To explore the indications of cholangiography for children with cholestasis and biliary atresia (BA) younger than 3 months old.

A total of 110 children who underwent cholangiography from April 2016 to April 2017 in Department of Neonatology, Children′s Hospital, Capital Institute of Pediatrics were selected into this study. According to different results of cholangiography, they were divided into BA group (n=36) and cholestasis group (n=74). The general clinical data, biochemical indexes and ultrasonic examination results of 2 groups were analyzed retrospectively. Independent-samples t test was used to statistically analyze the age, duration of jaundice, and aspartate aminotransferase (AST), total bilirubin (TBIL), direct bilirubin (DBIL), albumin, globulin, serum Ca²⁺ , bile acid, blood ammonia, lactic acid and alkaline phosphatase (ALP) levels between 2 groups. Mann-Whitney U rank sum test was used to analyze the levels of gamma-glutamyl transferase (GGT) and alanine transaminase (ALT) statistically. The gender composition ratio, incidence of kaoline stool, and the proportions of liver and spleen enlargement, GGT level ≥300 U/L, TBIL concentration ≥150.0 μmol/L and DBIL concentration ≥100.0 μmol/L were statistically analyzed by chi-square test. The study was in accordance with World Medical Association Declaration of Helsinki revised in 2013.

①There were no statistical differences between 2 groups in the general clinical data, such as age, the onset time of jaundice, duration of jaundice, and proportions of liver and spleen enlargement (P>0.05). ②There was statistically significant difference between two groups in gender composition ratio (χ²=13.580, P<0.001). The incidence of kaoline stool, and proportions of GGT level ≥300 U/L, TBIL concentration ≥ 150.0 μmol/L, DBIL concentration ≥ 100.0 μmol/L and the proportion of BA revealed by ultrasound examination in BA groups all were statistically higher than those in cholestasis group (χ²=18.396, 37.914, 25.728, 27.957, 53.606; all P<0.001). GGT level and DBIL concentration in BA group were 526.0 U/L (409.9-1 187.4 U/L) and (116.3±32.2) μmol/L, respectively, which were significantly higher than 101.9 U/L (71.8-440.7 U/L) and (88.8±38.8) μmol/L in cholestasis group. The differences between 2 groups were statistically significant (Z=2.955, P=0.003; t=3.214, P=0.006). ③The sensitivities of kaoline stool, GGT level, ultrasonic examination results indicating that gallbladder contraction was less than 50%, and gallbladder with shrinkage and stiffness or without gallbladder, and ultrasound examination results indicating hepatic portal triangular fibrosis in diagnosis of BA were 55.6%, 76.4%, 75.0% and 17.6%, respectively, and the specificities were 81.1%, 86.5%, 91.9% and 95.9%, respectively, and the accuracy were 72.7%, 83.3%, 86.4% and 71.3%, respectively, and the positive predictive values were 58.8%, 72.2%, 81.8% and 66.7%, respectively, and the negative predictive values were 78.9%, 88.9%, 88.3% and 71.7%, respectively.

Female children, kaoline stool, GGT level ≥300 U/L, TBIL concentration ≥150.0 μmoL/L, DBIL concentration ≥100.0 μmoL/L, ultrasonic examination results showed that poor gallbladder contraction or stiffness, shrinkage or absence of gallbladder, hepatic portal triangle fibrosis are all high risk factors for BA patients younger than 3 months old, but those items still cannot be used as the basis for diagnosis of BA. Cholangiography is required for patients younger than 3 months combined with the above-mentioned high-risk factors of BA to make final diagnosis.