Discharge against medical advice among infants with 24–31 weeks’ gestation admitted to Chinese neonatal intensive care units: A multicenter cohort study

BACKGROUND: Previous studies demonstrated high rates of discharge against medical advice (DAMA) among very preterm infants (VPIs) in China. OBJECTIVES: The aim of this study was to investigate the concurrent incidence, variation, and predictors of DAMA, along with the effect of DAMA on mortality of VPIs in China using data from the Chinese Neonatal Network (CHNN). METHODS: All infants born at 24–31 completed weeks’ gestation and admitted to 57 CHNN neonatal intensive care units (NICUs) in 2019 were included for this cohort study, excluding infants with major congenital anomalies. Patient information was prospectively collected using the CHNN database. Multivariable log-linear regression analysis was used to assess the association of perinatal factors and DAMA. RESULTS: A total of 9,442 infants born at 24–31 completed weeks’ gestation and admitted to 57 CHNN participating sites in 2019 were included in the study. Overall, 1,341 infants (14.2%) were discharged against medical advice. Rates of DAMA decreased with increasing gestational age (GA), and infants with lower GA were discharged earlier. DAMA infants had significantly higher rates of necrotizing enterocolitis, severe brain impairment, and bronchopulmonary dysplasia than non-DAMA infants. A total of 58.2% DAMA infants were predicted to die after discharge. The attributable risk percentage of mortality among DAMA infants was 92.4%. Younger maternal age, lower gestational age, small for gestational age, and Apgar score ≤3 at 5 min were independently associated with an increased risk of DAMA, while infants with antenatal steroids were less likely to be DAMA. CONCLUSION: The rate of DAMA in preterm infants between 24 and 31 weeks’ gestation remained high in China with a significant impact on the mortality rates. Continuous efforts to reduce DAMA would result in substantial improvement of outcomes for VPIs in China.