Variceal hemorrhage is an important complication of chronic liver disease or portal vein obstruction in children (1–8). Its incidence among children with biliary atresia ranges from 17% to 29% in a 5- to 10-year period (3,4,9) and is 50% in children who survive more than 10 years without liver transplantation (6). The associated mortality ranges from 2.5% to 20% (2,4,7,8).
For adults with medium or large varices, nonselective β-blockers and endoscopic variceal ligation (EVL) offer effective prevention of a first episode of variceal bleeding (primary prophylaxis) (10,11). Practice guidelines in the United States and Europe (12,13) recommend endoscopy at the time of diagnosis of cirrhosis and subsequent prophylactic treatment for patients with medium or large varices.
There are little pediatric data to help determine the effectiveness of primary prophylaxis of variceal bleeding in children or to guide therapeutic choices for the management of acute variceal bleeding or secondary prophylactic therapy to prevent rebleeding in children (14,15). In a previous survey of 30 pediatric gastroenterologists in the United States, approaches to screening endoscopy and primary prophylaxis of variceal bleeding varied considerably among the respondents (16). Variation of care provided by physicians for a patient with a specific clinical presentation suggests underuse, overuse, or misuse of tests or therapies by some practitioners and is therefore a marker of poor quality of care (17,18).
In the present study, we aimed to describe the approaches taken by pediatric gastroenterologists and hepatologists to the diagnosis and management of esophageal varices in children with portal hypertension, and to determine the attitudes of children with portal hypertension, their families, and their physicians to screening endoscopy for quantification of variceal hemorrhage risk.
A questionnaire was constructed to evaluate physicians' approaches to screening for esophageal varices, primary prophylactic therapy, acute management of variceal bleeding, and secondary prophylaxis of variceal bleed. Approval was obtained from the author of a previous survey of pediatric gastroenterologists in the United States to use certain questions from that survey to facilitate comparison of results (16). All of the pediatric hepatologists and gastroenterologists currently working in Canada were contacted via e-mail and asked to undertake the questionnaire via an Internet-based survey instrument. Response rates were optimized by sending 3 reminder e-mails in 2- to 4-week intervals after the initial request. No incentives for survey completion were provided.
The questionnaire comprised 19 questions and was pretested among the co-investigators to ensure clarity of the questions. It was divided into 3 sections. The first section included 3 questions about the gastroenterologist's professional experience; the second section included 12 questions on the gastroenterologist's approach to screening and treatment of esophageal varices and variceal bleeding; and the third section included 4 questions about attitudes to the risks and benefits of endoscopy. The total time required for completion of the questionnaire was estimated at 15 minutes.
We also questioned consecutive children and their parents with portal hypertension caused by cirrhosis or portal vein thrombosis at risk of varices who attended the Liver Clinic of the Hospital for Sick Children, Toronto, on their attitude toward the risks and benefits of screening endoscopy. Exclusion criteria were prior gastrointestinal bleeding, portal-systemic shunt surgery, ligation or sclerotherapy of varices, organ transplantation, or other significant comorbidities. Standard information about portal hypertension and endoscopy was provided, including mention of the lack of data to guide the most appropriate management approaches in children. The questionnaire was administered by a nurse practitioner following provision of standardized written and pictorial information about portal hypertension, variceal bleeding, and endoscopy. The questionnaire was composed of 19 questions requiring a response on a 5-point Likert scale.
Institutional research ethics board approval was obtained from the authors' participating institutions for both parts of the study. Consent was obtained from all of the participants, and all of the responses were anonymous.
Data were summarized as mean and standard deviation or median and range. Differences in responses were assessed by chi-square tests, as appropriate, using Microsoft Excel (Microsoft, Redmond, WA). Some questions were qualitative, and answers were analyzed via thematic analysis.
Between February and May 2009, 72 physicians were contacted, and 47 (65%) responded to the survey. Four physicians denied consent. Respondents had a mean of 13.1 years (SD 8.6, range 1–30 years) of experience as a pediatric gastroenterologist and followed a mean of 10 patients (SD 11, range 0–50) with probable or definite portal hypertension. Twelve respondents (26.1%) considered pediatric liver disease to be their primary area of interest or expertise.
The majority of respondents reported performing screening procedures for esophageal varices (70%). Respondents selected patients for screening according to the presence of liver disease with sonographic evidence of portosystemic collaterals (51%), liver disease and splenomegaly (47%), cirrhosis (42%), and liver disease and thrombocytopenia (37%). Patients solely with chronic liver disease were generally not selected for screening (5%).
The following vignette was presented in the questionnaire, based on the previously published survey (16): “An 8 year old child is status post portoenterostomy for biliary atresia. He is clinically well with the exception of portal hypertension. In particular, he has normal growth and development and no history of complications of his liver disease. Pertinent laboratory findings include: platelet count 65,000, albumin 33 g/l (3.3 g/dl), total bilirubin 27 μmol/L (1.6 mg/dl), INR 1.1 (prothrombin time 13.2 s).”
Twenty-nine physicians (63%) agreed they would perform screening procedures to look for esophageal varices in this case, including endoscopy (77%) and ultrasound (23%), but not esophagrams (0%) or other methods. If no varices were found, then repeat endoscopy would be performed after an interval of 12 months (53%) or 24 months (33%). Of the 29 respondents who would perform screening endoscopy, 17 (58%) stated that they would provide primary prophylactic treatment if varices were found. For the prophylaxis options, EVL was preferred by the majority (56%), β-blockers by 37%, and endoscopic sclerotherapy by 7%. The responses for the use of screening endoscopy in this clinical case were similar to responses from the previous survey of American pediatric gastroenterologists.
Respondents provided information about therapies included in their approach to management of a hemodynamically significant acute gastrointestinal bleed in a child with portal hypertension (Table 1). Endoscopy would be performed within 24 hours by 76% of respondents.
The vignette with the same 8-year-old child status postportoenterostomy for biliary atresia was further developed as follows to inquire about secondary management options: “He has had a hemodynamically significant esophageal varix bleed, which required a single red blood cell transfusion. He is clinically stable after initial medical management.” The majority of respondents recommended secondary prophylactic therapy with EVL for this situation. Management with transjugular intrahepatic shunt (TIPS), surgical shunt, or liver transplantation was more likely after recurrent versus first variceal hemorrhage (Table 2).
In the event that a patient with portal hypertension lived a far distance from the tertiary care center (such that transport by plane would be required), many respondents would use different practices regarding esophageal varices for screening (55%), primary prophylaxis (63%), and secondary prophylaxis (50%). Physicians reported being more likely to provide earlier or more frequent screening (15.9%), more likely to provide primary or secondary prophylaxis (34.1%), and more likely to use band ligation (27.3%). The use of β-blockers in this population was more variable, however, being reported more likely by some respondents (13.6%) and less likely by others (13.6%). The majority of pediatric gastroenterologists (68%) use a 4-point scale for the classification of variceal appearance, first described by Paquet (19).
Physician attitudes toward the risks associated with endoscopy were evaluated by the physician questionnaire and are reported below in comparison to the responses of patients and parents.
Patient and Family Questionnaire
Of 33 children and families approached for enrollment, 29 (mean age 11 years, range 2–17 years, 15 boys) agreed to participate. The children's diagnoses included biliary atresia in 7 (24%), overlap syndrome in 3 (10%), congenital hepatic fibrosis in 3, portal vein thrombosis in 3, primary sclerosing cholangitis in 2 (7%), and others in 7 (24%). Using a previously established predictive score based on platelet/spleen size ratio (20), all 29 patients were predicted to have esophageal varices. Eighteen patients and 28 family members (father or mother) completed the questionnaire.
When asked about their desire for a screening endoscopy to help determine the risk of variceal bleeding, 63% of parents and 50% of patients wanted an endoscopy, compared with 16% of both groups who did not want the procedure. Patients and parents were also asked whether they would accept endoscopy specifically if a treatment was available to reduce the risk of bleeding of any varices that may be identified. Under these circumstances, 67% of parents and children would opt for endoscopy.
The questionnaire also explored the extent to which the perceived importance of the risks of endoscopy contributed to the decision to perform or agree to endoscopy. The risk of bleeding caused by endoscopy was a factor that contributed to concern about performing endoscopy in 22% of physicians, 67% of patients, and 54% of parents, whereas the risk of perforation was deemed important by 11% of physicians, 44% of patients, and 52% of parents. The risks of anesthesia and the discomfort of the intravenous cannula were not concerning for physicians when deciding to perform endoscopy (3% and 0%, respectively), but were concerns for a minority of patients (22% and 22%, respectively) and parents (11% and 15%, respectively).
The questionnaire also evaluated factors that may contribute to the desire for endoscopy, such as finding that the patient either has a low or a high risk of bleeding, or knowledge that a treatment was available to prevent bleeding. These criteria were found to be of similar importance within each group of respondents (Table 3).
Parents and physicians were then asked to quantify the level of risk that they would be willing to accept to undertake an endoscopy. All of the parents and physicians would accept or perform a screening endoscopy (for diagnostic purposes only) if the risk of bleeding or perforation were <1 in 100,000 cases; 62% of the parents and 92% of physicians would accept a risk of <1 in 10,000; 27% of parents and 68% of physicians would accept a risk of <1 in 1000. If endoscopic treatment for varices was offered, then all of the parents and physicians would accept a risk of bleeding or perforation of <1 in 100,000, 81% of the parents and 100% of physicians would accept a risk of <1 in 10,000, and 54% of parents and 87% of physicians would accept a risk of <1 in 1000.
The results of this study provide an important overview of current practice among a spectrum of pediatric gastroenterologists and hepatologists in the investigation and management of esophageal varices in children. The data enable important comparisons with adult practice guidelines and with a previously published survey of American pediatric gastroenterologists conducted in the late 1990s (16).
The majority of pediatric gastroenterologists in the present survey (70%) reported that they would consider screening for esophageal varices in children with liver disease and evidence for cirrhosis or of portal hypertension (eg, splenomegaly, thrombocytopenia, or portosystemic collaterals on sonography). Endoscopy was the preferred screening modality.
The results of the present survey suggest that some pediatric gastroenterologists and hepatologists apply to children selected components of the guidelines for the management of adults with portal hypertension, such as screening endoscopy and/or primary prophylactic therapy. Our survey did not explore the reasons why practice differs between these specialists and those who choose not to offer screening endoscopy or other interventions recommended for adults. An important reason is likely to be the lack of pediatric studies that enable a clear understanding of the “correct” treatment option for the clinical vignette presented in our survey. It is not clear that the physiopathology, hemodynamic changes, and the appropriate approaches to clinical management of portal hypertension in adults can be directly extrapolated to children. There are preliminary data that suggest that infants with portal hypertension may have different hemodynamic changes compared with those expected from studies of adults, although further research is required (21). Furthermore, the pharmacokinetics and adverse event profile of drugs may differ significantly between adults and children.
Variation in reported approaches by physicians may imply that tests and therapies are being used inappropriately by some physicians and therefore the quality of care received by some patients is suboptimal (22); however, many elements are involved in the complex decision to undertake screening endoscopy. These include the perceived pretest probability of identifying varices, the individual patient's level of risk associated with endoscopy, the availability of endoscopy, and the perceived strength of evidence that screening endoscopy is beneficial and that primary prophylaxis is indicated in the individual patient. Additional preference-based elements important in making this decision include the patient's and parent's approach to the poorly defined, complex balance of risks and benefits.
Actual care provided in practice may therefore vary for appropriate reasons and unwarranted variation reflecting poor-quality care. There is, therefore, an urgent need for an improved evidence base to support the decision-making process for physicians and patients and to enable the generation of clinical practice guidelines for the management of children with portal hypertension.
Certain responses to this physician survey differed from those obtained in the previous published survey (16). Respondents to this survey were less likely to offer annual follow-up endoscopies if no varices were seen (53% vs 78% in the previous survey), were more likely to withhold primary prophylactic therapy (63% vs 16%), and were much less likely to use β-blockers if primary prophylaxis was offered (37% vs 88%). The reasons for these differences are not revealed by the survey results; possible explanations include national differences in health care systems, the time frame between the surveys, the smaller group of selected specialists surveyed in the prior US survey (30 participants in 30 different centers), or the different opinions of thought leaders used for advice between Canada (the location of the current survey) and the United States (location of the previous survey). To our knowledge, no significant pediatric research studies have been published in the intervening years that include data that are strong enough to change the evidence on which treatment decisions are based.
Although contemporary adult guidelines discourage the use of sclerotherapy for secondary prophylaxis, the results of this survey show that pediatric gastroenterologists/hepatologists continue to be prepared to use this treatment modality. It is likely that this is used primarily for infants or small children, in whom band ligation is typically not feasible because of the small caliber of the esophagus.
For management of a recurrent variceal hemorrhage in a child with biliary atresia, surgical shunting was not favored by either surveyed group (5% current survey, 15% previous survey). This is probably due to several reasons, including the lack of adequate research studies and the limited experience in many centers with use of these treatment modalities in pediatric patients. TIPS or transplantation were preferred options, perhaps because of the high risk of encephalopathy after surgical shunting or the complications arising for subsequent liver transplantation caused by adhesions and the need to reverse a surgical shunt.
An important difference was noted between the perception of risks of endoscopy among patients and families compared with pediatric gastroenterologists/hepatologists. One of the most interesting findings of this survey is the high desire of patients and their parents to perform screening endoscopy even if no treatment option is offered. Patients and families were more concerned about the risks of endoscopy, such as bleeding or perforation, than were their doctors. Moreover, the majority of physicians would accept a complication rate of 1 in 1000 for screening (68%) or treatment (87%) via endoscopy; however, the majority of parents would only accept 1 in 10,000 risk for screening (62%) or treatment (81%) via endoscopy, when asked in the context of this survey. This discrepancy is important because the actual risk associated with endoscopy is likely to be between 1 in 100 and 1 in 1000, if extrapolation of adult data is appropriate (23,24). Thus, physicians and parents may not be congruous in their expectations for an “acceptable risk” and this may have implications for the process of obtaining informed consent and constructing an appropriate treatment plan.
The limitations of this study include the response rate below 100%, which introduces the possibility of bias and results that are therefore not truly representative of practice by all of the pediatric gastroenterologists in Canada. The total number of respondents was inadequate to perform any meaningful subgroup analysis (eg subspecialist hepatologists vs gastroenterologists, more experienced vs less experienced). In the interest of maintaining a questionnaire of manageable length, we were not able to include additional questions to determine the reasons behind responses.
Of importance, this study represents one of the few attempts to describe the approaches taken by pediatric gastroenterologists/hepatologists to the management of esophageal varices in children. It is also the only study to our knowledge that describes patient and family attitudes to screening endoscopy and identifies that many families appear to value the information obtained from screening endoscopy that helps to quantify their child's risk of bleeding, independent of any plan to provide prophylactic treatment. Physicians should therefore consider this important point when considering whether to offer screening endoscopy and not base the decision solely on their intent to provide prophylaxis if varices are found. Our results highlight the variation in care provided to children with similar clinical presentations and thus the urgent need for further research to provide an evidence base to support quality improvement.
1. Lykavieris P, Gauthier F, Hadchouel P, et al
. Risk of gastrointestinal bleeding during adolescence and early adulthood in children with portal vein obstruction. J Pediatr 2000; 136:805–808.
2. Goncalves ME, Cardoso SR, Maksoud JG. Prophylactic sclerotherapy in children with esophageal varices: long-term results of a controlled prospective randomized trial. J Pediatr Surg 2000; 35:401–405.
3. Miga D, Sokol RJ, Mackenzie T, et al
. Survival after first esophageal variceal hemorrhage in patients with biliary atresia. J Pediatr 2001; 139:291–296.
4. van Heurn LW, Saing H, Tam PK. Portoenterostomy for biliary atresia: long-term survival and prognosis after esophageal variceal bleeding. J Pediatr Surg 2004; 39:6–9.
5. Kobayashi A, Itabashi F, Ohbe Y. Long-term prognosis in biliary atresia after hepatic portoenterostomy: analysis of 35 patients who survived beyond 5 years of age. J Pediatr 1984; 105:243–246.
6. Toyosaka A, Okamoto E, Okasora T, et al
. Outcome of 21 patients with biliary atresia living more than 10 years. J Pediatr Surg 1993; 28:1498–1501.
7. Mitra SK, Kumar V, Datta DV, et al
. Extrahepatic portal hypertension: a review of 70 cases. J Pediatr Surg 1978; 13:51–57.
8. Webb LJ, Sherlock S. The aetiology, presentation and natural history of extra-hepatic portal venous obstruction. Q J Med 1979; 48:627–639.
9. Kobayashi A, Itabashi F, Ohbe Y. Long-term prognosis in biliary atresia after hepatic portoenterostomy: analysis of 35 patients who survived beyond 5 years of age. J Pediatr 1984; 105:243–246.
10. D'Amico G, Pagliaro L, Bosch J. Pharmacological treatment of portal hypertension: an evidence-based approach. Semin Liver Dis 1999; 19:475–505.
11. Imperiale TF, Chalasani N. A meta-analysis of endoscopic variceal ligation for primary prophylaxis of esophageal variceal bleeding. Hepatology 2001; 33:802–807.
12. Garcia-Tsao G, Sanyal AJ, Grace ND, et al
. Prevention and management of gastroesophageal varices and variceal hemorrhage in cirrhosis. Hepatology 2007; 46:922–938.
13. Garcia-Tsao G, Bosch J, Groszmann RJ. Portal hypertension and variceal bleeding—unresolved issues. Summary of an American Association for the Study of Liver Diseases and European Association for the Study of the Liver single-topic conference. Hepatology 2008; 47:1764–1772.
14. Ling SC. Should children with esophageal varices receive beta-blockers for the primary prevention of variceal hemorrhage? Can J Gastroenterol 2005; 19:661–666.
15. Molleston JP. Variceal bleeding in children. J Pediatr Gastroenterol Nutr 2003; 37:538–545.
16. Shneider BL. Approaches to the management of pediatric portal hypertension: results of an informal survey. In: Groszmann RJ, Bosch J, editors. Portal Hypertension in the 21st Century. Dordrecht, The Netherlands: Kluwer Academic; 2004. pp. 167–172.
17. Chassin MR, Galvin RW. The urgent need to improve health care quality. Institute of Medicine National Roundtable on Health Care Quality. JAMA 1998; 280:1000–1005.
18. Colletti RB, Baldassano RN, Milov DE, et al
. Variation in care in pediatric Crohn disease. J Pediatr Gastroenterol Nutr 2009; 49:297–303.
19. Paquet KJ. Prophylactic endoscopic sclerosing treatment of the esophageal wall in varices—a prospective controlled randomized trial. Endoscopy 1982; 14:4–5.
20. Gana JC, Turner D, Roberts EA, et al
. Derivation of a clinical prediction rule for the noninvasive diagnosis of varices in children. J Pediatr Gastroenterol Nutr 2010; 50:188–193.
21. Arnon R, Srivastava S, Fahey M, et al
. Hyperdynamic circulation in infants with biliary atresia and portal hypertension [Abstract]. Hepatology 2007; 46:720A.
22. Goodman DC. Unwarranted variation in pediatric medical care. Pediatr Clin North Am 2009; 56:745–755.
23. Geraci G, Pisello F, Modica G, et al
. Complications of elective esophago-gastro-duodenoscopy (EGDS). Personal experience and literature review. G Chir 2009; 30:502–506.
24. Merchea A, Cullinane DC, Sawyer MD, et al
. Esophagogastroduodenoscopy-associated gastrointestinal perforations: a single-center experience. Surgery 2010; 148:876–880.