Objective To investigate the predictive value of serum lactate dehydrogenase (LDH) levels within 24 hours (hereinafter referred to as 24 h-LDH level) on the degree and prognosis of sepsis for the children admitted to the pediatric intensive care unit (PICU).
Methods A total of 326 children with sepsis who were hospitalized in the PICU of West China Second University Hospital, Sichuan University from May 4, 2015 to December 31, 2016 were collected as research subjects. Clinical data were retrospectively analyzed, including age, hospital stay, 24 h-LDH levels, affected organs /systems of primary infectious disease, organ damage, discharge outcomes, etc.. The Wilcoxon rank sum test or Kruskal-Wallis H rank sum test were used to compare the age, hospital stay, 24 h-LDH levels of children, and so on. Spearman rank correlation analysis was used to analyze the correlation between 24 h-LDH levels, sepsis severity, the occurrence of multiple organ dysfunction (MODS) and discharge outcomes, and so on. The receiver operating characteristic (ROC) curve was drawn to predict severe sepsis, septic shock (SS), MODS and death of children with sepsis by 24 h-LDH levels, and the area under the ROC curve (AUC) was calculated. According to the maximum principle of Youden index, the optimal cut-off value of 24-h LDH levels were determined to predict severe sepsis, SS, MODS and death, and the sensitivity and specificity of the prediction of the above indicators were calculated. This study met the requirements of the World Medical Association Declaration of Helsinki revised in 2013.
Results ①The incidences of MODS in children with sepsis, severe sepsis and SS were 27.5% (30/109), 68.8% (121/176) and 78.0%(32/41), and the mortality rates were 0, 8.0%(14/176) and 22.0% (9/41), respectively. ②The median 24 h-LDH levels of children with sepsis, severe sepsis, and SS were 718 U/L (560-976 U/L) and 871 U/L (666-1 972 U/L) and 1 588 U/L (799-2 150 U/L), respectively, the differences among them was statistically significant (χ2 = 27.445, P=0.008). The median 24 h-LDH levels of sepsis children with MODS was 713 U/L (571-967 U/L), which was significantly lower than that of sepsis patients without MODS 1 002 U/L (706-2 150 U/L), and the difference was statistically significant (Z=—5.745, P<0.001). The median 24 h-LDH levels of children with recovery, improvement, abandonment treatment and death were 712 U/L (584-945 U/L), 824 U/L (605-1 200 U/L), 1 311 U/L (716-2 150 U/L) and 2 150 U/L (1 380-2 150 U/L), respectively, the difference was statistically significant (χ2 =38.105, P<0.001). ③The results of Spearman rank correlation analysis showed that 24 h-LDH levels were positively correlated with sepsis severity and the occurrence of MODS (rs=0.290, 0.319; P<0.001). The level of 24 h-LDH were negatively correlated with the discharge outcomes of children with sepsis (rs=-0.387, P<0.001). ④ROC curve analysis results showed that the ROC-AUC of 24 h-LDH levels predicted severe sepsis, SS, MODS were 0.658 (95%CI: 0.598-0.719, P<0.001), 0.671 (95%CI: 0.583-0.759, P<0.001), 0.685 (95%CI: 0.628-0.742, P<0.001) and 0.788 (95%CI: 0.698-0.879, P<0.001), respectively. According to the maximum principle of the Youden index, when the above four conditions occur, the optimal cutoff values of serum LDH in children were 1 294.0 U/L, 1 307.0 U/L, 1 318.5 U/L, and 1 323.0 U/L (the upper limit of normal reference value of serum LDH level was 618 U/L). And the sensitivity for predicting severe sepsis, septic shock, MODS and death in children were 39.2%, 63.4%, 43.7%, 82.6%, respectively, and the specificity were 87.2%, 74.7%, 78.2% and 73.9%, respectively. ⑤The incidence of MODS in sepsis children with sepsis with 24 h-LDH levels > 1 236 U/L was 4.565 times higher than those with 24 h-LDH levels ≤ 1 236 U/L (OR=4.565, 95%CI: 2.662-7.830, P<0.001). ⑥Compared with the probability of death, abandonment, improvement and complete recovery of patients, the 24 h-LDH levels > 1 236 U/L, which were 25.175 times (OR=25.175, 95%CI: 7.741-81.875, P<0.001), 5.979 times (OR=5.979, 95%CI: 3.142-11.380, P<0.001), 1.619 times (OR=1.619, 95%CI: 0.824-3.182, P=0.162) higher than the 24 h-LDH levels ≤1 236 U/L, respectively.
Conclusions The 24 h-LDH levels in children with sepsis is related to the severity of sepsis, the occurrence of MODS and discharge outcomes. Especially, the 24 h-LDH levels > 1 236 U/L, which can be considered as a potential biomarker for the prognosis and predicting whether MODS occurs in children with sepsis.