In short
- South African parents face barriers to trustworthy child-health information.
- Parents judge advice by fit with their child, values, routines and the giver’s intent and competence.
- Message-based tools (SMS/chatbots) can deliver low-data, personalised, culturally sensitive guidance.
- Effective systems must tailor to the child, use medical history, respect beliefs and include visual aids.
Parents act on advice that is delivered with respect and fits their child, values and routines. Drawing on workshops that took place in South Africa and Portugal, a new research paper turns insights on how parents receive health advice into practical design rules for message-based tools that could close maternal and child health information gaps.
Image: EasypeasyAI
In 2023, South Africa’s under-5 mortality rate (the mortality rate of children under five years of age) was 34.7 deaths per 1,000 live births. The country’s infant mortality rate was 24 deaths per 1,000 live births. For an upper-middle-income country, these figures are high: the average infant mortality rate for upper-middle-income countries is 11.5, and the under-5 mortality rate is 14 deaths per 1,000 live births.
According to Stats SA’s Statistical Release on Mortality and Causes of Death in South Africa (2022), influenza and pneumonia—preventable and treatable diseases—were the first leading causes of death among children aged 1–14. The 4th Triennial Report Of The Committee on Morbidity and Mortality in Children Under 5 Years found that many child deaths in South Africa between 2017 and 2020 were caused by “modifiable factors” such as delayed healthcare seeking.
Recently, the HSRC, in collaboration with the University of Cape Town (UCT) and Fraunhofer AICOS in Porto, Portugal, conducted workshops that aimed to inform the design of message-based (SMS/chatbot) information systems for maternal and child health, as few concrete design guidelines for creating such systems existed. To help create these guidelines, researchers sought to understand how parents assess the child-health advice they receive and held 14 participatory workshops with 42 parents in Portugal and South Africa. The South African workshops took place in the communities of Oceanview, Mowbray Maternity Hospital and Sweetwaters.
Why message-based systems?
While most South Africans have access to smart technology and the internet, barriers to accessing digital knowledge remain. In 2023, 96.1% of households owned at least one mobile phone, but only 78.6% of households had internet access from any location.
Limited access to data or Wi-Fi connectivity remains a key barrier to maternal and child health information. According to HSRC researcher Xolani Ntinga, South African parents face language, cultural and digital-literacy barriers that make trustworthy maternal and child health information hard to find and use. As a result, many turn to traditional healers, which can delay evidence-based care, lead to unsafe practices, and worsen child health outcomes.
Given these barriers, researchers suggest that message-based systems, such as SMS texts or chatbots in apps like WhatsApp or Telegram, could deliver short, timed health messages and two-way prompts to caregivers, providing accessible and personalised content, even in areas without regular internet connections.
Assessment of health advice
The workshop participants reported receiving overwhelming amounts of advice from family, friends, healthcare professionals and strangers. To tailor messages to families’ real contexts, researchers wanted to learn what made advice acceptable and actionable. Therefore, the goal of the qualitative study was to gain an in-depth understanding of parents’ assessment processes and translate this understanding into design implications.
One size does not fit all
While some parents were quick to take on and try out new advice, others said they felt hesitant if the advice-giver was not familiar with their child. Participants viewed advice that applied directly to their child’s unique characteristics as more useful. Such characteristics included age, diagnoses (for example, skin conditions), allergies, sleep habits and feeding temperament. When the advice conflicted with a child’s history, mothers were more likely to reject it. For example, one participant in Sweetwaters had a child with a skin condition that prevented the use of soap, making it impossible for her to follow advice related to bathing or skin-care products.
Values and beliefs
Caregivers also weighed advice against their parenting philosophy and cultural or religious practices. When guidance conflicted with deeply held beliefs or with trusted intergenerational knowledge, the advice was likely to be rejected. One participant in Portugal had been told that to help her child sleep, she should put him to bed and let him cry. “[This] was something that was entirely contrary to my philosophy, and they [specialists] actually say it shouldn’t be done,” she said.
Trusted sociocultural practices were also prominent among participants in South Africa. For example, in Sweetwaters, elders commonly recommended burning izinyamazane (incense) to soothe babies, and parents often tried it first because they trusted family knowledge.
Lifestyle and habits
Many participants noted that advice was often ignored if it was not compatible with a family’s routines, time, finances, skills or comfort level. For example, even when advice was generally agreed upon, such as the amount of sleep a child needed, parents did not always follow these recommendations if they interfered with child–parent bonding time or family routines.
Perceived intention
How advice was given (the delivery) also mattered to the participants, as delivery was associated with the intention behind it. If delivery felt judgemental or dismissive, parents would disengage, even from professionals. Empathetic delivery, continuity and familiarity with the child’s history signalled competence and made advice feel more trustworthy and actionable.
For example, one mother from Sweetwaters said that she had stopped going to a clinic because one of its nurses had told her she had not been feeding her children correctly.
Perceived competence
Parents mostly followed advice from people they trusted, such as family, friends and health workers who knew their child and spoke respectfully. When guidance came from these trusted voices, parents were more likely to try it, even if they had heard mixed messages elsewhere.
Message-based design
Researchers proposed four design options for an SMS/chatbot system:
1. Tailor to the child. The system should ask for just-enough details at the right moment in a chat
so responses match the child, without forcing a long sign-up form that many parents skip.
2. Draw on medical history. If parents voluntarily record recent vaccinations or allergies, the system could offer situated advice while being transparent about privacy and data security.
3. Respect beliefs and offer options. Instead of lecturing, systems should present evidence-based choices that acknowledge traditional practices where they’re not harmful and point parents to safe, locally validated guidance.
4. Go beyond text. Some caregiving skills were perceived to be easier to learn by watching. Systems should complement SMS or chatbot texts with images or short, low-data videos (in low-data formats) to demonstrate techniques such as nasal rinses or safe breastfeeding positions. This would also help where reading in a second language is difficult.
Impact
“Accessible and accurate information can play a crucial role in empowering parents, families and communities to make informed decisions regarding their health needs and in learning how to care for their children,” wrote the researchers.
“These findings highlight how intervention developers could harness the power of digital tools to bridge long-standing gaps in maternal and child healthcare,” Ntinga told the HSRC Review. “By providing low-data, multilingual and culturally sensitive support, digital tools can support first-time parents by providing them access to evidence-based support, especially in low-resource communities.”
Researchers encouraged designing these systems with parents and local clinicians, using low-data audio or video formats, building trust by respecting traditional practices and guiding people to safe choices. In places like South Africa, chat-style tools could answer non-urgent questions, teach in small steps and flag when someone should see a clinician, making advice easier to access where clinics are hard to reach.
Research contacts and acknowledgements
This article is based on Understanding How Parents Deal With the Health Advice They Receive: A Qualitative Study and Implications for the Design of Message-based Health Dissemination Systems for Child Health. It was summarised by Jessie-Lee Smith (HSRC science writer) with inputs from Xolani Ntinga (project director at the HSRC’s Centre for Community-Based Research).For more information, contact: Ntinga at XNtinga@hsrc.ac.za.
The study was conducted by the HSRC in collaboration with the University of Cape Town (UCT) and Fraunhofer AICOS in Porto, Portugal. Contributing researchers and authors include Beatriz Félix, Cristiana Braga, Xolani Ntinga, Sarina Till, Leina Meoli, Prof. Alastair van Heerden, Dr Ricardo Melo, Dr Nervo Verdezoto, Prof. Melissa Densmore and Dr Francisco Nunes.
If you enjoyed reading this article, please click here to subscribe to the HSRC Review quarterly magazine.
