Province-wide Biliary Atresia Home Screening Program in British Columbia: Evaluation of First 2 Years.

BACKGROUND AND OBJECTIVES: Biliary atresia (BA), a rare newborn liver disease, is the leading cause of liver-related death in children. Early disease recognition and timely surgical Kasai hepatoportoenterostomy (KP) offers long-term survival without liver transplant. Universal BA screening in Taiwan using infant stool color cards (ISCCs) has proven effectiveness. We report our experience with infant stool color card (ISCC) BA screening in a province-wide program in British Columbia (BC). The objective of this study is to assess program performance and cost from launch April 1, 2014 to March 31, 2016. METHODS: ISCCs distributed to families upon maternity ward discharge. Parents were instructed to monitor their infant's stool color for 1 month and contacted the screening center with concerns. The number of live births, ISCC distribution, BA cases, and costs were recorded. Cases with Program screen success had both acholic stool recognition (ISCC screen success) and timely referral for BA. RESULTS: All 126 maternity units received ISCCs. Of 87,583 live births there were 6 BA cases. Of the 5 cases with ISCC Screen Success 3 had Program Screen Success. The median KP age in the program screen success and failure groups was 49 (42-52) and 116 (49-184) days, respectively. Program sensitivity was 50%, specificity 99%, positive predictive value 4%, and negative predictive value 99%. A random sample of 1054 charts at BC Children's Hospital found an ISCC distribution rate of 94%. After a phase-in period, the annual program cost was $30,033.82, and the ISCC cost per birth was $0.68. CONCLUSIONS: The screening program has high specificity and distribution with low cost. Successful program case identification had earlier age at KP. Program modifications aim to improve sensitivity. Longer-term studies will determine program impact on health outcomes.

Cost-effective analysis of screening for biliary atresia with the stool color card.

BACKGROUND: Biliary atresia (BA) is the leading cause of pediatric end-stage liver disease and liver transplantation in the United States. Early diagnosis leads to improved outcomes, but diagnosis is often delayed, leading to increased rates of transplantation and mortality. METHODS: A Markov model was developed to simulate the natural history and transplant-related outcomes of patients with BA in a US cohort studied for 20 years. Data regarding proportions of individuals in different health states, including transplant and death, were obtained from published literature. Costs were derived from the literature and the Johns Hopkins database of charges using the cost-to-charge ratio. Strategy A represented the status quo and assumed no screening. Strategy B used nationwide screening with the stool color card developed by the Taiwan Health Bureau. The cost associated with both strategies was compared with the number of life-years gained, deaths, and the number of transplants for a 20-year interval. A dominant strategy was one that was associated with lower cost alongside improved outcomes, including increases in life-years gained, reductions in number of deaths, and reductions in number of transplants. One-way and probabilistic sensitivity analyses were performed. RESULTS: In strategy A, the 20-year cost was $142,479,725 with 3702 life-years, 74 deaths and 158 liver transplants. For strategy B, the cost was $133,893,563 with 3731.7 life-years, 71 deaths and 147 liver transplants. There was a >97% probability that screening with the stool color card would be cost saving and associated with an increase in life-years gained. Among all parameters, only stool color card specificity was associated with the potential for screening to no longer be cost saving. CONCLUSIONS: Compared with no screening, screening with the stool color card is a dominant strategy associated with lower costs and better outcomes. These findings suggest that screening with the stool color card could be an important, economically feasible strategy for improving outcomes in BA in the United States.

Outcome of hepatobiliary scanning: preterm versus full-term cholestatic infants.

OBJECTIVES: The aims of this study were to evaluate the specificity of a non-draining hepatobiliary scintigraphy (HBS) for biliary atresia (BA) in preterm and full-term babies, to verify the relationship between non-draining scan and higher levels of direct bilirubin and to find an objective criterion to guide the time in performing HBS. METHODS: A total of 175 infants (113 males and 62 females, median age of 45 days) with 181 HBS performed in Tuen Mun Hospital between January 1998 and May 2010 were retrospectively analysed. A 'non-draining' scan was defined as one showing no excretion of radiolabelled tracer into the small bowel 24 h after injection. The disease category, epidemiological and laboratory data were compared between infants having non-draining and draining scans. In addition, the predictive value of a negative scan for BA was compared between preterm and full-term infants. RESULTS: Twenty infants (11.4%) were surgically confirmed to have BA. A non-draining scan was found to be 100% sensitive for BA, and the specificity was 96% and 78% among full-term infants and preterm infants, respectively. The mean direct bilirubin values of infants with BA and intrahepatic cholestasis were 141.9 and 111.3 µmol/L, respectively, which were significantly higher than 67.2 µmol/L seen in infants with draining scans. This analysis shows that using direct bilirubin ≥63 µmol/L as an objective criterion in guiding the time to perform HBS is most cost-effective. CONCLUSION: Our data supported that using direct bilirubin ≥63 µmol/L as an objective criterion in guiding the time to perform HBS will avoid unnecessary scans.

Living donor liver transplantation in children: a single North American center experience over two decades.

Little data concerning hospital charges and long-term outcomes of LDLT in North American children according to transplant indications have been published. To compare outcomes of patient and graft survival and healthcare charges for LDLT for those with BA vs. other diagnoses (non-BA). A retrospective review of 52 children receiving 53 LDLT (38 BA and 14 non-BA) from 1992 to 2010 at our institution was performed. One-, five-, and 10-yr patient and graft survival data were comparable to national figures reported to UNOS. Average one-yr charges for recipients and donors were $242 849 for BA patients and $183 614 for non-BA (p = 0.074). BA patients were 1.23 ± 1.20 yr of age vs. 4.25 ± 5.02 for non-BA, p = 0.045. Examination of the total population of patients who were alive in 2010 in five chronological groupings showed that the crude five-yr survival rates were 1992-1995: 9/11 (82%); 1995-1997: 6/10 (60%); 1997-1999: 8/10 (80%); 1999-2001: 9/10 (90%); and 2001-2003: 7/7 (100%). Thus, examination of the clinical and financial data together over the entire period of the transplant program suggests that the dramatic improvement in patient survival was accomplished without a dramatic increase in indexed charges. All 53 donors survived, and only 10% had complications requiring hospitalization. LDLT in children results in excellent outcomes for patients and donors. Ways to lower costs and maximize graft outcome should be investigated.

Impact of pretransplant growth retardation in pediatric liver transplantation.

OBJECTIVE: Malnutrition frequently complicates end-stage liver disease and orthotopic liver transplantation (OLT) in pediatric patients. Pretransplant malnutrition has been associated with increased post-OLT mortality and length of stay in adults. The relationship between pre-OLT nutritional status and post-OLT outcomes in pediatric liver transplant recipients, however, is not well studied. METHODS: The records of 65 pediatric patients who underwent OLT at a single institution were reviewed. Univariate analyses were used to investigate the relationship between anthropomorphic data (expressed as z-scores) and post-OLT hospital length-of-stay, hospital costs and clinical outcomes. A multivariate model was then used to identify peri-OLT variables independently correlated with post-OLT length-of-stay. RESULTS: A decreased height z-score was correlated with an increased post-OLT hospital length-of-stay (r = -0.30; P = 0.015) and increased hospital costs (r = -0.49; P = 0.0004). The mean length-of-stay was 20.5 days for patients with a height z-score of <-1.5 and 10.7 days for patients with a height z-score of >1.5 (P = 0.038). Likewise, hospital costs were about $40,000 higher (25% increased) for patients with growth retardation. A weak direct correlation was seen between weight z-score and post-OLT length-of-stay (r = 0.18; P = 0.15). Height z-score, biliary atresia and pre-OLT protime were independently and significantly correlated with post-OLT length-of-stay in a multivariate model. CONCLUSIONS: Height z-score is a better indicator of pretransplant malnutrition than weight z-score. Pretransplant growth retardation is associated with increased post-OLT hospital length-of-stay and increased hospitalization costs.