Prevalence of cerebral palsy in Uganda: a population-based study.

BACKGROUND: Few population-based studies of cerebral palsy have been done in low-income and middle-income countries. We aimed to examine cerebral palsy prevalence and subtypes, functional impairments, and presumed time of injury in children in Uganda.

METHODS: In this population-based study, we used a nested, three-stage, cross-sectional method (Iganga-Mayuge Health and Demographic Surveillance System [HDSS]) to screen for cerebral palsy in children aged 2-17 years in a rural eastern Uganda district. A specialist team confirmed the diagnosis and determined the subtype, motor function (according to the Gross Motor Function Classification System [GMFCS]), and possible time of brain injury for each child. Triangulation and interviews with key village informants were used to identify additional cases of suspected cerebral palsy. We estimated crude and adjusted cerebral palsy prevalence. We did χ2 analyses to examine differences between the group screened at stage 1 and the entire population and regression analyses to investigate associations between the number of cases and age, GMFCS level, subtype, and time of injury.

FINDINGS: We used data from the March 1, 2015, to June 30, 2015, surveillance round of the Iganga-Mayuge HDSS. 31 756 children were screened for cerebral palsy, which was confirmed in 86 (19%) of 442 children who screened positive in the first screening stage. The crude cerebral palsy prevalence was 2·7 (95% CI 2·2-3·3) per 1000 children, and prevalence increased to 2·9 (2·4-3·6) per 1000 children after adjustment for attrition. The prevalence was lower in older (8-17 years) than in younger (<8 years) children. Triangulation added 11 children to the cohort. Spastic unilateral cerebral palsy was the most common subtype (45 [46%] of 97 children) followed by bilateral cerebral palsy (39 [40%] of 97 children). 14 (27%) of 51 children aged 2-7 years had severe cerebral palsy (GMFCS levels 4-5) compared with only five (12%) of 42 children aged 8-17 years. Few children (two [2%] of 97) diagnosed with cerebral palsy were born preterm. Post-neonatal events were the probable cause of cerebral palsy in 24 (25%) of 97 children.

INTERPRETATION: Cerebral palsy prevalence was higher in rural Uganda than in high-income countries (HICs), where prevalence is about 1·8-2·3 cases per 1000 children. Children younger than 8 years were more likely to have severe cerebral palsy than older children. Fewer older children than younger children with cerebral palsy suggested a high mortality in severely affected children. The small number of preterm-born children probably resulted from low preterm survival. About five times more children with post-neonatal cerebral palsy in Uganda than in HICs suggested that cerebral malaria and seizures were prevalent risk factors in this population.

FUNDING: Swedish Research Council, Promobilia.