Outcome ascertainment and follow-up strategies in large-scale population-based prospective studies of chronic disease in resource-limited settings: a scoping review

Background: Large population-based prospective studies are central to understanding chronic disease determinants, but their validity depends on accurate ascertainment of fatal and non-fatal outcomes. In many low-income and middle-income countries (LMICs), limited digital health records and incomplete routine data systems necessitate long-term active follow-up for outcome ascertainment. Despite a growing number of large prospective studies in LMICs, there is limited consensus on feasible and reliable approaches for outcome ascertainment in these settings. We aimed to map follow-up strategies and methods used to ascertain outcomes in large-scale population-based prospective studies in LMICs.

Methods: We conducted a scoping review following Joanna Briggs Institute methodology and PRISMA-ScR guidance. Medline, Embase, and Global Health were searched from inception to January 2026. Eligible studies were population-based prospective studies in LMICs, enrolling ≥5,000 participants with ≥12 months’ follow-up, and reporting incident fatal and/or non-fatal outcomes for ≥1 major chronic disease. Data were extracted on follow-up methods, retention strategies, and outcome ascertainment.

Results: Thirty-eight studies were included, spanning South Asia (n=12), East Asia (n=5), the Middle East (n=10), Latin America (n=5), sub-Saharan Africa (n=1), and multi-country (n=5); thirteen studies enrolled ≥50,000 participants. Active follow-up predominated: 74% (n=28) used active follow-up alone, while 26% (n=10) combined active follow-up with registry linkage. For non-fatal outcomes, 47% (n=18) relied solely on self-report and 37% (n=14) used clinician adjudication; only 13% (n=5) reported pre-specified adjudication forms. For fatal outcomes, 26% (n=10) reported mortality-registry linkage, but only one study commented on registry completeness. Verbal autopsy was used in 50% (n=19) of studies; physician certification was most common, while five used automated algorithms.

Conclusion: Large LMIC prospective studies rely mainly on active follow-up but use heterogeneous and often under-reported outcome-ascertainment methods. Improved standardisation and reporting of follow-up strategies, registry coverage, and adjudication processes could strengthen the validity of evidence generated from these studies.