Przystosowanie kulturowe i walidacja szpitalna sinhalskiego i tamilskiego wariantu ankiety do badania epilepsji AIIMS-INCLEN u dzieci i nastolatków ze Sri Lanki

BACKGROUND: Early recognition of childhood epilepsy remains challenging in settings with limited access to specialist services. Culturally and linguistically adapted screening tools may facilitate earlier identification of children requiring further diagnostic assessment. This study aimed to translate, culturally adapt, and evaluate the diagnostic accuracy and test-retest reliability of Sinhala and Tamil versions of the AIIMS-modified INCLEN epilepsy screening questionnaire in Sri Lankan children and adolescents aged 1 month to 18 years. METHODS: A hospital-based, two-gate diagnostic accuracy study was conducted from February to May 2026 at Lady Ridgeway Hospital for Children and the National Epilepsy Centre, National Hospital of Sri Lanka. Children with epilepsy were recruited from neurology clinics, and children without epilepsy were recruited from outpatient clinics. The questionnaire underwent forward-backward translation, expert review, cognitive interviews, and pilot testing. Trained medical graduates blinded to clinical diagnosis administered the caregiver questionnaires. The reference standard was a clinical diagnosis of epilepsy established by a paediatric neurologist. Diagnostic accuracy measures and test-retest reliability were assessed. RESULTS: A total of 319 children were included: 163 in the Sinhala cohort and 156 in the Tamil cohort. Epilepsy was present in 77 (47.2%) Sinhala-speaking participants and 76 (48.7%) Tamil-speaking participants. The Sinhala version demonstrated a sensitivity of 84.4% (95% CI 74.2-91.3) and specificity of 82.6% (95% CI 72.9-89.4). The Tamil version demonstrated a sensitivity of 96.1% (95% CI 88.9-99.2) and specificity of 96.3% (95% CI 89.4-99.2). Positive and negative predictive values were 81.3% and 85.5% for the Sinhala version and 96.1% and 96.3% for the Tamil version, respectively; these estimates reflect the enriched validation sample and should not be interpreted as community prevalence-based estimates. For the derived questionnaire classification, test-retest agreement was 100.0% (κ = 1.00) for the Sinhala version and 97.5% (κ = 0.950) for the Tamil version. CONCLUSION: The Sinhala and Tamil versions of the AIIMS-modified INCLEN epilepsy screening questionnaire demonstrated acceptable diagnostic accuracy and high reliability of the derived classification variable in this hospital-based validation sample. The questionnaire may support structured identification of children requiring further diagnostic assessment but should not replace clinical evaluation. Prospective community-based validation using consecutive recruitment is required before routine implementation in primary care or population screening programmes.