Excess Child Mortality Associated With Colombia's Armed Conflict, 1998-2019.

Importance: Armed conflicts are directly and indirectly associated with morbidity and mortality due to destruction of health infrastructure and diversion of resources, forced displacement, environmental damage, and erosion of social and economic security. Colombia's conflict began in the 1940s and has been uniquely long-lasting and geographically dynamic. Objective: To estimate the proportion of infant and child mortality associated with armed conflict exposure from 1998 to 2019 in Colombia. Design, Setting, and Participants: This ecological cohort study includes data from all 1122 municipalities in Colombia from 1998 to 2019. Statistical analysis was conducted from February 2022 to June 2023. Exposure: Armed conflict exposure was measured dichotomously by the occurrence of conflict-related events in each municipality-year, enumerated and reported by the Colombian National Center for Historic Memory. Main Outcomes and Measures: Deaths among children younger than 5 years and deaths among infants younger than 1 year, offset by the number of births in that municipality-year, enumerated by Colombia's national vital statistics. Results: The analytical sample included 24 157 municipality-years and 223 101 conflict events covering the period from 1998 to 2019. Overall, the presence of armed conflict in a municipality was associated with a 52% increased risk of death for children younger than 5 years of age (relative risk, 1.52 [95% CI, 1.34-1.72]), with similar results for 1- and 5-year lagged analyses. Armed conflict was associated with a 61% increased risk in infant (aged <1 year) death (relative risk, 1.61 [95% CI, 1.43-1.82]). On the absolute scale, this translates to a risk difference of 3.7 excess child deaths per 1000 births (95% CI, 2.7-4.7 per 1000 births) and 3.0 excess infant deaths per 1000 births (95% CI, 2.3-3.6 per 1000 births) per year, beyond what would be expected in the absence of armed conflict. Across the 22-year study period, the population attributable risk was 31.7% (95% CI, 23.5%-39.1%) for child deaths and 35.3% (95% CI, 27.8%-42.0%) for infant deaths. Conclusions and Relevance: This ecological cohort study of Colombia's spatiotemporally dynamic armed conflict suggests that municipal exposure to armed conflict was associated with excess child and infant deaths. With a record number of children living near active conflict zones in 2020, policy makers and health professionals should understand the magnitude of and manner in which armed conflicts directly and indirectly undermine child health.