Hepatic and non-hepatic hydrothorax in pediatric ascites.

BACKGROUND: Hydrothorax in the presence of ascites is a serious condition, but it is not well studied, particularly in pediatrics. We aim to identify risk factors for having hydrothorax, compare morbidity and mortality, and report the prevalence of hepatic hydrothorax and non-hepatic hydrothorax in pediatric patients with diagnosis of ascites and hydrothorax. METHODS: This is a retrospective study of pediatric patients under 22 years of age with both ascites and hydrothorax. Hydrothorax was categorized into hepatic and non-hepatic hydrothorax. Demographic data and clinical data including ascites grade, ascites etiology, treatments, length of stay, and death were collected and analyzed using logistic regression. RESULTS: We identified 120 patients with ascites and hydrothorax, 63 (53%) being female. The median age was 13 years (IQR: 4-18). Patients 6 years of age or older (OR=1.90; 95% CI=1.16-3.17; p = 0.012), patients with higher grades of ascites (OR=1.77; 95% CI=1.27-2.47; p < 0.001), those treated with furosemide (OR=2.27; 95% CI=1.37-3.76; p = 0.001), and those with hepatorenal syndrome (OR=4.22; 95% CI=1.19-15.63; p = 0.025) had increased risk of hydrothorax. The underlying etiology of ascites was not associated with mortality, but it was associated with length of stay (p = 0.013), with veno-occlusive disease being the largest contributor. Hepatic versus non-hepatic hydrothorax was also not found to be associated with mortality, but length of stay was significantly greater in former (23 days; IQR=13-38) compared to the latter group (14 days; IQR=8-26) (p = 0.009). CONCLUSIONS: With pediatric ascites, there are  certain risk factors that are associated with having hydrothorax, and ascites etiology may be associated with morbidity.

Clinical Predictors of Morbidity and Mortality in Hospitalized Pediatric Patients With Ascites.

OBJECTIVES: Ascites is a pathologic buildup of fluid in the peritoneal cavity. Knowledge is lacking in clinical outcome in pediatric patients with ascites. We aim to identify and assess clinical variables, associated with morbidity and mortality in pediatric patients who are hospitalized with ascites. METHODS: A retrospective cohort study was performed on patients ages 0 to 21 hospitalized at Johns Hopkins Hospital between 1983 and 2010 with an ICD-9 discharge diagnosis of ascites (789.5, 789.51, 789.59). A total of 518 pediatric patients were studied, all with a diagnosis of ascites during hospitalization. Study outcomes included hospital length of stay (LOS) as a proxy for morbidity and death at hospital discharge for mortality. Variables analyzed included demographic data, ascites etiology and grade, comorbidities, and laboratory markers. Variables were analyzed by log-linear regression and competing risk model. RESULTS: Among the 3 age groups (0-5, 6-12, and 13-21), the 0 to 5 age group experienced significantly increased LOS (P < 0.001) and mortality (P = 0.027). Ascites etiology of veno-occlusive disease (VOD) and the presence of hydrothorax or thrombocytopenia was also significantly associated with increased LOS. Ascites with the etiology of congestive hepatopathy and the presence of grade 3 ascites, hepatic encephalopathy, hepatorenal syndrome, hydrothorax, hyponatremia, and thrombocytopenia were associated with increased mortality. Additionally, black pediatric patients have an increased risk of mortality (P = 0.027). Other factors including sex, leukopenia, portal vein thrombosis, and splenomegaly were not associated with LOS or mortality. CONCLUSIONS: Morbidity and mortality in pediatric patients hospitalized with ascites are associated with specific demographic and clinical factors. Further studies are required to apply this knowledge to predict the clinical outcomes.

Clinical usage of serum albumin to ascitic fluid albumin gradient and ascitic fluid total protein in pediatric ascites.

BACKGROUND: Abdominal paracentesis is performed as a diagnostic test in children with ascites. Serum albumin to ascitic fluid albumin gradient (SAAG) is frequently used in adults to distinguish types of portal hypertension. We aim to investigate the utilization of SAAG and other biomarkers in determining the etiology of significant ascites in children. METHODS: In this retrospective study, children aged 0-21 years with significant ascites were identified using International Classification of Diseases, Ninth Revision (ICD-9) codes and medical records during the period 1983-2010. Medical records of children who had abdominal paracentesis were examined in detail. RESULTS: 207 children had significant ascites and of those children, 20 (9.6%) had abdominal paracentesis. Our data showed that high albumin gradient (SAAG ≥ 1.1 g/dL) differentiates causes of ascites secondary to portal hypertension (cirrhosis, hepatic vein outflow obstruction, or congestive hepatopathy) from other causes. In addition, ascitic fluid total protein (AFTP) may help in differential diagnosis of ascites. Children with high SAAG manifest clinical features of portal hypertension including esophageal varices or variceal hemorrhage. CONCLUSION: Among patients with initially unclear causes of ascites, SAAG and AFTP can provide guidance for appropriate investigations.

A longitudinal assessment of non-invasive biomarkers to diagnose and predict cystic fibrosis-associated liver disease.

BACKGROUND & AIMS: A practical, inexpensive, and non-invasive biomarker of liver fibrosis is needed as a reliable screening test for cystic fibrosis-associated liver disease (CFLD). Studies have shown the utility of AST to Platelet Ratio Index (APRI), fibrosis index based on 4 factors (FIB-4), and gamma-glutamyl transferase (GGT) as good biomarkers for identifying CFLD. The goal of the study was to evaluate the effectiveness of APRI, FIB-4, AST/ALT ratio, platelet count, GGT, and GGT platelet ratio (GPR) in predicting CFLD development. METHODS: Data was collected from CF Foundation Patient Registry for patients aged 3-21 years at Johns Hopkins from January 1, 2002 to December 31, 2014. Collected data included demographic characteristics, presence of splenomegaly, hepatomegaly, ascites, and variceal bleeding, AST, ALT, GGT, platelet count, and FEV1. The sensitivity and specificity of each biomarker were analyzed and reported by the area under receiver operating characteristic (AUROC) curve. RESULTS: By the end of the study, 144 "healthy" CF, 12 CFLD, 19 CF-associated pulmonary disease (CFPD), and 4 CFLD with CFPD cases were identified. APRI scores were higher in CFLD, 0.85 versus 0.28 in "healthy" CF and 0.23 in CFPD groups (p<0.001). GPR had the highest AUROC curve at 0.91. CONCLUSIONS: GPR, GGT, APRI score, and platelet count were potentially useful biomarkers while FIB-4 did not predict CFLD development. Cost-effectiveness studies are needed to analyze the utility of these biomarkers in clinical practice.

Ascites in Children: A Single-Center Experience of 27 Years.

OBJECTIVES: The aim of our study was to describe the changing prevalence, demographic features, etiologies, and treatment of ascites in children hospitalized during a 27-year period at the Johns Hopkins Hospital (Baltimore, MD). METHODS: We retrospectively reviewed discharges from 1983 to 2010 to select patients whose records included a diagnosis of ascites. We assessed the etiologies and degrees of ascites (ascites grade 1 detectable only by radiologic tests; ascites grades 2 and 3 recognized by moderate and marked abdominal distension by physical examinations). RESULTS: We classified 518 children into 9 etiology groups: intrahepatic disease (IH) (105), hepatic vein outflow obstruction (HVOO) (45), congestive heart disease (CH) (33), nephrotic syndrome (NS) (36), pancreatitis (26), inflammatory and infectious diseases (77), malignancy (49), idiopathic (71), and miscellaneous (76). IH and CH were predominant in the younger age group (0-5 years) versus HVOO, pancreatitis, and malignancy in the older age group (13-21 years) (P < 0.001). The prevalence of ascites increased over time from 1983 to 2006 and declined thereafter. Ascites grade 1 was more common than ascites grades 2 and 3 in all the groups (P = 0.048). IH and NS were more likely to have ascites grade 2 and 3 (P = 0.02). Although spironolactone was more frequently used in the IH group versus other etiologies, furosemide was used more frequently in NS and CH versus other etiologies (P < 0.001). CONCLUSIONS: The increased prevalence of ascites during the initial study period could reflect improved detection radiologic detection. The proportion of severe ascites and the various medical treatments differed among the etiologic groups.

Predictors of severity in childhood pancreatitis: correlation with nutritional status and racial demographics.

OBJECTIVE: Acute pancreatitis is one of the leading causes of rising pediatric hospitalizations in North America. The aim of this study was to assess the role of nutritional status and racial influences on the severity of acute pancreatitis in children. METHODS: The institutional review board approved this retrospective chart review of children with the diagnosis of acute pancreatitis between the ages of 0 and 18 years hospitalized at the Johns Hopkins Hospital between 1998 and 2008. Parameters studied included biochemical markers associated with pancreatitis, review of severity of illness reflected through the length of stay, and pediatric intensive care unit admission. RESULTS: The length of in-patient hospitalization was longer for children with imaging findings of pseudocyst or pancreatic necrosis (23.1 ± 26.4 days vs 4.4 ± 10.6 days; P = 0.0074) and malnourished children versus normal weight and obese children (16.5 days for malnourished vs 10.6 days for normal weight vs 10.7 days for obese; P = 0.04). There was also a significant difference in the need for pediatric intensive care unit admission across ethnic groups (18% African American vs 7% white) (P = 0.04). CONCLUSIONS: Ethnicity and nutritional status may influence the severity and duration of hospitalization among children with pancreatitis.