An mHealth-based social support program to improve antenatal care engagement and facility-based births in Uganda: A type I hybrid effectiveness-implementation clinical trial

This type I hybrid effectiveness-implementation trial demonstrates that an mHealth-based social support intervention (Support-Moms) significantly improved antenatal care engagement, skilled birth attendance, and various maternal and neonatal health outcomes compared to standard care among pregnant women in rural Uganda.

The Gist

Imagine you are trying to climb a steep, rocky hill to reach a safe, well-equipped shelter at the top. For many pregnant women in rural Uganda, this "hill" represents the journey to giving birth in a hospital with a skilled doctor or nurse, rather than at home with untrained help. The climb is often blocked by fog (lack of information), heavy backpacks (financial or social barriers), and a feeling of walking alone.

This paper describes a study that tested a new tool to help women make that climb: a mobile phone app called Support-Moms.

The Problem: The Fog and the Lonely Path

In the study area, many women were stuck in the fog. They didn't always know the danger signs of pregnancy, they lacked the money for transport, or they felt they couldn't make the decision to go to the hospital without permission from their husbands or families. As a result, many ended up giving birth at home, which is riskier, like trying to fix a broken engine in the middle of a storm without a mechanic.

The Solution: A Digital Village

The researchers didn't just build a map; they built a digital village.

They created an app that sent automated voice messages and text messages to pregnant women. But here is the clever part: the app didn't just talk to the woman. It also invited her to pick two "digital guardians" from her real life—like her husband, a mother, a friend, or a community health worker.

Think of these guardians as co-climbers. The app sent them messages too, reminding them: "Hey, your friend is due for a check-up soon. Please help her get there."

The app acted like a 24/7 tour guide that:

1. Shone a flashlight: It gave clear, simple health advice in the local language (via voice and text) so women knew why they needed to go to the hospital.
2. Tied a rope: It connected the woman to her support network, making it harder for her to stay home because her "co-climbers" were checking in on her.
3. Carried the backpack: It reminded everyone about appointments and birth dates, reducing the mental load of remembering everything.

The Experiment: Two Groups, One Hill

The researchers gathered 824 pregnant women who were just starting their journey (less than 5 months pregnant). They split them into two groups, like two teams starting the same hike:

• Team A (Standard Care): These women got the usual care from the local health center. They were treated well, but they didn't get the special app or the extra digital reminders.
• Team B (Support-Moms): These women got the usual care plus the app. They received the voice/text messages and had their "co-climbers" (social supporters) notified by the app to help them.

The Results: Who Reached the Top?

When the researchers looked at the finish line, the difference was clear.

• Reaching the Shelter (Skilled Birth): In the group with the app, 93% of women gave birth in a hospital with a skilled professional. In the group without the app, only 84% did. That might sound like a small number, but in the world of health, that extra 9% represents hundreds of lives saved from danger.
• The Check-up Stops (ANC Visits): Women with the app were much more likely to make it to their 4 required check-up stops along the way (84% vs. 75%).
• The Safety Net: The app group had significantly fewer emergencies.
  • Bleeding after birth: This dangerous complication happened to only 9% of the app group, compared to 23% of the standard group. It's like having a better safety harness that prevented many falls.
  • Baby needing help: Fewer babies in the app group needed emergency resuscitation.
  • Breastfeeding: Almost all women in the app group started breastfeeding within an hour of birth, compared to less than three-quarters in the other group.
  • Mental Health: The women with the app felt less depressed after giving birth. The "digital village" seemed to make them feel less alone.

What Didn't Change?

The study was honest about what the app didn't do. It didn't change the weight of the baby, the length of the pregnancy, or the number of babies born. It also didn't cause any new problems. The app was a tool for navigation and support, not a magic cure for everything.

The Takeaway

Think of this study as proving that a smart, connected walking stick can help more people reach the summit safely. By combining simple technology (voice and text messages) with the power of human relationships (asking friends and family to help), the researchers created a system that worked better than just telling women to "go to the doctor."

The paper concludes that in places where people rely heavily on their families and communities, a digital tool that strengthens those existing bonds is a powerful way to save lives. It suggests that if you want to help people climb a difficult hill, don't just give them a map; give them a guide and a rope to their friends.

Technical Summary

1. Problem Statement

Despite the availability of skilled birth attendants and referral systems, Uganda continues to face high maternal and perinatal mortality rates (MMR: 189/100,000; Perinatal mortality: 41/1,000). A primary driver of these outcomes is low engagement with Antenatal Care (ANC) and a preference for unskilled home births.

• Barriers: The challenges are multi-dimensional, including lack of health education, limited social support, financial dependence, lack of decision-making autonomy, and poor birth preparedness.
• Gap in Current Interventions: While mobile health (mHealth) interventions exist in Uganda, few focus on reproductive health, and even fewer have been evaluated at scale in the public sector. Previous failures are often attributed to a mismatch between intervention design and user needs, or a lack of integration with existing social networks.

2. Methodology

Study Design:

• Type: Type I hybrid effectiveness-implementation clinical trial.
• Setting: Two rural districts in Southwestern Uganda (Mbarara and Mitooma), characterized by subsistence agriculture, food insecurity, and reliance on family networks.
• Participants: 824 pregnant women (<20 weeks gestation) recruited via Community Health Workers (CHWs).
  • Inclusion Criteria: Age ≥18 (or emancipated minors), residence within 10km of a public facility, access to a cell phone, ability to identify two social supporters, and informed consent.
• Randomization: 1:1 allocation to Standard of Care (SOC) or the Support-Moms intervention, stratified by district and health center level (blocks of 10).

Intervention (Support-Moms App):

• Platform: A user-centered, audio-SMS messaging application developed in partnership with iStreams-Uganda and Africas' Talking.
• Mechanism: Based on Andersen's Healthcare Utilization Model, the intervention targets information gaps and leverages social support networks.
• Components:
  • Automated Messaging: Scheduled SMS and audio voice messages (in preferred languages) regarding ANC appointments, danger signs, and birth preparedness.
  • Social Support Engagement: Two pre-identified social supporters (e.g., spouse, family, friend, CHW) received weekly notifications to encourage the pregnant woman's attendance at ANC and facility delivery.
  • Duration: Messages were delivered for at least 4 months (covering pregnancy and postpartum).
• Control Group: Received routine care and standard information giving per Ugandan Ministry of Health (MOH) guidelines.

Data Collection & Analysis:

• Primary Outcome: Completion of a Healthcare Provider (HCP)-led skilled birth (verified by discharge cards/facility records).
• Secondary Outcomes: ANC visit count (≥4), postpartum hemorrhage (PPH), neonatal resuscitation, breastfeeding initiation, postnatal depression, birth preparedness, and social support scores (Duke-UNC Functional Social Support Scale).
• Statistical Approach: Intention-to-treat (ITT) analysis. Crude odds ratios (OR) with 95% confidence intervals (CI) were calculated using chi-squared tests and logistic regression. Sub-group analyses were performed to test for interaction effects.

3. Key Contributions

• Integration of Social Support: Unlike many mHealth trials that focus solely on information delivery to the patient, this study successfully integrated active social network engagement. It utilized existing community relationships to create accountability and instrumental support.
• Multimodal Accessibility: The use of both SMS and audio messages addressed low literacy barriers common in rural Uganda, ensuring the intervention was accessible to women who could not read text messages.
• Hybrid Design: The study simultaneously evaluated clinical effectiveness and implementation feasibility, providing data necessary for future scale-up and integration into routine public health systems.
• Scalability Framework: The intervention was built on existing public sector infrastructure (public health centers) and utilized low-cost, widely available mobile technology, making it highly replicable in similar low-resource settings.

4. Key Results

Demographics:

• 824 women enrolled (412 per arm); 818 (99%) had complete outcomes.
• Baseline characteristics (age, gestation, education, HIV status) were balanced between arms.

Primary Outcome:

• HCP-led Skilled Birth: Significantly higher in the intervention group compared to control.
  • Intervention: 92.9% (379/408) vs. Control: 84.1% (345/410).
  • OR 2.51 (95% CI 1.57–4.03, p < 0.001).

Secondary Outcomes (Significant Improvements in Intervention Group):

• ANC Attendance (≥4 visits): 84.1% vs. 75.1% (OR 1.76, p = 0.001).
• Postpartum Hemorrhage (PPH): Reduced from 22.7% in control to 9.1% in intervention (OR 0.34, p < 0.001).
• Neonatal Resuscitation: Reduced from 13.7% to 9.8% (OR 0.69, p = 0.037).
• Early Breastfeeding Initiation (<1 hour): Increased from 71.7% to 97.1% (OR 1.76, p = 0.003).
• Postnatal Depression: Reduced from 27.1% to 20.1% (OR 0.68, p = 0.019).
• Birth Preparedness: 35.5% vs. 19.0% (OR 1.82, p < 0.001).
• Presence of Companion at Birth: 87.3% vs. 82.2% (OR 1.49, p = 0.013).
• Reported Adequate Social Support: 79.7% vs. 70.5% (OR 1.65, p = 0.002).

Non-Significant Outcomes:

• No statistically significant differences were found in term births, birth weight, mode of delivery (vaginal vs. C-section), perinatal mortality, or maternal mortality (zero deaths in both arms).
• Sub-group analysis showed the intervention was effective across most demographics, with no significant interaction effects except for inter-pregnancy interval <2 years.

5. Significance and Conclusion

Clinical and Public Health Impact:
The study demonstrates that a low-cost, mHealth-based intervention combining automated health education with active social support engagement can significantly improve maternal health behaviors and outcomes in rural, low-literacy settings.

• Magnitude of Effect: The intervention resulted in a 9% absolute increase in facility-based skilled births and a 60% relative reduction in postpartum hemorrhage.
• Scalability Potential: The authors estimate that if scaled nationally, this intervention could reduce home births by ~144,000–155,000 annually and prevent ~22,000–24,000 cases of PPH in Uganda.

Limitations:

• Reliance on self-reported data for some secondary outcomes (mitigated by chart reviews).
• Potential for contamination if clinic staff improved care for all participants due to study presence.
• Lack of multiple-testing adjustment for secondary outcomes increases the risk of Type I error.

Future Directions:
The authors recommend further evaluation of cost-effectiveness, long-term maternal-child outcomes, and sustainability mechanisms (e.g., data privacy, integration into routine MOH workflows) to facilitate national scale-up.

Conclusion:
The Support-Moms intervention represents a promising, evidence-based strategy to bridge the gap between community needs and facility-based care in sub-Saharan Africa, proving that technology works best when it leverages, rather than replaces, existing social support structures.