Predictive value of STAMP in Health Information System in children with critical congenital heart disease / 中国胸心血管外科临床杂志

ABSTRACT

@#Objective    To analyze the perdictive value of Screening Tool for the Assessment of Malnutrition in Pediatrics (STAMP) for malnutrition or postoperative complications in children with critical congenital heart disease (CHD). Methods    A total of 875 children with critical CHD who were hospitalized in West China Hospital, Sichuan University form August 2019 to February 2021, including 442 males and 433 females with a median age of 30 (12, 48) months, were assessed by STAMP in Health Information System. Clinical data of postoperative complications were collected. Results    (1) Based on World Health Organization Z-score as gold standard, 24.5% had malnutrition risk, and 34.3% were diagnosed with malnutrition. According to STAMP, the children were with medium malnutrition risk of 37.9% and high malnutrition risk of 62.1%. There was a statistical difference of incidence rate of malnutrition and detection rate of STAMP malnutrition risk in gender, age, ICU stay or length of mechanical ventilation (P<0.05); (2) with the optimal cut-off point of 5.5 in STAMP for malnutrition, the sensitivity, specificity, positive predictive value, negative  predictive value and area under the curve (AUC) were 68.3%, 84.3%, 48.1%, 88.3% and 0.82, respectively; (3) 12.0% of the children were with postoperative complications; (4) with the optimal cut-off point of 5.5 in STAMP for postoperative complications, the sensitivity, specificity, positive predictive value, negative predictive value and AUC were 83.8%, 73.1%, 18.8%, 99.1% and 0.85, respectively. Conclusion    Children with critical CHD have a higher incidence of malnutrition risk and postoperative complications. STAMP has a good perdictive value for malnutrition or postoperative complications, however, the sensitivity and specificity of STAMP are affected by the gold standard or the cut-off point.