Biliary atresia (BA) is an idiopathic fibrosclerosing obliterative disease of the large intrahepatic and extrahepatic bile ducts (1–3). This uniquely paediatric liver disease manifests exclusively during the neonatal period, in the first 2 to 4 weeks of life, with persistent jaundice due to a conjugated hyperbilirubinaemia and pale acholic stools. BA is a leading cause of cirrhosis and liver disease–related death in children. It is also the most common indication for liver transplantation in the paediatric population. At least 2 BA clinical phenotypes have been recognised, yet the present standard of care for affected patients, regardless of phenotype, is sequential surgical treatment with an initial hepatoportoenterostomy, also known as the Kasai procedure (KP), at the time of diagnosis in an effort to restore the bile flow, followed by liver transplantation for those patients who progress to liver failure.
The survival of patients with BA with their native liver, without the need for liver transplantation, depends largely on the success of the initial KP. The most important and well-established prognostic factor for the KP is the patient age at the time of the KP (4–7). Previous studies in Canada and elsewhere have shown that when the KP is performed within the first month of life, the patient survival with his or her own native liver may be as high as 50% (6,8,9). In contrast, the native liver survival is <20% for those cases who present late, having the KP after 90 days of age. Several other strategies aimed to improve the KP prognosis are related to postoperative patient management. These include administration of ursodeoxycholic acid to stimulate bile flow, aggressive nutritional support and fat-soluble vitamin replacement to maximise growth potential, and the use of antibiotics as prophylaxis for ascending cholangitis (10). Whether steroid therapy is beneficial to KP outcome remains controversial: a multicentre randomised trial is under way in the United States to evaluate the effect of short-term adjuvant corticosteroid therapy on the post-KP native liver survival (11).
Of considerable interest are the recent national reports from the United Kingdom and France demonstrating that paediatric centres managing larger BA caseloads, >5 cases per year (UK) and ≥20 cases per year (France), have significantly improved post-KP patient survival with their own native liver compared with those managing only a few cases [<5 cases per year (UK) or <2 cases per year (France)] (4,5,12,13).
In Canada, paediatric patients with BA are customarily managed at university-based paediatric hospital centres. We recently established a national database to evaluate the experience with BA in Canadian children (8). The primary aim of the present study was to determine the role of centre experience in the outcome of children with BA in Canada. A secondary aim was to assess for factors that may influence the timing of the KP among the treatment centres.
PATIENTS AND METHODS
Under the auspices of the Canadian Pediatric Hepatology Research Group, the medical records of patients with BA, born between 1 January 1992 and 31 December 2002, who were diagnosed and followed at 1 of 12 Canadian university-based paediatric tertiary care institutions were reviewed (Appendix 1) (8). This cohort represents a subset of the patients entered into our large Canadian database of patients with BA diagnosed between 1985 and 2002. The 1992 to 2002 subset was chosen to allow comparison of our data with data from other published studies.
BA cases were identified through the health record department at each respective institution using the standard national numerical coding systems for the diagnosis of BA or the surgical intervention of a KP. Data collected from each chart as detailed elsewhere (8) included date of birth, age at referral, type of imaging studies, liver histology, whether a Kasai operation was performed, age at Kasai operation, whether a liver transplant (LT) was performed, date of transplantation, date of last follow-up, and final outcome. All of the data were recorded onto a uniform spreadsheet by the investigator at each centre and then, after eliminating patient identifiers, was electronically transferred to the central registry at BC Children's Hospital. The database was critically reassessed by 1 investigator (R.S.) to ensure accuracy and completeness and then further anonymised. The diagnosis of BA was confirmed based on standard clinical, biochemical, radiological, histological, and operative findings. Despite the retrospective nature of this study, complete data collection was possible in most cases.
The study patients were first analysed as a single group. They were then grouped according to centre size. The centres were initially grouped into A, B, or C, respectively, if they managed on average <1, 1 to 3, or >3 cases annually. Patients were then regrouped by centre size managing >5 or <5 cases annually (there were no centres managing 5 cases per year) to facilitate direct comparison with previously published data. Survival curves were calculated according to the Kaplan-Meier method with overall patient survival defined as starting at birth and ending at death or last follow-up; liver transplant (LT) survival beginning at transplant and ending at death or last follow-up; and native liver survival starting at birth and ending at death, transplant, or last follow-up. Results were expressed as calculated survival rate with 95% confidence intervals (CI). The proportional hazard ratio for the Kaplan-Meier curves was verified graphically. Outcome analyses were also performed after stratifying the patients for age at the time of KP at less than or equal to 30, 45, 60, and 90 days. These ranges for patient age at the time of the KP were chosen based on the most favourable prognostic age cutoffs for post-Kasai native liver survival previously reported in Canada, France, and Taiwan (4,8,13). Univariate analysis used the rank sum and log-rank tests. Comparisons between the entire cohort and the centre-size groups were made to see whether outcomes differed between them. Median and ranges were measured for the various time variables because the data had a nonparametric distribution. The t tests and the χ2 tests were used to compare descriptive measures and categorical data, respectively, between the groups. All of the significance tests were 2-tailed with alpha fixed at 0.05. All of the statistical analyses were done with Stata software (version 8.1, StataCorp, College Station, TX). The study received approval from each respective university and hospital ethical review board.
There were 230 patients identified (135 were girls; girls:boys, 1.42:1). Of the 12 university-based hospitals there were 6 group A-, 4 group B-, and 2 group C-size centres managing 36, 74, and 120 cases, respectively, through the study period (Table 1). The overall median patient age at last follow-up was 66 months (4–170 months) with no significant difference in the length of follow-up between the centre-size groups. The KP was performed in 90% of the cases overall, and there was no difference in the percentage of cases having the KP between the centre groups. The other 10% of the patients were directly listed for transplant (Fig. 1, in the online-only Supplemental Digital Content, available at http://links.lww.com/MPG/A17). Through the study period, 4 patients were lost to follow-up, all from group A centres. Each of the 4 patients had received a KP and 3 of them had undergone LT. The median patient age at the time of Kasai was 56 days in the group B centres, significantly earlier than in groups A and C at 66 days (P = 0.05). Of note, although the median length of time from initial presentation to the Kasai operation was 8 days for groups A and B, it was significantly longer at 12 days for group C (P = 0.05). Although the diagnostic evaluation, which included an abdominal ultrasound, hepatobiliary scanning, and liver biopsy, was similar between the centre groups, hepatobiliary scanning was performed less often at the group C centres, whereas ultrasound imaging was completed in only 80% of the patients in group B centres. LT was ultimately required in 60% of the cases overall, with no difference in the frequency of transplant or the median age at transplant between the sized-centre groups (Table 1).
The overall 4-year patient survival was 83% (95% CI 78–88) without any significant difference amongst the groups (Fig. 1). This survival rate was durable, remaining essentially unchanged with only 1 death in 10 years. The overall LT recipient survival at 4 years was 84%. The 4-year post-Kasai native liver survival was 37%, 35%, and 43% for the size A, B, and C centres, respectively, with the overall survival curves not being significantly different (P = 0.54). Of the patients alive with their native liver at 4 years, 81% had ongoing survival without requiring transplantation through the next 5-year follow-up period.
Only 1 centre in Canada managed >5 cases annually (Toronto). Following the regrouping of patients into the largest Canadian centre (>5 cases per year) and the other smaller centres (<5 cases per year), there were no significant differences in the overall patient survival or post-KP native liver survival between these groups (Fig. 2). However, we noted a trend towards improved native liver survival in the largest Canadian centre compared with the other centres. Because patient age at the time of KP is an important prognostic factor for the post-KP native liver survival, we more closely examined these cohorts by stratifying for the patient age at the time of KP. We noted significantly improved post-KP native liver survival in the largest centre for those patients having their KP at less than or equal to 60 and less than or equal to 90 days of age (Table 2).
The influence of hospital caseload experience on patient outcomes has attracted intense interest for those designing strategies for the best delivery of health care. Strongest associations between high patient volume and improved outcomes have been observed for HIV treatment, oesophageal and pancreatic cancer, congenital heart disease, and organ transplantation (14). For BA, recent studies from the United Kingdom and France demonstrated improved patient outcomes in those centres managing a larger patient load. In the United Kingdom through the 1990s, the overall 4-year post-KP native liver survival rate was 30%, with the rates being highest (60%) in the 2 centres managing the largest patient loads of >5 cases per year (5). In 1999, the United Kingdom instituted a national centralised policy for the care of all BA cases, with referral and management restricted to 3 liver units, in London, Birmingham, and Leeds. A recent follow-up study has shown that the overall 5-year post-KP native liver survival subsequently improved to 48% (15). In France, the introduction of a decentralised policy for BA care, in which the smaller centres receive consultative support from the larger ones, led to an improvement in overall patient survival largely because of better LT survival rates, because the post-KP native liver survival rate did not change significantly (13).
It is important to examine such strategies in other countries. In this retrospective study, we found no significant differences in the overall patient, LT, or post-KP native liver survivals based on centre caseload experience among treatment centres in Canada. Several factors may account for our findings. Paediatric care for disorders such as BA is, in effect, already highly centralised in Canada. These complex cases are managed almost exclusively at university-based tertiary or quaternary paediatric centres. In contrast, before centralisation in the United Kingdom, children were managed in 15 individual surgical centres with only 2 centres treating >5 cases per year (5). In France, introduction of a decentralised policy for BA care led to a reduction in the overall number of BA centres from 30 to 22, yet 80% of the centres still manage <2 cases per year with only 3 centres managing >3 patients per year (13). Liver transplantation is also highly centralised in Canada, with only 4 regional paediatric transplant programmes across the country. Thus, in Canada, unlike elsewhere in North America or Europe, treatment of children with BA is provided almost entirely by paediatric medical and surgical subspecialists at large academic institutions, and not at community or regional hospitals. National health insurance policies in Canada guarantee health care regardless of economic or social status. Consequently, every newborn infant in our country has universal access to complex care. Indeed, the post-KP native liver survival in Canada, even for those patients managed in the smallest centres (37%), is much better than the rate (14%) previously reported from the smaller centres in the United Kingdom (<5 cases per year) before introduction of a centralisation policy (Table 3). Canadian rates are also similar to the rates in France either before or after decentralization, or the rates in other centres in Europe and elsewhere (4,7,13). In this context, advocacy for a more centralised biliary care policy in Canada, similar to the current policies in the United Kingdom or France, would be difficult to justify.
In the present Canadian study, the median patient age at KP was 64 days (range 6–173 days) compared with 54, 55, and 61 days in recent US, UK, and France studies, respectively (4,15,16). We found that the median age of Kasai differed with the size of treatment centres. Within the group C centres, the time from the initial consultation to the Kasai operation was significantly longer, perhaps because of logistical issues including the scheduling of investigative procedures and the demands for operating room time. Therefore, any potential outcome advantage for the larger centres may have been offset by the older median age at Kasai. Indeed, after stratifying for age at Kasai we did find significantly improved native liver survival only among those cases having their Kasai at less than or equal to 60 and less than or equal to 90 days in the largest centre. We did not observe any differences in outcome among the groups who had Kasai at an earlier age, although here the sample sizes were small and lack of significance may have been subject to beta error.
The trend towards significantly improved post-KP native liver survival in the largest Canadian centre, which may have as many as 8 cases per year, raises the possibility that caseload volumes as high as 20 cases per year may not be required to achieve best outcomes. Further studies to determine the minimum caseload volumes for optimal post-KP native liver survival may prove beneficial, especially to support BA care policy in those jurisdictions where these minimum ranges could be realistically achieved.
Another confounding factor, which we were unable to examine in this retrospective study, is whether having a dedicated surgeon and medical team at any given treatment centre contributes to the best outcomes, as is currently practiced in the United Kingdom. Even within the larger Canadian centres several different surgeons perform the KP, and the postoperative care, although similar, is not uniform.
Canada differs from the United Kingdom and France in being geographically a large country. The population is highly dispersed and comparatively small (UK and France: 60 million; Canada: 33 million). Based on the 1:19,000 incidence we reported for BA in Canada, we estimate 18 to 22 new cases each year (8). For caseload experience in Canada to approximate the UK experience (40–45 new cases per year) there would have to be only 1 centre with a dedicated team of paediatric surgeons and hepatologists managing all of the patients in the country. Implementation of such a highly centralised programme for the care of infants with BA in Canada would be extremely difficult given the great distances that patients and parents would be required to travel expeditiously for the initial KP, as well as the scheduling demands that would exist between the referral and treatment centres. Even if the entire diagnostic assessment were performed at 1 academic centre and the surgery and postoperative management were performed at another, preceded by only the most cursory clinical reevaluation, unnecessary delays would be introduced into the management of BA precisely at the time when delays affect most adversely prognosis. We believe that similar challenges exist in other health care jurisdictions including regions of the United States, Australia, and some South American countries, but this possibility requires separate local assessment. For Canada, a centralised policy to ensure caseload volumes at levels currently in place in the United Kingdom is impractical for Canadian patients with BA. Our data suggest, moreover, that it is not imperative.
Given the major logistical challenges associated with a centralised policy in Canada and the trend already noted towards unduly late referral (8), we need to look at alternate ways to improve the outcomes for BA in our country. National public health strategies aimed to prompt earlier referral for patients with BA and shift the current timing of the KP to an earlier age would have considerable influence on native liver survival outcomes (17). Most promising is the recent Taiwan experience in which introduction of an infant stool colour card system as a universal screening tool for BA has virtually eliminated occurrence of KP performed beyond 90 days of age (18). Introduction of national BA programmes to educate both parents and physicians towards earlier disease recognition, to initiate timely testing of total and conjugated serum bilirubin for all of the infants with persistent jaundice beyond 2 weeks of age, and to develop a stool colour card screening programme as reported by the group in Taiwan may prove to be the most straightforward, effective, and important intervention for Canadian patients with BA. Further studies that aim to examine the cost benefit and efficacy of these intervention strategies for timely BA diagnosis are needed.
In summary, caseload experience did not affect the post-KP native liver survival rates for Canadian children with BA. Although the majority of paediatric university-based hospitals in Canada are low-volume centres, our findings do not support the referral of patients to the larger Canadian centres to improve outcome. Indeed, such a policy would impose substantial hardship on Canadian patients and their families. Our study conclusions are not meant to refute the UK or French results showing that large caseload volume improves outcome for BA. Instead, we conclude that in a large country such as Canada, with a relatively small population, nationwide centralisation of surgical and medical management of BA to approximate the few UK or French centres having extremely high caseload volumes and best outcomes is almost impossible. Moreover, advocacy for public health policy to create 1 or 2 large-volume centres in Canada is not required because our current national results with KP are comparable to those from other countries. Outcomes in Canada, however, could be further improved by earlier referral of infants with BA for treatment. For countries such as Canada, studies should be performed to evaluate screening programmes whose principal objective is to prompt earlier disease recognition, in an effort to eliminate late referrals and delayed treatment for BA.
1. Balistreri WF, Grand R, Hoofnagle JH, et al
. Biliary atresia
: current concepts and research directions. Summary of a symposium. Hepatology 1996; 23:1682–1692.
2. Petersen C. Pathogenesis and treatment opportunities for biliary atresia
. Clin Liver Dis 2006; 10:73–88.
3. Schreiber RA, Kleinman RE. Biliary atresia
. J Pediatr Gastroenterol Nutr 2002; 35:S11–S16.
4. Chardot C, Carton M, Spire-Bendelac N, et al
. Prognosis of biliary atresia
in the era of liver transplantation: French national study from 1986 to 1996. Hepatology 1999; 30:606–611.
5. McKiernan PJ, Baker AJ, Kelly DA. The frequency and outcome of biliary atresia
in the UK and Ireland. Lancet 2000; 355:25–29.
6. Nio M, Ohi R, Miyano T, et al
. Five- and 10-year survival rates after surgery for biliary atresia
: a report from the Japanese Biliary Atresia
Registry. J Pediatr Surg 2003; 38:997–1000.
7. Petersen C, Harder D, Abola Z, et al
. European biliary atresia
registries: summary of a symposium. Eur J Pediatr Surg 2008; 18:111–116.
8. Schreiber RA, Barker CC, Roberts EA, et al
. Biliary atresia
: The Canadian experience. J Pediatr 2007; 151:659–665.
9. Serinet MO, Wildhaber BE, Broué P, et al
. Impact of age at Kasai operation for biliary atresia
screening. Pediatrics 2009; 123:1280–1286.
10. Meyers RL, Book LS, O'Gorman MA, et al
. High-dose steroids, ursodeoxycholic acid, and chronic intravenous antibiotics improve bile flow after Kasai procedure
in infants with biliary atresia
. J Pediatr Surg 2003; 38:406–411.
12. Davenport M, De Ville de Goyet J, Stringer MD, et al
. Seamless management of biliary atresia
in England and Wales (1999–2002). Lancet 2004; 363:1354–1357.
13. Serinet MO, Broue P, Jacquemin E, et al
. Management of patients with biliary atresia
in France: results of a decentralized policy 1986–2002. Hepatology 2006; 44:75–84.
14. Myers RP, Papay KD, Shaheen AA, et al
. Relationship between hospital volume and outcomes of esophageal variceal bleeding in the United States. Clin Gastroenterol Hepatol 2008; 6:789–798.
15. Ong E, Sharif K, Alizai N, Kelly D, McClean P, Davenport M. Biliary atresia
in the 21st century—the centralization experience. Paper presented at the British Association of Paediatric Surgeons meeting, June 17–20, 2009 Graz, Austria.
16. Shneider BL, Brown MB, Haber B, et al
. A multicenter study of the outcome of biliary atresia
in the United States, 1997 to 2000. J Pediatr 2006; 148:467–474.
17. Chitsaz E, Schreiber RA, Collet JP, et al
. Biliary atresia
: the timing needs a changin'. Can J Public Health 2009; 100:475–477.
The Canadian university-based paediatric tertiary care centres were The Janeway Child Health and Rehabilitation Centre, Memorial University; the IWK Health Centre, Dalhousie University; Centre Hospitalier Universitaire Sainte-Justine, Université de Montréal; the Montreal Children's Hospital, McGill University; The Children's Hospital of Eastern Ontario, University of Ottawa; the Hamilton Health Sciences Center, McMaster University; The Hospital for Sick Children, University of Toronto; Children's Hospital, The University of Manitoba; the Royal University Hospital, University of Saskatchewan; Alberta Children's Hospital, University of Calgary; Stollery Children's Hospital, University of Alberta; and BC Children's Hospital, University of British Columbia.
18. Hsiao CH, Chang MH, Chen HL, et al
. Universal screening for biliary atresia
using an infant stool color card in Taiwan. Hepatology 2008; 47:1233–1240.
biliary atresia; Kasai procedure; liver transplant; newborn health
Supplemental Digital Content
Copyright 2010 by ESPGHAN and NASPGHAN