Strategies for data harmonisation in preterm health research: Bridging the gap

Preterm birth remains a significant health challenge, impacting neonatal mortality and long-term health outcomes. Globally, about 10% of births are preterm (13.4 million), with about two-thirds of preterm births occurring in southern Asia and sub-Saharan Africa.1, 2 Despite efforts to reduce preterm births, no measurable changes have been observed in the last decade. A major concern is the lack of national routine data for preterm birth (especially in southern Asia and sub-Saharan Africa) that would enable evidence-based quality care when preterm birth occurs and tracking of trends over time to inform progress towards national and global targets. The work by Powell et al.3 in this issue of Paediatric and Perinatal Epidemiology must be applauded for acknowledging the need to have a standardised core dataset for preterm birth that will enable unified data collection and reporting across countries. A standardised core dataset will help inform care and management of preterm births by determining effective interventions, prevention strategies and policies. The Research on European Children and Adults Born Preterm (RECAP) core dataset initiative aims to fulfil this need by proposing a comprehensive dataset tailored for preterm cohort studies, emphasising standardisation and comparability across diverse settings.3 The RECAP Preterm Core Dataset comprises 160 core data items designed to capture essential information from neonatal care through adulthood. This life-course framework facilitates longitudinal studies tracking developmental outcomes and health trajectories across diverse populations. This initiative seeks to mitigate variability in data collection methods among preterm cohort studies, enhancing the reliability and utility of research findings across different geographical and healthcare contexts and appraisal using similar methodologies.3 The RECAP initiative introduces several pivotal recommendations to improve data collection in preterm cohort studies. These include standardising definitions to minimise variability in preterm birth categorisation, identifying core variables such as birthweight and maternal age essential for comprehensive analysis, incorporating longitudinal data to track developmental trajectories and implementing data harmonisation protocols to integrate diverse dataset findings accurately. These recommendations are underpinned by evidence supporting their role in enhancing study validity and reliability, underscoring the importance of adhering to standardised data collection protocols. The RECAP initiative employed Delphi rounds involving stakeholders with varied expertise. However, concerns arise regarding the declining participation of individuals with lived experiences of preterm birth in later rounds, highlighting the need for sustained engagement throughout dataset development processes.3 Efforts should be made to maintain diverse stakeholder engagement throughout all stages of dataset refinement. While the RECAP initiative addresses the need for standardised data collection, it is solely focused on high-income countries (HICs). The consultative process was targeted at stakeholders from HICs. This begs the question of its applicability to low- and middle-income countries (LMIC). The dataset's comprehensive nature (with 160 variables) may pose challenges in resource-constrained settings where prioritisation of essential data elements is crucial. A core dataset must strike the delicate balance (trade-off) between an irreducible minimum set of covariates for the provision of quality care for preterm births and 'nice-to-have' variables. The end goal is essential. Efforts such as the RECAP aim to achieve a standardised core dataset that is pragmatic to scale up across countries without compromising on quality, adding to clinical staff workload or affecting the quality of collected data—less is more! Perhaps a modular approach, allowing for flexible adaptation based on local healthcare capacities, could optimise the dataset's utility while reducing the burden on clinical staff. An example of similar multi-country efforts that might be useful for cross-learning is the NEST360 core Neonatal Inpatient Dataset (NID).4 The NEST360 Alliance is dedicated to improving neonatal outcomes in LMICs by emphasising context-specific data collection and operational strategies. A core principle for developing the NID core dataset was to prioritise clinical care and newborn outcomes data, a parsimonious variable list and electronic data entry. The NID underscores the critical need for adapting data collection protocols to local settings to effectively address resource constraints and enhance healthcare delivery.4 Another pertinent example is the life-course multisectoral approach discussed by Swartling Peterson et al.5 in LMICs. This approach advocates for integrating socio-economic and environmental factors into data collection frameworks to bolster precision health strategies. By tailoring data collection methods to local contexts, this approach not only addresses specific health disparities but also navigates the intricacies of healthcare delivery in diverse global settings.5 Comparing the RECAP Preterm Core Dataset with these global health initiatives5 underscores strengths and limitations. While initiatives like the life-course multisectoral approach enhance the dataset's relevance by addressing local health disparities and delivery challenges in LMICs,5 they also highlight the ongoing need for flexible, adaptable data collection frameworks to effectively support diverse healthcare contexts worldwide. A notable critique of the RECAP initiative is its predominantly Euro-centric stakeholder composition, which limits representation from regions with high burdens of preterm birth, like sub-Saharan Africa and Southeast Asia. Inclusive stakeholder engagement is crucial for developing a dataset that reflects global epidemiological realities and enhances its applicability across diverse populations.1, 2 Future iterations should prioritise the involvement of stakeholders from LMICs to ensure the dataset's relevance and effectiveness in these settings and as a global good. The RECAP Preterm Core Dataset significantly advances standardising data collection practices for preterm cohort studies. While its comprehensive framework holds promise for advancing research and clinical practices in HICs, efforts to enhance its global relevance and applicability to LMICs are imperative. By addressing these challenges through inclusive stakeholder engagement and adaptive data collection methodologies, the dataset can better serve its intended purpose of improving outcomes for preterm-born individuals worldwide (Figure 1). To enhance global applicability, recommendations include refining the dataset to a minimum core set that accommodates diverse healthcare settings while supporting modular expansions for specialised research needs. Collaboration with global health organisations and advocacy for funding allocations are essential to support data collection efforts in LMICs, thereby broadening the dataset's impact and relevance. Modular versions of the dataset allow for the optional inclusion of additional data elements based on the level of care and research focus, optimising resource utilisation without compromising data quality. EOO was invited to write the commentary. EOO wrote the first draft of the article with inputs from JA. None. None. None. The comments and views of the authors do not necessarily represent the views of the London School of Hygiene & Tropical Medicine. Eric O. Ohuma is a Professor of Medical Statistics and Epidemiology and Deputy Associate Dean of Research in the Faculty of Epidemiology and Population Health at the London School of Hygiene & Tropical Medicine, UK. He completed his DPhil in Medical Statistics from the University of Oxford (UK), MSc in Medical Statistics from the London School of Hygiene & Tropical Medicine (LSHTM) and BSc in Applied Statistics from Maseno University, Kenya. Joseph Akuze is a Research Fellow and Statistician in the Department of Infectious Disease Epidemiology and International Health at the LSHTM, UK. He is also affiliated to the Department of Health Policy Planning and Management at Makerere University as a Research Associate. Dr Akuze completed his PhD in Epidemiology and Population Health from the LSHTM (UK), Master/Postgraduate Diploma in Computational Statistics and Biostatistics from Makerere University, Uganda, Master of International Public Health from the Hebrew University of Jerusalem, Israel, and BSc in Statistics (Business) from Makerere University, Uganda. No data are presented in this commentary.