| Objective:To investigate the prevalence of vitamin D deficiency(VDD)in critically ill children,to assess the relationship between VDD and the severity and prognosis of critically ill patients,and to determine the factors that influence serum vitamin D levels in critically ill children.Methods:From October 1,2017 to October 1,2018,eligible children in the Pediatric Intensive Care Unit(PICU)of the Hainan Provincial Maternal and Child Health Hospital were selected as subjects.Serum 25-hydroxyvitamin D[25(OH)D]’s levels were measured within 24 hours of PICU,and according to the 25(OH)D level,the patients were divided into vitamin D deficiency group,insufficient group and sufficient group,and a study was conducted to track the characteristics of disease outcome in each group.General information and 25(OH)D levels were compared between critically ill children and healthy children who were in the outpatient clinic at the same time.General data and clinical symptoms of critically ill children were collected.The pediatric critical illness score(PCIS)and early childhood warning score(PEWS)were used to assess the severity of the disease and to identify and predict the risk of the disease.Follow-up the child until he was transferred out of the PICU and discharged,the incidence of multiple organ dysfunction syndrome(MODS),sepsis,and shock was recorded,as well as PICU retention time,total hospital stay,mechanical ventilation rate and time,mortality,etc.Results:Serum 25(OH)D levels were significantly lower in critically ill children than in healthy children(t=8.332,P=0.000),and the prevalence of VDD was higher than healthy children(15.0%vs 4.0%,P=0.000).The level of 25(OH)D in children was closely related to age(F=6.222,P=0.002),and the older children had lower25(OH)D.The level of serum 25(OH)D in children with sepsis(t=-2.631,P=0.009)or shock(t=-3.001,P=0.003)were significantly lower than those without sepsis and shock.Although the 25(OH)D level decreased with the combination of MODS,the difference was not statistically significant(t=-1.455,P=0.147).The vitamin D deficiency group in children with PICU was more likely to have shock(χ~2=9.710,P=0.002)and sepsis(χ~2=6.621,P=0.010)than the adequate group.There were no statistically significant differences in PCIS score,PEWS score,MODS,PICU retention time,total hospital stay,mechanical ventilation rate,mechanical ventilation time,and mortality between the three groups at the 25(OH)D level(P Both>0.05).In the PCIS score,there was a downward trend in the level of extremely dangerous reorganization compared with the dangerous reorganization and non-hazardous reorganization serum 25(OH)D,but no significant difference between the three groups(F=1.322,P=0.269).There was no significant difference in serum 25(OH)D levels between the death group and the survival group(t=-0.562,P=0.575).In multivariate analysis,age(OR=1.630,P=0.005),sepsis(OR=3.214,P=0.038),and shock(OR=7.243,P=0.009)were significantly associated with 25(OH)D deficiency.Conclusion:Serum 25(OH)D levels were significantly lower in critically ill children,and the prevalence of VDD was significantly higher than healthy children.As the age increases,serum 25(OH)D levels are significantly lower in critically ill children.Serum 25(OH)D levels were significantly lower in patients with sepsis or shock than those without sepsis and shock,whereas serum 25(OH)D levels were only decreased when MODS was combined.A comparison between the vitamin D deficiency group,the insufficient group,and the sufficient group found that the risk of sepsis and shock was significantly higher in children with VDD than in the sufficient group,and vitamin D groups were not associated with PCIS score,PEWS score,MODS,mechanical ventilation rate and time,PICU retention time,total hospital stay,and mortality.At the PCIS score,the risk of extremely dangerous reorganization decreased compared to the level of dangerous reorganization and non-hazardous reorganization serum 25(OH)D.In multivariate analysis,age,presence or absence of sepsis or shock were independent risk factors for VDD. |