Validade do questionário Marcos do desenvolvimento do SurveyofWellbeingof Young Children (SWYC-Brasil) e normas para interpretação dos resultados em crianças brasileiras

Abstract

Introduction: The Developmental Milestones Questionnaire of the Survey of Wellbeing of Young Children – Brazil version (DM-SWYC-BR) is a screening tool to identify global development delays in children aged 1 to 65 months. Based on the caregivers’ report, it is free of charge, brief and easy to respond, holding a promise to be used in children´shealth care in Brazil. Objectives: To investigate the DM-SWYC-BR reliability, construct and concurrent validity, and to establish the preliminary standards for the interpretation of the results in Brazilian children. Methods: Methodological study with two investigative axes. Samples were recruited in primary health care centers and public daycares. The normative study involved 1535 children aged less than 65 months from Santa Catarina (SC), Minas Gerais (MG) and Ceará (CE). Parents answered a characterization questionnaire and the DM-SWYC-BR. DM-SWYC-BR’sinternal consistency and exploratory factor analysis were calculated. The cutoff point were defined as the value corresponding to 85% of the average score for each age (approximately the 15th percentile). The prevalence of suspected delay was calculated using both Brazilian and original standards. The concurrent validity study was conducted with a 465 children aged 23 to 58 months from Minas Gerais e Santa Catarina. All caregivers answered the DM-SWYC-BR and the Ages and Stages Questionnaire – Brazil version (ASQ-BR), and 167 children were also assessedusing the Bayley Scales of Infant and Toddler Development III(Bayley III). Spearman’s correlation coefficient and accuracy measures were calculated, comparing children’s scoreson the DM-SWYC-BR versus ASQ-BR and also versus Bayley III. Results: The DM-SWYC-BR questionnaire was considered consistent (Cronbach’s Alpha = 0.97) and unidimensional, with items’ factorial loads > 0.78. Small differences were found between the Brazilian and original cutoff points for children of most ages. At 18, 23 and 29 months, the cutoff points were higher for Brazilian sample. The cutoff points were higher for the North American sample betweern 44 to 46 and 54 to 58 months. The general prevalence of suspected developmental delay was 27.5% using Brazilian standards and 28.2% with the original ones. The correlation between the DM-SWYC-BR and ASQ-BR was moderate (rho = 0.50) and was weak with the Bayley III (rho=0.32), although significant (p<0.001). The accuracy of DM-SWYC-BR using ASQ-BR or the Bayley III as a reference test was similar. The negative predictive values (NPV) and specificity were above 70%. The positive predictive values (PPV) and sensitivity ranged from 53 to 81%. The Area Under the ROC curve (AUC)was 0.81 compared to the ASQ-BR and 0.70compared to theBayley III. Conclusion: The DM-SWYC-BR questionnaire showed satisfactory psychometric properties, supporting its use for developmental delays screening in Brazilian children. The standards for interpretation reported in the present study are those recommended when using the DM-SWYC-BR in Brazil.