The Impact of an mHealth Voice Message Service (mMitra) on Infant Care Knowledge, and Practices Among Low-Income Women in India: Findings from a Pseudo-Randomized Controlled Trial

Objectives mHealth interventions for MNCH have been shown to improve uptake of antenatal and neonatal services in low- and middle-income countries (LMICs). However, little systematic analysis is available about their impact on infant health outcomes, such as reducing low birth weight or malnutrition among children under the age of five. The objective of this study is to determine if an age- and stage-based mobile phone voice messaging initiative for women, during pregnancy and up to 1 year after delivery, can reduce low birth weight and child malnutrition and improve women’s infant care knowledge and practices. Methods We conducted a pseudo-randomized controlled trial among pregnant women from urban slums and low-income areas in Mumbai, India. Pregnant women, 18 years and older, speaking Hindi or Marathi were enrolled and assigned to receive mMitra messages (intervention group N = 1516) or not (Control group N = 500). Women in the intervention group received mMitra voice messages two times per week throughout their pregnancy and until their infant turned 1 year of age. Infant’s birth weight, anthropometric data at 1 year of age, and status of immunization were obtained from Maternal Child Health (MCH) cards to assess impact on primary infant health outcomes. Women’s infant health care practices and knowledge were assessed through interviews administered immediately after women enrolled in the study (Time 1), after they delivered their babies (Time 2), and after their babies turned 1 year old (Time 3). 15 infant care practices self-reported by women (Time 3) and knowledge on ten infant care topics (Time 2) were also compared between intervention and control arms. Results We observed a trend for increased odds of a baby being born at or above the ideal birth weight of 2.5 kg in the intervention group compared to controls (odds ratio (OR) 1.334, 95% confidence interval (CI) 0.983–1.839, p = 0.064). The intervention group performed significantly better on two infant care practice indicators: giving the infant supplementary feeding at 6 months of age (OR 1.4, 95% CI 1.08–1.82, p = 0.009) and fully immunizing the infant as prescribed under the Government of India’s child immunization program (OR 1.531, 95% CI 1.141–2.055, p = 0.005). Women in the intervention group had increased odds of knowing that the baby should be given solid food by 6 months (OR 1.89, 95% CI 1.371–2.605, p < 0.01), that the baby needs to be given vaccines (OR 1.567, 95% CI 1.047–2.345, p = 0.028), and that the ideal birth weight is > 2.5 kg (OR 2.279, 95% CI 1.617–3.213, p < 0.01). Conclusions for Practice This study provides robust evidence that tailored mobile voice messages can significantly improve infant care practices and maternal knowledge that can positively impact infant child health. Furthermore, this is the first prospective study of a voice-based mHealth intervention to demonstrate a positive impact on infant birth weight, a health outcome of public health importance in many LMICs. A pseudo-randomized controlled trial of the Mobile Alliance for Maternal Action (MAMA) implementation in India, called mMitra, was conducted from January 2015 to December 2017 with data collection beginning in June 2015 and ending in January 2017 (Mobile Alliance for Maternal Action Research Agenda 2015; ARMMAN, n.d.). MAMA was a four-year global initiative that aimed to improve the health and well-being of pregnant women and their newborns and infants through age- and stage-based tailored voice or text messages delivered via mobile phone (Mobile Alliance for Maternal Action Research Agenda 2015). MAMA supported the non-profit organization, Advancing Reduction in Mortality and Morbidity of Mothers, Children and Neonates (ARMMAN), to pilot a mobile messaging service and program called mMitra (ARMMAN, n.d.). The program was built on the premise that if women receive educational messages on their phone that are interesting, easy to understand, and aligned with the physiological stage of pregnancy or infant development, they will be motivated to engage in recommended self-care and seek recommended health services (Mobile Alliance for Maternal Action Research Agenda 2015). mMitra engaged pregnant women living in urban slums in Mumbai during pregnancy and through the first year of their infants’ lives. The overall aim of the program was to improve self-care and uptake of effective MNCH practices and clinical services through digital behavior change communication. The mMitra impact evaluation is registered with ISRCTN under Registration # 88968111 (See https://www.isrctn.com/ISRCTN88968111). Participants were pregnant women from urban slum areas of Mumbai. Mumbai is divided into 27 municipal wards, or administrative units, each with a population of approximately 800,000–900,000 people. Each ward is typically served by one maternity home and five or six health posts that provide pregnancy and infant health services. Each ward appoints roughly 100 community health workers who make home visits, register pregnant women and motivate them to seek health care for themselves and their children. For this study, two such wards (F North and M East) were purposely selected due to their large slum area, high population proportion classified as low-income and no prior exposure to mMitra. Women speaking Hindi or Marathi language—which are spoken by over 80% of the population in the city—were enrolled in the study, and mMitra voice messages were delivered in those two languages. Women without access to a mobile phone at home or not likely to be in Mumbai for four to five months during the pregnancy and post-delivery period (i.e., those planning to visit natal homes outside Mumbai for delivery, a common cultural practice in India) were not enrolled in the study. Pregnant women were identified and enrolled into the study by research team members. They were systematically assigned to either the intervention or control group. Group assignment was based on gestational age at the time of enrollment. For every four women enrolled consecutively, the first three were assigned to the intervention group and the last woman was assigned to the control group. The aim of the sampling was to enroll a sufficient number of women in each trimester to ensure that a dose response could be measured. The intervention group received mMitra messages; the control group did not. Women enrolled in the first trimester had a longer exposure to the messages than those enrolled in the third. All women gave their informed written consent prior to inclusion in the study. All women were followed until their infants turned 1 year of age. The mMitra package consisted of 145 voice messages designed to be shared from when a woman was 6 weeks pregnant until the infant reached 1 year of age. Messages were delivered two times per week during pregnancy; they were clustered at one message per day immediately post-partum for 7 days, and then reduced in frequency back to two messages per week from the second week of infancy. mMitra also provided a free call-back service within 2 days after the original call was received, in case women wanted to hear the messages they missed or listen to messages again. There were no text messages delivered through this program unlike in other programs. The audio messages, designed by BabyCenter (BabyCenter, n.d.), were timed to the gestational age and developmental stage of the fetus and infant and based on global [World Health Organization (WHO)] and local (National Health Mission) guidelines. The messages were adapted to local practices in partnership with ARMMAN and representatives from the Federation of Obstetric and Gynecological Societies of India and Indian Academy of Pediatrics. The translations were tested for appropriateness and cultural nuances with local health experts and community focus groups. Finally, the voice and tone of the recording artist were field-tested to ensure that the messages were delivered in the reassuring tone necessary to promote the desired behavior change. The final message product was approximately 2 min in length, beginning with a recognizable ‘jingle’ to alert family members to pass a shared household phone to the pregnant woman or mother (or to place the call on speaker phone) and ending by reiterating the key element of the message. Messages were recorded in a female voice designed to represent an educated but approachable female relative. The sampling procedure and sample size were determined to ensure that (a) the study population was representative of the target population and (b) the study sample size was adequate to detect a 10% reduction in the proportion of infants weighing < 2.5 kg at birth, in the intervention group as compared to the control group at an alpha of 0.05 and 80% power. The baseline for infants born with weight < 2.5 kg was determined to be 12.5% using data from the Government of India District Level Household and Facility Survey report (http://rchiips.org/DLHS-4.html). We estimated that there would be 30% attrition overall. Using z proportion test, the required sample size was estimated as 500 pregnant women per trimester in the intervention group (total n = 1500 in the intervention group) and 500 pregnant women in the control group. The intervention sample aimed to include sufficient women in the first, second, and third trimester of pregnancy to assess dose response against the same size control group which would not have any exposure to the messages and therefore were not stratified by trimester. From June to October 2015, research team members went from house-to-house in the two Mumbai wards to identify and recruit eligible pregnant women into the study. Each investigator aimed to enroll four to five pregnant women per day. This required visiting 100–110 homes per day, and 2000 women were enrolled in the study in 5 months. At the time of enrollment, the investigators administered a pregnancy survey (baseline survey) with all study participants. There were three rounds of data collection: Pregnancy (baseline/Time 1), Post-delivery (Time 2), and when the infant was 1 year old (Time 3). The survey instruments were digitized and available in Hindi and Marathi on the Kobo Collect Android-based platform. In addition to administering the surveys, the investigators also collected data from the participant’s Mother and Child Health (MCH) card, which is issued to every pregnant woman at the local health facility and updated each visit. Women are advised to retain these cards at home and bring them to every antenatal care and child health visit. The MCH cards contain information on services provided and clinical/laboratory findings (e.g., weight, BP, hemoglobin level) of the woman and the infant until the infant reaches 1 year of age. When faced with connectivity issues or drained batteries, the investigators completed paper-based surveys and entered the data into Kobo Collect later the same day. Every evening, investigators submitted their tablets to supervisors who checked the number of completed interviews and uploaded the data onto the central server. Every day, the data manager examined the uploaded data for completeness and consistency in responses. Any problems identified were discussed in the daily morning briefings with investigators and subsequently resolved. The first primary outcome of interest was number of full-term infants born at or above the ideal birth weight of 2.5 kg. This outcome was selected as it is a marker of the baby’s health, serves as a proxy for the nutritional status of the mother throughout her pregnancy, and the data are routinely collected at birth and are available from the woman’s MCH card. The second primary outcome of interest was nutritional status at 1 year of age (Time 3). This outcome was also assessed by collecting data recorded on the MCH cards. The nutritional status of each infant was determined using weight-for-age criteria and graphing the values over time on the WHO’s z-score graph for growth monitoring. Immunization status of the infant was assessed (at Time 3) as a secondary outcome. Being fully immunized was defined as the infant having received the schedule of vaccines under the Government of India’s Child Immunization Program, which are: one dose of Bacillus Calmette–Guérin (BCG) for tuberculosis; three doses of the pentavalent vaccine for diphtheria, pertussis, tetanus, hepatitis B and Haemophilus influenzae type B; three doses of polio and one dose of measles. Additional outcomes focused on knowledge, attitudes and practices of women. The impact of the intervention on infant care practices was determined by comparing survey results of the final round of interviews when the infants reached 1 year of age (Time 3) in the intervention group versus control group. The changes in knowledge over time about infant care within, and across the groups (intervention, control) were also assessed by comparing the responses to ten infant care knowledge questions included in surveys, conducted during pregnancy (baseline/Time 1) and shortly after delivery (Time 2). The data were analyzed using SPSS version 18.0. Descriptive analyses were conducted to ascertain the distribution of the data. Continuous data following a parametric distribution were compared between the intervention and control groups using two-paired t test. For the categorical outcomes, Chi square tests were conducted. Simple and binary logistic regressions were also conducted to assess the outcomes data and account for socio-demographic factors. A per protocol analysis was conducted and compared. We were unable to conduct an intention to treat analysis due to the unavailability of data. Analyses comparing the duration of exposure to the intervention were conducted by stratifying women in the intervention group by gestational age and categorizing them into three groups reflecting the total duration of exposure to mMitra by Time 3 (1–3, 4–6 and 7–9 months).