Alan Thevenet N. Tita

UNSTRUCTURED Artificial intelligence (AI) is rapidly transforming health systems, research, and education, creating new opportunities to address global health challenges while raising questions about equity, ethics, and governance. The symposium Artificial Intelligence and Technology in Global Health: Bridging Divides, hosted by the University of Alabama at Birmingham and international partners in October 2025, brought together researchers, clinicians, policymakers, industry leaders, implementers, trainees, and students from over 20 countries to examine the current and future role of AI in global health. A central theme was the need to ensure that AI promotes health equity rather than deepening health disparities. Discussions emphasized human-centered design, community engagement, explainable and trustworthy AI, data governance and privacy, and strengthening local capacity, particularly in low-resource settings. Speakers highlighted the importance of aligning AI tools with local health priorities, infrastructure, and cultural contexts while fostering multidisciplinary partnerships and sustainable implementation strategies. The symposium concluded that AI's greatest promise lies not in technological advancement alone, but in its responsible, ethical, and equitable integration into health systems through inclusive governance and global collaboration.

Background: This report summarizes the final results from a randomized, placebo-controlled, phase 2b trial evaluating the safety, antibody persistence, and potential efficacy of a bivalent respiratory syncytial virus prefusion F (RSVpreF) vaccine in pregnant individuals and their infants. Methods: Maternal participants were randomized from 24 to 36 weeks’ gestation to receive RSVpreF (120 or 240 µg ± aluminum hydroxide) or placebo. Results: This analysis included 579 pregnant individuals and their 572 infants from 4 countries (Argentina, Chile, South Africa, United States); 462 maternal participants received RSVpreF vaccine. Adverse events in the month following receipt of either RSVpreF vaccination or placebo (maternal) or birth (infant) were mostly anticipated events in pregnancy and the neonatal period. For all RSVpreF groups, combined RSV-A/B 50% neutralizing titers peaked 2 weeks after vaccination. At delivery, geometric mean titer ratios between RSVpreF and placebo recipients' infants were 10.9 to 13.6. Transplacental transfer ratios (all groups) were 1.39 to 1.83. Neutralizing geometric mean titers remained higher in infants whose mothers received RSVpreF vs placebo through their first 6 months of life, with an estimated half-life of 39 to 44 days. In an exploratory analysis, observed efficacy (95% CI) for the combined RSVpreF groups against medically attended and severe medically attended infant RSV-associated lower respiratory tract illness through the first 180 days of life was 75% (−11% to 94%) and 83% (−48% to 99%), respectively. Conclusions: RSVpreF had a favorable safety profile and elicited robust neutralizing responses with efficient transplacental transfer. The potential to prevent infant RSV-associated lower respiratory tract illness was subsequently confirmed in the pivotal phase 3 efficacy trial. Clinical Trials Registration: ClinicalTrials.gov (NCT04032093).

Congenital cytomegalovirus (CMV) infection, due to both nonprimary and primary maternal infection, affects nearly 1% of neonates. In conducting a randomized trial of hyperimmune globulin for prevention of congenital CMV infection, we screened over 200,000 pregnancies across the United States for evidence of prior or primary CMV infection and evaluated their geographic variation. Although the majority were seropositive, seroprevalence and primary infection risk varied significantly by geographic region, with seroprevalence the highest (88%) and primary infection (0.2%) the lowest in the Southwest. This variation in seroprevalence by geography may inform planning of public health measures to decrease congenital CMV.

Background: Early relational health (ERH) is thought to buffer the association between early life stress (ELS) and child psychopathology, but limited work has directly tested this hypothesis. Objective: We evaluate mother-infant emotional connection, a facet of ERH, as a buffer of combined and individual impacts of specific ELS exposures (maternal mental health and interpartner conflict) on child psychopathology. Methods: Participants included mother-infant dyads (n=100) followed longitudinally in the COMBO cohort, a convenience sample recruited during the COVID-19 pandemic. ERH was assessed via a remote mother-infant face-to-face interaction at ~4mo postpartum coded for emotional connection. An ELS Index was estimated using measures of maternal self-reported postpartum anxiety, depression, stress, and inter-partner conflict. Mothers rated emerging signs of child psychopathology symptoms at 2-3yrs on the Child Behavior Checklist for Ages 1½−5 (CBCL/1½−5). Main and interactive effects of ELS and ERH on emerging signs of child psychopathology were tested in generalized linear models. Results: Greater ELS Index scores were associated with a higher rate of emerging psychopathology symptoms (aRR=1.32, p<.001), but this association was moderated by a significant interaction between the ELS Index and emotional connection (aRR=0.99, p=.03), such that at higher levels of emotional connection, the association of ELS with child psychopathology symptoms was weaker (aRR=1.16, p <.001). Conclusion: Parent-infant emotional connection may buffer the impact of ELS exposure in infancy on child emerging symptoms of psychopathology in toddlerhood, supporting efforts to invest in pediatric interventions that target ERH.

Objective: A single oral dose of azithromycin (AZM) given during labor to women planning a vaginal delivery reduced maternal infections including sepsis, with a stronger effect in sub‐Saharan Africa than South Asia. Since maternal infection contributes to labor dysfunction and postpartum hemorrhage (PPH), we evaluated the effect of AZM on the risk of PPH and blood transfusion. Methods: This was an unplanned secondary analysis of the Azithromycin Prevention in Labor Use Study (A‐PLUS) randomized controlled trial at eight sites in seven low‐ and middle‐income countries in sub‐Saharan Africa, South Asia, and Latin America. The population consisted of pregnant women in labor at ≥28 weeks' gestation in health facilities randomized to either 2 g AZM or placebo. Based on an intent‐to‐treat analysis, the risk of PPH and blood transfusion was compared between AZM and placebo arms using Poisson regression adjusting for arm and site as fixed effects. The main outcome measures were (1) PPH (500 mL or greater) after delivery; and (2) postpartum blood transfusion after delivery. Results: A total of 29 278 participants were randomized to APLUS; 14 590 to AZM and 14 688 to placebo. The risk of PPH did not significantly differ between AZM and placebo arms (1.4% in AZM; 1.6% in placebo; relative risk [RR] = 0.88; 95% confidence interval [CI]: 0.73, 1.07). The risk of blood transfusion also did not significantly differ between AZM and placebo arms (0.5% in AZM; 0.5% in placebo; RR = 0.90; 95% CI: 0.65, 1.25). There was also evidence indicating that the effect of AZM on the risk of blood transfusion, but not PPH, was beneficial in sub‐Saharan Africa but not in South Asia ( P value for two‐way interaction = 0.002). Conclusion: A single intrapartum oral dose of AZM did not significantly reduce the overall risk of PPH or blood transfusion. ClinicalTrials.gov Identifier: NCT03871491.

Objective: The randomized trial of azithromycin to reduce maternal and neonatal sepsis (the A‐PLUS Trial) found substantial reduction in maternal sepsis among women receiving azithromycin and substantial non‐study antibiotic use. This secondary analysis explored the effect modification of non‐study antibiotics on azithromycin versus placebo on maternal and newborn infection among A‐PLUS participants. Methods: Women ≥28 weeks gestation in labor and planning a vaginal delivery at a study hospital in seven low‐ and middle‐income countries (Bangladesh, India [two sites], Pakistan, Guatemala, Kenya, Democratic Republic of Congo, and Zambia) were eligible for inclusion. Non‐study antibiotic use was collected prospectively. We estimated the interaction of non‐study antibiotics with azithromycin versus placebo on maternal and newborn sepsis. Results: A total of 29 287 participants were randomized (14 590 to azithromycin; 14 688 to placebo). Maternal infection was reduced among the azithromycin group compared to placebo among those who did not receive non‐study antibiotics, with estimated relative risk (RR) 0.58 (95% confidence interval [CI] 0.48, 0.70), and among those who received non‐study antibiotics, with RR 0.80 (95% CI 0.70, 0.91). Similar results were observed for maternal sepsis. Neonatal infection was not significantly reduced in any group. These results were similar when stratified by African and Asian region but not statistically significant. Conclusion: Our results suggest a benefit of azithromycin in reducing maternal infection or sepsis across all groups, with a larger reduction in risk among participants who had not received other antibiotics. Given the concerns of inappropriate use of antibiotics, further research is warranted to determine the most effective strategies of reducing risk of infection.

Objective: The United States (US) reports disparities in maternal health outcomes across states, regions, and racial/ethnic groups. Few studies in the US South have sought to understand the perspectives of Non-Hispanic Black (NHB) women about their experiences of maternity care. This study examines NHB women's maternity care experiences in three Southern states. Methods: Through in-depth interviews (February—September 2023), we explored maternity care experiences of NHB women experiencing adverse maternal outcomes ( n = 18), and family members of deceased NHB women or infants ( n = 2), in Alabama, Georgia and North Carolina. Grounded in the World Health Organization’s Quality of Care Framework for Maternal and Newborn Health Care, our study elucidated the experiences of these NHB women and family members with racial and other discrimination, and their preferences for treatment during maternity care. We adopted a mixed deductive and inductive thematic approach to analyze the data using NVivo software. Results: Three major themes emerged: 1) experience of care, 2) perceptions of racial and other discrimination during maternity care, and 3) preferred manner of treatment in maternity services. Perceptions about discrimination due to race and other identities were shared. Issues related to provider-patient communication, respect, and emotional support were identified as main factors that influenced NHB women’s experiences of maternity care. Conclusion: Both positive and negative maternity care experiences, including instances of overt racism and discrimination were reported by NHB women and families who had experienced adverse pregnancy outcomes. Our findings highlight the need to prioritize respectful maternity care through targeted quality measures.