Can Health Information through Mobile Phones Close the Divide in Health Behaviours among the Marginalised? An Equity Analysis of Kilkari in Madhya Pradesh, India

"Understanding inequities in who participates and benefits from mobile health programmes is vital for designing strategies to reach the most marginalised."

Established in 2013 in Bihar, India, the maternal mobile messaging programme Kilkari had scaled to 13 states to reach over 10 million subscribers by April 2019. Available free of charge to subscribers, Kilkari is an example of a "direct to beneficiary" digital health communications programmes as a potential mechanism for addressing gaps in access to and receipt of health information, particularly in low- and middle-income countries (LMICs). To assess who actually participates in and benefits from Kilkari, this study assesses differentials in eligibility, enrolment, reach, exposure, and impact across three proxies of socioeconomic position: education, caste, and wealth.

Kilkari was originally designed as part of an integrated social and behavioural change (SBC) communication programme in the state of Bihar to strengthen reproductive, maternal, newborn, and child health (RMNCH) practices and generate demand for public health services. (See Related Summaries, below.) Subscribers receive up to 90 minutes of content via up to 72 weekly calls from the fourth month of pregnancy until the child is 1 year old. Communication is layered through different channels, including face-to-face communication by frontline health workers, to increase reach and exposure and improve diffusion of information.

Across the 13 states where Kilkari implementation is ongoing, 2015 Indian National Family Health Survey (NFHS) data indicate that fertility rates are highest among the economically poorest and among those with no schooling, which may suggest greater representation of marginalised groups among pregnant and postpartum women in Kilkari than in the broader population at any given time. "While it is difficult to quantify which subgroups have the most 'need for information', it would not be unreasonable to assume that the less educated, poorest and most marginalised castes are likely to benefit from such an intervention."

Data from 2015 NFHS indicate that the gap in household mobile phone ownership between the advantaged general caste (GC) and least advantaged scheduled tribe (ST) is 15.4%, while the gap in women's access to mobile phones between the GC and ST is 27.1%. These differences in household ownership of phones and women's reported phone access are different in magnitude between the economically poorest and richest (21% and 47%) segments of the population, and those with higher education and no schooling (13.9% and 55.3%).

To determine the impact of Kilkari on key RMNCH behaviours, the researchers conducted an individually randomised controlled trial (RCT) in 4 districts of Madhya Pradesh, India from 2017 to 2020. The questions considered for the analysis include whether a household lists a mobile phone in its assets and if women report having a mobile phone they can use. The health outcomes considered in this analysis are based on self-reports at endline by women in the trial who received the intervention.

A comparison of the NFHS and RCT samples conveys 2 key points: (i) The RCT sample is similar in profile to the population-based NHFS sample of women 15-49 years of age with access to a phone, and (ii) there are stark differences in educational levels and caste between those women enrolled in the Kilkari RCT and those without access to a mobile phone according to the NFHS survey. "These differences underscore the higher educational level of those enrolled in even the most basic of 'direct to beneficiary' mobile health communication programmes - that is, programmes that can be accessed from the most basic mobile phone with no additional software or skill beyond answering a phone call."

Subscribers are "reached" if the Kilkari call not only arrives to the handset but is answered. To optimise reach, Kilkari attempts to call the same subscribed mobile phone numbers up to 9 times each week. The RCT found that, on average, 60% of subscribers from the economically richest quintile have been reached by the fifth call attempt, compared to only 40% of those from the economically poorest quintile. The number of attempts needed to successfully reach a subscriber may vary based on factors including sociodemographic characteristics, phone characteristics, environmental factors, and social norms underpinning phone access and use.

In this study, to be considered "exposed", subscribers needed to listen to 50% or more of the cumulative content they were eligible to receive. Those in the most marginalised groups in the intervention arm of the Kilkari RCT sample - that is, those in scheduled castes and ST, those with no education, and the economically poorest - were less likely to be exposed to Kilkari. This finding could be due to factors including more frequent changes in mobile phone numbers among these groups, and these groups being less able to answer calls during the day, less able to maintain the battery charge of their phones, less able to retain sufficient phone credit to receive calls, and more likely to live in areas with poor network connectivity.

Figure 4 in the paper includes summary measures of equity for different health behaviours between subscribers who were exposed and not exposed to Kilkari in the intervention arm of the RCT. Three key takeaways include: (i) Not all health behaviours are similar in their equity distribution; (ii) among those exposed to Kilkari when compared with those not exposed, the inequities in wealth linked to many health behaviours are lesser than the inequities in education; and (iii) Kilkari appears to allow the "worse off" to make proportional gains in line with the "better off".

In short, the results suggest that exposure to Kilkari has improved levels of some health behaviours across marginalised groups but has not completely closed preexisting gaps in indicators such as wealth and education. "Unless the gender gap in mobile phone access is addressed, inequities in the population coverage and reach of Kilkari will persist."

The researchers suggest that: "Digital 'direct to beneficiary' communication programmes that seek to improve health through behavioural change communication with the population may have differential impact on those who can understand and act on the information provided. Hence the presence of some level of education...appears to act as an effect modifier on the pathway to programme impact. The way forward for digital 'direct to beneficiary' communication programmes may be to customise and target health information based on the characteristics of specific segments of the population to help close certain gaps in health behaviours...."

In conclusion: "An analysis of who is left behind by such interventions and how to reach those groups through alternative communication channels and platforms should be undertaken at the intervention design phase to set reasonable expectations of impact."