 1 year in Nsanje were offered oral cholera vaccine during a mass vaccination campaign between March 30 and April 30, 2015, but no vaccines were administered in Chikwawa. We enrolled women who were exposed to oral cholera vaccine during pregnancy in Nsanje district, and women who were pregnant in Chikwawa district (and thus not exposed to oral cholera vaccine) during the same period. The primary endpoint of our analysis was pregnancy loss (spontaneous miscarriage or stillbirth), and the secondary endpoints were neonatal deaths and malformations. We evaluated these endpoints using log-binomial regression, adjusting for the imbalanced baseline characteristics between the groups. This study is registered with ClinicalTrials.gov, number NCT02499172. Findings We recruited 900 women exposed to oral cholera vaccine and 899 women not exposed to the vaccine between June 16 and Oct 10, 2015, and analysed 835 in each group. 361 women exposed to the vaccine and 327 not exposed to the vaccine were recruited after their pregnancies had ended. The incidence of pregnancy loss was 27·54 (95% CI 18·41–41·23) per 1000 pregnancies among those exposed to the vaccine and 21·56 (13·65–34·04) per 1000 among those not exposed. The adjusted relative risk for pregnancy loss among those exposed to oral cholera vaccine was 1·24 (95% CI 0·64–2·43; p=0·52) compared with those not exposed to the vaccine. The neonatal mortality rate was 11·78 (95% CI 5·92–23·46) per 1000 livebirths for infants whose mothers were exposed to oral cholera vaccine versus 8·91 (4·02–19·77) per 1000 livebirths for infants whose mothers were not exposed to the vaccine (crude relative risk 1·32, 95% CI 0·46–3·84; p=0·60). Only three newborn babies had malformations, two in the vaccine exposure group and one in the no-exposure group, yielding a relative risk of 2·00 (95% CI 0·18–22·04; p=0·57), although this estimate is unreliable because of the small number of outcomes. Interpretation Our study provides evidence that fetal exposure to oral cholera vaccine confers no significantly increased risk of pregnancy loss, neonatal mortality, or malformation. These data, along with findings from two retrospective studies, support use of oral cholera vaccine in pregnant women in cholera-affected regions. Funding Bill & Melinda Gates Foundation.
In this observational cohort study, we recruited women who received at least one dose of the oral cholera vaccine and who were pregnant at the time of vaccination (ie, whose fetuses were exposed to oral cholera vaccine), and women who did not receive oral cholera vaccine until their time of delivery and who were pregnant on March 30, 2015. The study was done in two neighbouring districts in the southern region of Malawi, Nsanje and Chikwawa (figure 1). Malawi has one of the highest maternal mortality ratios in the world (634 per 100 000 livebirths).11 Miscarriage is estimated to occur in about 15% of all recognised pregnancies.12 The stillbirth rate was estimated to be 24 per 1000 births in 2009,13 and the neonatal mortality rate was estimated to be 27 per 1000 livebirths from 2011 to 2016.14 Spatial distributions of the enrolled women in the study area Digital maps were obtained from DIVA-GIS. We recruited pregnant women from Nsanje who had their last menstrual period at least 3 weeks before their first dose of oral cholera vaccine and who received at least one dose, and women from Chikwawa who had their last menstrual period at least 3 weeks before March 30, 2015 (the start of the vaccination campaign), but who did not receive any doses of oral cholera vaccine on March 30, 2015. Trained study staff obtained written informed consent or assent from all study participants in the local language. For women who were not literate, the study staff read the entire consent form and participants were permitted to use their fingerprint in place of their signature; an impartial witness verified the consent. Unmarried women younger than 18 years were included after receipt of consent from their parent or legal guardian. Ethical clearance and oversight was provided by the Johns Hopkins Bloomberg School of Public Health institutional review board and the National Health Sciences Research Committee of Malawi. Before enrolment, community meetings were organised where the objectives of the study were explained to community leaders and local health and administration officials. Recruitment visits were made between June 16 and Oct 10, 2015. Initially, female village health volunteers (approximately one per village) enumerated the households in their villages and identified the women who were pregnant at any point between March 30 and April 30, 2015. Using this list, 16 trained interviewers who had completed a 4-day training programme (including a primer on human research) visited the households of women with identified pregnancies. Household visits continued until the desired sample size was reached. After enrolment, the study staff collected information about sociodemographic characteristics, current pregnancy, obstetric history, cholera vaccination status, and the global positioning satellite location of the household. Vaccination status was verified by vaccination card or entry in the vaccination register. For women who had not delivered by the time of enrolment, pregnancy status was determined by either documentation of gestational age and due date in their National Health Passport and visible signs of pregnancy, or a pregnancy test. Pregnancy tests were done at the household and the test result was individually communicated to the enrolled women. For women whose pregnancies ended between March 30, 2015, and the date of enrolment, we collected additional information including date of delivery, type of outcome, and the risk factors related to their pregnancy. The village health volunteers conducted monthly home visits to determine pregnancy status. After delivery, study staff visited the woman to collect information about delivery outcome and the health of the newborn baby for livebirths. All newborn babies presenting with health problems, per mothers’ reports, were referred to the designated health facility where a qualified medical professional did a medical examination to detect and manage any malformations or other health issues. Clinicians at these facilities caring for these babies completed a standard questionnaire. The primary endpoint was pregnancy loss (spontaneous miscarriage or stillbirth) among women exposed to oral cholera vaccine while they were pregnant compared with women not exposed but who were pregnant during the same period. We defined spontaneous miscarriage as the expulsion of an embryo or fetus from its mother at 20 weeks’ gestation or earlier.7 We defined stillbirth as a pregnancy loss which occurred on or after 21 weeks of gestation.15 The secondary endpoints were neonatal deaths and malformations. Neonatal deaths were defined as deaths occurring within 28 days of delivery. A malformation was defined as a physical defect in a live infant that was identified by a clinician at designated health facilities. Because this study was observational, we developed a clinical examination form that included a screening process to detect abnormalities, and then referred the children with abnormalities to the routine government health-care system to establish a diagnosis and manage the health issue. We estimated that 800 women exposed to oral cholera vaccine and 800 women not exposed to oral cholera vaccine would be necessary to have 80% power to detect a 1·5-times increase in the risk of pregnancy loss among vaccinated women with α of 0·05. We assumed that 7% of pregnancies would result in pregnancy loss in unvaccinated women. Assuming 10% loss to follow-up, our target sample size was 900 in each group. We compared the occurrence of pregnancy loss, neonatal death, and newborn malformation between the two groups. We compared individual-level baseline variables judged to be potentially related to the risk of pregnancy loss between the two groups using χ2 or Fisher’s exact tests when expected values in any of the cells of a contingency table are below 5, for categorical variables and Student’s t or Mann-Whitney U tests for continuous variables. In a crude analysis, we estimated the rate ratio of outcomes in women exposed versus those not exposed to oral cholera vaccine using a generalised linear model, assuming a binomial family and a link log (log-binomial regression). In a multivariable model, we adjusted the rate ratio for significantly imbalanced baseline characteristics between the two groups. We followed the rule of ten events per covariate to maximise the coverage of the confidence interval of the estimate from a regression model.16 We classified associations as statistically significant if p was less than 0·05. Data were analysed with SAS version 9.3. This study is registered at ClinicalTrials.gov, number {“type”:”clinical-trial”,”attrs”:{“text”:”NCT02499172″,”term_id”:”NCT02499172″}}NCT02499172. The sponsor of the study had no role in study design, data collection, analysis, or interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.
