Validade preditiva do Denver II: teste de triagem do desenvolvimento para detecção de alterações motoras leves a moderadas em crianças pré-termo aos 7 anos de idade

Abstract

In Brazil, it is very common the use of Denver Developmental Screening Test II (Denver-II) as a screening tool for monitoring preterm children, being widely used by health professionals, as, in addition to its low cost, it is a tool that is easy and quick to apply, both in scientific research and in clinical practice. However, few studies analyze the psychometric properties of the Denver-II test, such as its predictive validity, and there are no studies that assess its ability to predict mild to moderate motor impairments in preterm children. The study aimed was to determine the predictive validity of the Denver-II, applied at 1, 2, 3, 4 and 5 years of age, to identify mild to moderate motor impairments in preterm children at 7 years of age. Data from 121 children born in the maternity hospital of the Hospital das Clínicas (UFMG) with a gestational age of less than 34 weeks and/or birth weight less than or equal to 1.500 grams, who were followed up at the Ambulatório da Criança de Risco (ACRIAR/UFMG) were analyzed, from January 2005 to December 2019. Univariate binary logistic regression analyzes were performed to confirm the association between the Denver-II results and the MABC-2 results at 7 years. Receiver Operating Characteristic (ROC) curve analyzes and calculated area under the curve (AUC) were also evaluated to determine sensitivity, specificity, positive and negative prediction values, considering 95% of the confidence interval. Probability calculation was performed to check there are more chances of presenting a result of motor impairment in the MABC-2 according to the greater number of suspect results in the Denver-II. The Denver-II results at 2 (X² = 8.90; P = 0.003), 3 (X² = 19.670; P <0.001) and 4 years (X² = 16.010; P <0.001) were significantly associated with MABC-2 at 7 years old. Children identified as suspect by Denver-II at 2, 3, and 4 years were, respectively, 3.45, 7.40, and 6.06 times more likely to have a risk for Developmental coordination disorder (DCD) result on the MABC-2 at 7 years. The greater the number of suspect results in Denver-II over time, the greater the probability of risk of DCD in the MABC-2 in 7 years (p<0.001). The AUC was considered fair for ages 2 (0.60; 95% CI 0.50–0.70), 3 (0.61; 95% CI 0.51-0.71) and 4 years (0.64; 95% CI 0.54-0.74). The Denver-II sensitivity was 66%, 70% and 69% and its specificity 55%, 51% and 54%, with a cutoff value of 78.50, 91.50 and 111.50 points for the age of 2, 3 and 4 years, respectively. This is the first study to assess the predictive validity and accuracy of the Denver-II to identify mild to moderate motor impairments in preterm children. The results obtained that the Denver-II cutoff points can discriminate children between 2 and 4 years old who are at risk for DCD at 7 years of age, supporting the use of the Denver-II for screening motor changes at 2, 3 and 4 years of age.