Saleh Shadi, El Arnaout Nour, Sabra Nadine, El Dakdouki Asmaa, El Iskandarani Khaled, Chamseddine Zahraa, Hamadeh Randa, Shanaa Abed, Alameddine Mohamad
Global Health Institute, American University of Beirut, Beirut, Lebanon.
Faculty of Health Sciences, American University of Beirut (AUB), Beirut, Lebanon.
Front Digit Health. 2025 Aug 12;7:1574946. doi: 10.3389/fdgth.2025.1574946. eCollection 2025.
Maternal health in Lebanon faces significant challenges, particularly among disadvantaged populations, due to limited access to antenatal care (ANC) and a strained healthcare system. While mHealth interventions have improved maternal outcomes globally, few engage healthcare providers (HCPs) or incorporate advanced tools like artificial intelligence (AI) and gamification. This study evaluated the effectiveness of an AI-based, gamified mHealth intervention, Gamification and AI and mHealth Network for Maternal Health Improvement (GAIN MHI), on ANC utilization and maternal and neonatal outcomes in Lebanon.
The intervention included two arms: one targeting pregnant women and their spouses without HCP engagement and another involving HCPs. A post-intervention analysis was conducted with 2,880 pregnant women divided into three groups: control ( = 1,315), non-HCP intervention ( = 668), and HCP intervention ( = 897). Intervention components included AI-driven, gamified HCP professional development via the GAIN MHI app, weekly WhatsApp-based educational messages, and ANC visit reminders. Data on healthcare access (ANC visits, supplement intake, ultrasounds, and lab tests) and outcomes (term delivery, maternal/neonatal complications) were analyzed using logistic regression to calculate adjusted odds ratios (OR).
The HCP arm significantly improved healthcare access, with higher odds of attending ≥4 ANC visits (OR = 1.968, 95% CI: 1.575-2.459), completing ≥2 ultrasounds (OR = 3.026, 95% CI: 2.301-3.981), lab test completion (OR = 2.828, 95% CI: 1.894-4.221), and supplement intake (OR = 1.467, 95% CI: 1.221-1.762). Term deliveries were more likely in the HCP arm (OR = 1.360, 95% CI: 1.011-1.289), and neonatal morbidity decreased by 52.15% (OR = 1.521, 95% CI: 1.127-2.051). No improvements were seen in abortion rates, and normal deliveries decreased across intervention arms. Significant baseline demographic differences, including nationality and chronic disease prevalence, were observed between groups.
Integrating HCPs into an mHealth intervention significantly enhanced ANC uptake and maternal and neonatal outcomes in disadvantaged populations in Lebanon. These findings underscore the importance of combining digital tools with clinical support to address systemic barriers and improve maternal health in resource-limited settings. Future interventions should address delivery practices and broader social determinants of health to achieve sustainable impacts.
黎巴嫩的孕产妇健康面临重大挑战，尤其是在弱势群体中，这是由于获得产前护理（ANC）的机会有限以及医疗系统紧张所致。虽然移动健康（mHealth）干预措施在全球范围内改善了孕产妇结局，但很少有措施能让医疗服务提供者（HCP）参与其中，或纳入人工智能（AI）和游戏化等先进工具。本研究评估了一种基于AI的、具有游戏化元素的mHealth干预措施——改善孕产妇健康的游戏化、AI与mHealth网络（GAIN MHI）对黎巴嫩ANC利用率以及孕产妇和新生儿结局的有效性。
该干预措施包括两个分支：一个针对孕妇及其配偶，不涉及HCP参与；另一个涉及HCP。对2880名孕妇进行了干预后分析，这些孕妇分为三组：对照组（n = 1315）、非HCP干预组（n = 668）和HCP干预组（n = 897）。干预内容包括通过GAIN MHI应用程序进行的AI驱动的、具有游戏化元素的HCP专业发展、每周基于WhatsApp的教育信息以及ANC就诊提醒。使用逻辑回归分析医疗服务获取情况（ANC就诊、补充剂摄入、超声检查和实验室检查）和结局（足月分娩、孕产妇/新生儿并发症）的数据，以计算调整后的优势比（OR）。
HCP分支显著改善了医疗服务获取情况，参加≥4次ANC就诊的几率更高（OR = 1.968，95%置信区间：1.575 - 2.459），完成≥2次超声检查的几率更高（OR = 3.026，95%置信区间：2.301 - 3.981），实验室检查完成率更高（OR = 2.828，95%置信区间：1.894 - 4.221），补充剂摄入率更高（OR = 1.467，95%置信区间：1.221 - 1.762）。HCP分支的足月分娩可能性更大（OR = 1.360，95%置信区间：1.011 - 1.289），新生儿发病率降低了52.15%（OR = 1.521，95%置信区间：1.127 - 2.051）。流产率没有改善，各干预组的顺产率均有所下降。各群体之间观察到显著的基线人口统计学差异，包括国籍和慢性病患病率。
将HCP纳入mHealth干预措施显著提高了黎巴嫩弱势群体中ANC的利用率以及孕产妇和新生儿结局。这些发现强调了将数字工具与临床支持相结合以解决系统性障碍并改善资源有限环境中孕产妇健康的重要性。未来的干预措施应解决分娩实践和更广泛的健康社会决定因素，以实现可持续影响。
