Edit Content

Pilot results: Testing an mHealth approach to reach the last mile of Eswatini’s under-immunized infants

Completing a year-long pilot in partnership with Eswatini’s Ministry of Health (MoH) and the Clinton Health Access Initiative (CHAI), Eswatini to test PROMPTS as a cost-effective tool for infant immunization tracking.

Completing a year-long pilot in partnership with Eswatini’s Ministry of Health (MoH) and the Clinton Health Access Initiative (CHAI), Eswatini to test PROMPTS as a cost-effective tool for infant immunization tracking.

The Immunization ‘Data Gap’ in Eswatini

Immunization is one of the lowest cost, highest impact public health interventions to reduce child mortality, currently preventing 2-3 million deaths every year. Yet global vaccine coverage has remained stagnant over the last few years due, in part, to COVID-19-related service disruptions and the need for tailored immunization tracking and outreach strategies to reach the “last mile” of under-immunized and unimmunized children.

In Eswatini, a decline in routine immunization coverage was reported for most antigens in 2020, with approximately 24% of infants U1 not receiving full dose. A continued focus on the COVID-19 response, persistent gaps in operational funds, stockouts, and skilled provider shortages remains a barrier to the country meeting its 90% coverage targets.

Increasing access to immunization directly correlates to declining rates of under-5 mortality. Yet in the path to 100% coverage, a data challenge emerges – namely, how to identify and reach under-immunized babies and infants when coverage is already high.

An mHealth solution to bridge this gap

In late 2021, Jacaranda Health, the Expanded Program on Immunization (EPI), MoH, and CHAI Eswatini announced a new partnership to test and adapt PROMPTS, Jacaranda’s AI-enabled digital health tool, for the immunization use case in Eswatini.

PROMPTS is a two-way SMS service that empowers new and expecting mothers with gestation stage-specific information to make informed decisions, uses AI to power real-time exchanges, and generates nuanced data on their experiences of the services that support them.

Already operational at scale in Kenya (including for the immunization use case), the platform’s two-way messaging capabilities and sophisticated data infrastructure made it well positioned to:

  1. Identify under-immunized infants and ensure that mothers take the babies to the clinic for missed vaccinations by schedule
  2. Offer EPI the data tools to improve immunization coverage in geographic ‘hotspots’ in a resource-efficient manner.

Adapting PROMPTS for the context and use case

In early 2022, Jacaranda Health and CHAI Eswatini staff engaged in a participatory process with facility staff to understand supply-side challenges impacting immunization (such as stock-outs) and their input on the tool.

The team also conducted a series of Focus Group Discussions (FDGs) with mothers to better understand barriers leading to defaulting or hesitancy (such as a lack of information around the immunization schedule), and their preferences for message frequency, timing, and content.

The FGDs resulted in considerations around;

  1. Tailored content to address knowledge gaps, including information about the immunization schedule, and timed reminders for when to take their newborns for vaccines.
  2. Translation to meet linguistic preferences, including working with local translators to translate messaging into Siswati and adapting PROMPTS’ AI-based system to read and respond to mothers’ questions in their preferred language.
  3. Data collection to inform personalization, including collecting gender-specific data and tailoring messages accordingly towards both parents to incentivize fathers to share responsibilities for infant immunization.

How did PROMPTS perform as an immunization tracking tool in Eswatini?

The pilot took place from August 2022 through March 2023 and covered childhood immunizations provided during the first 16 weeks of life. While the pilot had originally aimed to reach 500 women in the four regions, by its end, over 1,300 mothers had been enrolled on PROMPTS, with notably strong results, showing that:

1. PROMPTS utilization among new mothers is high.

  • Nearly all mothers (88%) reported having read all messages they received, and the majority (97%) found them useful. 95% said they would recommend PROMPTS to a friend.
  • Most mothers reported that the immunization appointment reminders were the most useful component of service.
  • 17% of mothers asked a question on PROMPTS. Some users reported they didn’t know they could ask a question, or had concerns about the SMS charges.

 

2. There were some positive behavior changes around immunization

  • An increase in uptake of 1+ dose of BCG vaccine between baseline (96%) and endline (99%)
  • Knowledge change was mixed. Positive results on knowledge showed that more mothers:
    • Recognized they could come back for missed vaccine doses
    • Knew that it is common for an infant to develop a mild fever after a vaccine
    • Understood to avoid giving pain relievers to the infant for swelling / pain.
    • Reported that timing of vaccines is important for immunity
    • Knew that the first vaccine is given at birth

 

It’s important to remember that the results above have been generated through a simple pre-post comparison from a small pilot sample, so the data are not implying causality. However, they do indicate that PROMPTS is both acceptable and relevant to mothers, as well as effective in shifting some behaviors related to vaccine uptake and adoption.

The findings also offer some actionable insights to further adapt both the messaging functionality and data collection to increase relevance for mothers, and the granularity of immunization data generated for EPI.

For example, concerns raised from participants around message costs indicate a need to reiterate the free nature of the service, and its capability to offer two-way exchange and answer any questions they might have.

Additionally the reported value of the platform was reflected in the behavior change trends, and so a greater emphasis on appointment reminders that respond to mother’s preferences on timing, frequency and sequence could help promote better adherence to a childhood vaccination schedule.

Expanding PROMPTS to support a broader MNCH use case in Eswatini

Following the conclusion of the pilot, Jacaranda Health and CHAI, with MoH support, have continued to enroll 700+ additional mothers onto PROMPTS.

Yet beyond direct scale-up of the immunization work, the results have generated strong interest from the MoH to expand PROMPTS geographically (ie. to all facilities in the country), and for a broader use case to help move the needle on key MNCH priorities.

This would include all childhood immunizations up to age 5, as well as behavior change around other essential practices including antenatal and postnatal care (already an existing capability of the platform in Kenya), and prevention of mother-to-child transmission (PMTCT) of HIV, which is driven by low reported knowledge on practices which could expose children to HIV post-delivery.

Complementing these efforts, over the next year Jacaranda aims to explore models for increasing the sustainability of PROMPTS in the long-term, including increasing the level of government ‘cost-share’, or financial contributions towards the platform’s running costs, and other levers to increase economies of scale, such as partnering with telcos to zero-rate the cost of bulk SMS.

We welcome the opportunity to talk with partners about how to scale this program.

Share this resource