The General Movements Assessment (GMA) is one of the most important tools for early diagnosis of neurodevelopmental disorders. It is a reliable, quick, and non-invasive assessment of spontaneous movements in newborns, ideal for use in the Neonatal Intensive Care Unit (NICU). Previous studies have identified a strong influence of postnatal clinical factors on the classification of general movements using the GMA in the NICU. However, this literature is still scarce in developing countries, limiting the use of the tool and, consequently, the early diagnosis and the monitoring of developmental changes.

To develop a multicriteria index with the main clinical factors related to the occurrence of abnormal classification of general movements during the NICU stay; To verify the contribution of the index to explain the percentage of abnormal classifications of general movements and to identify babies at risk for developmental changes.

This is an exploratory cross-sectional study, with data from a prospective longitudinal study. Preterm newborns (PTNB) with less than 37 weeks of gestational age were included, according to admission to the NICU. Their spontaneous movements were classified as normal or abnormal through the GMA by 2 trained and certified evaluators. The babies' clinical variables were recorded on a data sheet. Data analysis was performed using the Multicriteria Decision Support, a method that allows the development of an index to identify risk factors related to the abnormal classification of the general movements of newborns.

Fifty-two PTNB were evaluated, of which 30 (57.7%) were male, with a mean gestational age of 31.63 (±2.38) and mean birth weight of 1560.13 (±412.86). The mean total hospitalization time of the babies was 32.84 days, with the mean use of mechanical ventilation for 2.05 days; 45 (86.5%) used non-invasive ventilatory support and/or oxygen therapy. Grade I-II peri-intraventricular hemorrhage was identified in 24 (44.8%) babies and grade III in just two (3.8%); 4 (7.7%) PTNB had patent ductus arteriosus and 7 (13.5%) had postnatal infection. As for socioeconomic level, 44 (84.6%) families had an average income of less than 2 minimum wages. The multicriteria index was calculated from the equation: Multicriteria Index child i = Evaluation criterion 1 child i weight criterion 1 + .... + Evaluation criterion n child i weight criterion n. A significant positive linear association was found between the multicriteria index and the abnormal trajectories of general movements (R2=0.27; β=0.51; p<0.0001).

The developed multicriteria index was able to identify PTNB with a higher risk of developmental changes, given its positive relationship with the percentage of abnormal general movements.

The results of the present study reinforce the possibility of using GMA for the early detection of neurodevelopmental disorders in PTNB even during their stay in the NICU, helping with postnatal follow-up and early intervention, if necessary.