Análisis comparativo de pruebas de tamiz para la detección de problemas en el desarrollo diseñadas y validadas en México

Abstract

BackgroundIn recent years a number of child development screening tools have been developed in Mexico; however, their properties have not been compared. The objective of this review was to compare the report quality and risk bias of the screening tools developed and validated in Mexico in their published versions. MethodsA search was conducted in databases, gray literature and cross references. The resultant tests were compared and analyzed using STARD, QUADAS and QUADAS-2 criteria. Results“Valoración Neuroconductual del Desarrollo del Lactante” (VANEDELA), “Evaluación del Desarrollo Infantil or EDI” (CDE in English), “Prueba de Tamiz del Neurodesarrollo infantil” (PTNI), “Cartillas de Vigilancia para identificar alteraciones en el Desarrollo del Lactante” (CVDL) and “Indicadores de riesgo del Perfil de Conductas de Desarrollo” (INDIPCD-R) were included for the comparison. No test fulfilled all STARD items. The most complete in their methodological description were VANEDELA and EDI. The areas lacking more data on the reports were recruiting and patient selection (VANEDELA, PTNI, CVDL, INDIPCD-R). In QUADAS evaluation, all had some risk bias, but some serious concerns of risk bias were raised by patient sampling and by the choice of gold standard in two tests (PTNI, INDIPCD-R). ConclusionsChild development screening tests created and validated in Mexico have variable report quality and risk bias. The test with the best validation report quality is VANEDELA and the one with the lowest risk of bias is EDI.