The of Evidence The Hierarchy of evidence is based on summaries from the National Health and Council (2009), the Oxford Centre for Evidence-based Medicine Levels of Evidence (2011) and Melynyk and Fineout-Overholt (2011).

Ι Evidence obtained from a of all relevant randomised control trials. ΙΙ Evidence obtained from at least one well designed randomised control trial. ΙΙΙ Evidence obtained from well-designed controlled trials without randomisation. IV Evidence obtained from well designed cohort studies, case control studies, interrupted with a control group, historically controlled studies, interrupted time series without a control group or with case- series V Evidence obtained from systematic reviews of descriptive and qualitative studies VI Evidence obtained from single descriptive and qualitative studies VII Expert opinion from clinicians, authorities and/or reports of expert committees or based on physiology

Melynyk, B. & Fineout-Overholt, E. (2011). Evidence-based practice in nursing & healthcare: A guide to best practice (2nd ed.). Philadelphia: Wolters Kluwer, Lippincott Williams & Wilkins. National Health and Medical Research Council (2009). NHMRC levels of evidence and grades for recommendations for developers of guidelines (2009). Australian Government: NHMRC. http://www.nhmrc.gov.au/_files_nhmrc/file/guidelines/evidence_statement_form.pdf OCEBM Levels of Evidence Working Group Oxford (2011).The Oxford 2011 Levels of Evidence. Oxford Centre for Evidence-Based Medicine. http://www.cebm.net/index.aspx?o=1025

Evidence Table Rationale for development of Clinical Guideline: Oesophageal variceal bleeds are a rare but serious complication of portal hypertension and a medical emergency. This new guideline is intended to assist nurses and other health professionals to provide evidence-based care to the infant or child in the acute phase of an oesophageal variceal bleed event, and those at high risk of experiencing such an event. Search Strategy: Research articles included in this review were sourced from CINAHL Plus and Medline databases (2005 - 2013). Searches were limited to English language and full text. MESH topics used included “variceal bleed”, “child*”, “p*ediatric”, “h*emorrhage”, “treatment” and “controlled trial”. Subsequently, articles cited in the identified literature were reviewed and included as appropriate. Where adequate recent evidence could not be identified, or where evidence was deemed to be foundational or paediatric-focused, earlier studies have been included.

Reference Evidence Methods, key findings, outcomes or recommendations level (I-VII) Avgerinos, A., Armonis, A., Stefanidis, G., II This study investigated the early effects of endoscopic injection sclerotherapy (EIS) compared to endoscopic Mathou, N., Vlachogiannakos, J., band ligation (EBL) on hepatic venous pressure gradient (HVPG) during acute variceal bleeding episodes. Kougioumtzian, A., Triantos, C., The study took place during a 30-month period (1998 –2001) at Evangelismos Hospital, Athens, Greece. Papaxoinis, C., Manolakopoulos, S., Participants were 50 cirrhotic patients with an endoscopic diagnosis of variceal bleeding. After Panani, A., & Raptis, S.A. (2004).

Page | 1 Reference Evidence Methods, key findings, outcomes or recommendations level (I-VII) Sustained rise of portal pressure after , all patients underwent femoral hepatic vein catheterization under local anaesthesia for the sclerotherapy, but not band ligation, in measurement of HVPG. Half the study group was treated with EIS and half with EBL. Conventional acute variceal bleeding in cirrhosis. supportive treatment such as intravenous fluids, crystalloid infusions, and fresh, frozen plasma were Hepatology, 39(6): 1623-1630. administered when required. All patients received oral antibiotics for selective intestinal decontamination and lactulose for prophylaxis against hepatic encephalopathy. In conclusion, during acute variceal bleeding EIS, but not EBL, causes a sustained increase in HVPG, which is followed by a higher rebleeding rate. The study justifies the early administration of a vasoactive drug in patients with variceal bleeds for at least 48 hours after EBL or 120 hours after EIS. The results also suggest that changes in HVPG may help predict the outcome of variceal bleed events, however this observation should only be considered exploratory. Bendtsen, F., Krag, A., & Møller, S. (2008). VII This Mini-Symposium report was prepared by staff at Copenhagen University and Hvidovre Hospital in Treatment of acute variceal bleeding (Mini- Copenhagen, Denmark. It makes numerous evidence-based recommendations related to the treatment of Symposium). Digestive and Liver Disease, acute variceal bleeding in adults. 40: 328–336. Cárdenas, A., Ginès, P., Uriz, J., Bessa, X., IV This retrospective review assessed the , clinical course, predictive factors and short-term prognosis Salmerón, J. M., Mas, A., Ortega, R., of renal failure in patients with cirrhosis and gastrointestinal bleeding over a 2-year period in 178 patients Calahorra, B., De Las Heras, D., Bosch, J., admitted to a specialised gastrointestinal unit of a tertiary referral centre. A control group of 73 patients with Arroyo, V. & Rodés, J. (2001). Renal failure upper-gastrointestinal bleeding without cirrhosis was identified from patients admitted during the same period after upper gastrointestinal bleeding in and matched blindly to the cirrhotic patients. The main findings of the current study are that renal failure is a cirrhosis: Incidence, clinical course, moderately frequent event in patients with cirrhosis and gastrointestinal bleeding, particularly in the setting of predictive factors, and short-term a severe bleed and advanced liver disease. prognosis. Hepatology, 34: 671–676. Chavez-Tapia, N., Barrientos-Gutierrez, T., I This systematic review of 17 RCTs compared the all-cause mortality and infection mortality, between cirrhotic Tellez-Avila, F., Soares-Weiser, K., (adult) patients with upper gastrointestinal bleeding receiving antibiotic prophylaxis or no Mendez-Sanchez, N., Gluud, C., & Uribe, intervention/placebo. Some of the RCTs were single-centre, some multi-centre. Most were not blinded. The M. (2011). Meta-analysis: antibiotic trials excluded patients with known infections. Twelve trials evaluated antibiotic prophylaxis compared with prophylaxis for cirrhotic patients with upper placebo or no intervention. gastrointestinal bleeding - an updated The meta-analysis found that use of prophylactic antibiotics in cirrhotic patients with upper gastrointestinal review. Alimentary Pharmacology bleeding significantly reduced bacterial infections, incidence of rebleeding events, and length of and Therapeutics, 34(5): 509-518. hospitalisation. These benefits were observed independent of which antibiotic was administered. The review did however find that all the examined trials had methodological weaknesses with high risk of bias. Lack of blinding and lack of proper sample size calculations were the most common sources of bias. Costaguta, A., & Alvarez, F. (2012). VII This review paper summarised current knowledge and expert opinion regarding the aetiology and Etiology and Management of Hemorrhagic management of haemorrhagic complications of portal hypertension in children. The paper was prepared by Complications of Portal Hypertension in

Page | 2 Reference Evidence Methods, key findings, outcomes or recommendations level (I-VII) Children. International Journal of researchers at Universit́e de Montŕeal, Canada. Hepatology, 1-8.

D'Amico, G., Pagliaro, L., Pietrosi, G., & I This systematic review of RCTs compared benefits and harms of sclerotherapy with vasoactive drugs Tarantino, I. (2010). Emergency (vasopressin, terlipressin, somatostatin or octreotide) for acute variceal bleeding in adult cirrhotic patients sclerotherapy versus vasoactive drugs for (regardless of the aetiology and the severity of cirrhosis).Seventeen RCTs were reviewed, none of which bleeding oesophageal varices in cirrhotic were blinded. No significant differences were found when comparing drugs. Combining all the trials patients. Cochrane Database Of Systematic irrespective of the drug, the risk differences were failure to control bleeding, five-day failure rate, rebleeding, Reviews, Mar 17(3):CD002233. DOI: mortality and transfused blood units. Adverse events and serious adverse events were significantly more 10.1002/14651858.CD002233.pub2. frequent with sclerotherapy. The authors report that the methodological quality of the included trials was generally unsatisfactory, with only five trials reporting adequate control of . However, four of the RCTs in the analysis have not been obtained in full; only abstracts have been read. Thus methodologies and limitations of these four studies cannot have been reasonably scrutinised. The meta-analysis found no convincing evidence to support the use of emergency sclerotherapy for variceal bleeding in cirrhosis as the first, single treatment compared to vasoactive drugs. De Franchis, R. (2005). Evolving VII This paper is a summary of the IV Baveno International Consensus workshop on definitions, methodology consensus in portal hypertension: Report of and therapeutic strategies related to portal hypertension and variceal bleeding, and provides guideline and the Baveno IV consensus workshop on recommendations for the conduct of trials in this field. The report makes key recommendations related to the methodology of diagnosis and therapy in acute management of a variceal bleed and adjunct therapies after a bleeding episode, including fluid portal hypertension. Journal of Hepatology, resuscitation; endoscopy; pharmacological therapy; secondary prophylaxis. 43: 167-176. Gana, J., Turner, D., Roberts, E., & Ling, S. IV The objective of this study was to derive a non-invasive clinical prediction rule able to identify children with (2010). Derivation of a Clinical Prediction oesophageal varices. Fifty-one consecutive children with liver disease or portal hypertension who underwent Rule for the Noninvasive Diagnosis of endoscopy were included in the retrospective study. At endoscopy, variceal size was graded on a 4-point Varices in Children. Journal of Pediatric Likert scale. Results of physical examination, blood tests, and abdominal ultrasound scan were recorded. Gastroenterology and Nutrition, 50(2): 188- Spleen length on scan was expressed as a standard deviation score (z score). A descriptive univariate 193. analysis was performed on variables that were potentially associated with oesophageal varices and multivariate logistic regression was then modelled to derive a clinical prediction rule. The variables clinical prediction rule is a simple non-invasive measure that may identify children at high risk for oesophageal varices. A prospective validation study is yet to be completed by the authors. The study found 33% of the children had oesophageal varices. Variables found to differ significantly between children with and without varices included platelet/spleen-length z score ratio, platelet count, international normalized ratio, aspartate aminotransferase/alanine aminotransferase ratio, and albumin. Garcia-Pagán, J., & Bosch, J. (2005). V This review focuses on the role of endoscopic band ligation in the primary and secondary prevention of Endoscopic band ligation in the treatment of variceal bleeding as well as in the treatment of the acute bleeding episode. The authors searched PubMed portal hypertension. Nature Clinical Practice for articles published between 1990 and 2005 using the term “band ligation” alone and in combination with Gastroenterology & Hepatology, 2(11): 526- the terms “acute variceal bleeding”, “variceal rebleeding”, and “first variceal bleeding”. In addition, the reports

Page | 3 Reference Evidence Methods, key findings, outcomes or recommendations level (I-VII) 535. of the portal hypertension consensus conferences held in Baveno in 1990, 1995 and 2000 (Baveno I, II and III) and in Reston in 1996, as well as new data presented at the 2005 Baveno IV meeting, were also considered. Papers cited in these studies, and reports and abstracts presented at the last five annual meetings of the European Associations for the Study of the Liver and of the American Association for the Study of Liver Diseases were also evaluated. An algorithm for management of acute variceal bleeding and primary and secondary prophylaxis of variceal bleeding is also provided. Garcia-Tsao, G., Sanyal, A.J., Grace, N.D., VII This guideline has been approved by the American Association for the Study of Liver Diseases and the & Carey, W. (2007). American Association American College of Gastroenterology. It provides a data-supported approach to the management of patients for the Study of Liver Diseases Practice with varices and variceal haemorrhage. The recommendations are based on: formal review and analysis of Guidelines for the Prevention and the recently published world literature on the topic and expert opinion. When little or no data exist from well- Management of Gastroesophageal Varices designed prospective trials, emphasis is given to results from large series and reports from recognised and Variceal Hemorrhage in Cirrhosis. experts. Further controlled clinical studies are needed to clarify aspects of this statement, and revision may Hepatology, 46(3): 922-938. be necessary as new data appear. The guideline’s key recommendations relate to the acute management of a variceal bleed and adjunct therapies after a bleeding episode, including initial resuscitation, fluid resuscitation, pharmacological treatment, endoscopy and secondary prophylaxis. Lacroix, J., Hebert, P.C., Hutchison, J.S., II In this prospective, randomised, controlled, single-centre non-inferiority trial, the researchers hypothesized Hume, H.A., Tucci, M., Ducruet, T., Gauvin, that a restrictive PRBC transfusion strategy would be as safe as a liberal transfusion strategy, as judged by F., Collet, J.P., Toledano, B.J., Robillard, the outcome of multiple-organ dysfunction. P., Joffe, A., Biarent, D., Meert, K., & The study included 637 stable, critically ill children who had haemoglobin concentrations below 9.5g/dL Peters, M.J. (2007). Transfusion strategies within seven days after admission to ICU. Blinding was not appropriate for this trial and methodology and for patients in pediatric intensive care units. analysis was robust. The researchers conclude from the trial that in stable, critically ill children a New England Journal of Medicine, 356: haemoglobin threshold of 7g/dL for PRBC transfusion can decrease transfusion requirements without 1609–1619. increasing adverse outcomes. Lam, J.C., Aters, S.S., & Tobias, J.D. IV A retrospective review of all patients 18 years of age or less who had received octreotide at a single-centre. (2001). Initial Experience with Octreotide in The sample included 10 patients aged from 14 days to 17 years with no control or comparison group. The the Pediatric Population. American Journal effectiveness of octreotide was evaluated by criteria based on the indication for its use. In patients with Of Therapeutics, 8: 409-416. gastrointestinal bleeding, transfusion requirements before and after octreotide therapy and the clinical evidence of ongoing bleeding were evaluated. Based on their study and review of other small paediatric and adult studies, the authors conclude that octreotide is a viable medication in children with gastrointestinal bleeding, pancreatitis, hypoglycaemia, and chylous ascites/ peritoneum because of its low risk profile and apparent effectiveness. They recommend further evaluation of octreotide in RCTs to make definitive conclusions and determine optimal dosing. It is notable that only three of the patients considered in this 12 year-old study had oesophageal bleeds and the age of the patients with oesophageal bleeds (14 and 17 years) also means their physiology is more adult than paediatric. Given the weaknesses in the study, no conclusions could reliably be applied to the clinical guideline under development. No higher-quality paediatric trials could be found (other than in abstract-form only) that investigated the use of octreotide.

Page | 4 Reference Evidence Methods, key findings, outcomes or recommendations level (I-VII) Sarin, S., Kumar, A., Angus, P., Baijal, S., VII The Asian Pacific Association for the Study of the Liver (APASL) set up a working party on Portal Baik, S., Bayraktar, Y., & ... Zhang, C. Hypertension in 2002, with a mandate to develop consensus guidelines on various clinical aspects of portal (2011). Diagnosis and management of hypertension, relevant to disease patterns and clinical practice in the Asia–Pacific region. This review acute variceal bleeding: Asian Pacific summarises the APASL consensus guidelines on acute variceal bleeds. In developing these guidelines, the Association for Study of the Liver working party reviewed the existing literature and deliberated to develop consensus guidelines. Only those recommendations. Hepatology statements that were unanimously approved by the experts were accepted. International, 5(2): 607-624. The guidelines makes key recommendations, as they relate to the acute management of a variceal bleed and adjunct therapies after a bleeding episode, including: clinical assessment, airway management, fluid resuscitation, and antibiotics. Shaheen, N.J., Stuart, E., Schmitz, S.M., II This double-blinded, randomised, placebo-controlled trial of assessed the efficacy of the proton pump Mitchell, K.L., Fried, M.W., Zacks, S., inhibitor (PPI), pantoprazole, as an adjunct to elective endoscopic variceal ligation. Participants were patients Russo, M.W., Galanko, J., & Shrestha, R. presenting at a single US hospital between 2000 and 2003 with portal hypertension, varices and a history of (2005). Pantoprazole reduces the size of prior variceal haemorrhage, aged 18 to 80 years. Monitoring during and after treatment was compulsive, and postbanding ulcers after variceal band no subject was lost to follow-up. Compliance was high with the intervention. The RCT concluded that ligation: a randomized, controlled trial. subjects receiving pantoprazole after elective ligation leads to a 50% reduction in the size of post-banding Hepatology, 41(3): 588-594. ulcers at follow-up endoscopy compared to placebo. Because of the infrequency of rebleeds post ligation, the researchers could not use this variable as a primary outcomes of the effect of the PPI on improved mortality cannot be measured with the sample size used in this study. The focus instead is on reducing the size and number of ulcers and the subjects’ quality of life reports (dysphagia, chest pain, and heartburn). Shneider, B.L., Bosch, J., de Franchis, R., VII In 2011, an international panel of experts revised and developed the Baveno V Consensus Report into a Emre, S.H., Groszmann, R.J., Ling, S.C., paediatric-specific commentary. The specialists’ qualifications are not described although it is assumed they Lorenz, J.M., Squires, R.H., Superina, R.A., are gastroenterologists and/or hepatoligists. No critique of the panel report was identified. This article reports Thompson, A.E., & Mazariegos, G.V. the recommendations of the panel. It describes the methodology of diagnosis and therapy in paediatric portal (2012). Portal Hypertension in Children: hypertension and its complications. The recommendations in the report are relevant primarily to pre- Expert Pediatric Opinion on the Report of pubescent children, where physiologic parameters are most distinct from adults. Recommendations for the Baveno V Consensus Workshop on therapeutic management include preventing the formation of varices, preventing a bleed, acute management Methodology of Diagnosis and Therapy in of a variceal bleed preventing a second bleed and management of treatment failures. This report is Portal Hypertension. Pediatric authoritative in its definitions for a bleeding episode, but for therapeutic management of a variceal bleed, it is Transplantation, 16: 426–437. less conclusive. Most of the statements are expert opinion or are derived from or cohorts. Tibballs, J., Kinney, S., Duke, T., Oakley, IV This analysis is a quality assurance exercise conducted by staff at Royal Children’s Hospital, Melbourne to E., & Hennessy, M. (2005). Reduction of determine the impact of the implementation of a paediatric medical emergency team (MET) on cardiac arrest, paediatric in-patient cardiac arrest and mortality, and unplanned admission to intensive care in a paediatric tertiary care hospital. death with a medical emergency team: The study made a comparison of the retrospective incidence of cardiac arrest and death during 41 months preliminary results. Arch Diseases of before introduction of a MET service with the prospective incidence of these events during 12 months after Childhood, 90(11): 1148-1152. its introduction. A comparison of transgression of MET call criteria in patients who arrested and died before and after introduction of MET was also made. Although clinically important reductions of cardiac arrest and

Page | 5 Reference Evidence Methods, key findings, outcomes or recommendations level (I-VII) death were observed after implementation of the MET service, these were not statistically significant. This probably reflects a relatively small incidence of paediatric cardiac arrest, a small MET sample size after introduction of MET, and the fact that paediatric arrest may occur without warning. Tsoi, K., Ma, T., & Sung, J. (2009). V This literature review evaluated the optimal timing of early endoscopy in adult patients with upper GI Endoscopy for upper gastrointestinal bleeding, by examining the findings of 3 RCTs and 5 retrospective cohort studies published between 1996 bleeding: how urgent is it? Nature Reviews and 2007. Outcome measurements included recurrent bleeding, surgery, mortality, length of hospital stay, Gastroenterology & Hepatology, 6(8): 463- and blood transfusion. Studies were categorised into those in which endoscopy was performed within 2–3 469. hours, 6–8 hours, 12 hours or 24 hours of the patient’s presentation to hospital. The authors conclude that appropriate endoscopic therapy helps to identify the source of bleeding, establishes the underlying aetiology, achieves haemostasis; prevents rebleeding, reduces transfusion requirements, shortens the length of hospital stay, and reduces mortality. A prompt initial clinical and endoscopic assessment should enable effective triage of patients who require active therapy versus those who need monitoring and preventive therapy. However, no evidence exists that very early endoscopy reduces the risk of rebleeding or improves survival. Watson, E.F., & Church, N.I. (2013). VII The following recommendations are made in this review paper, which summarises current knowledge and Haematemesis and melaena. Africa Health, expert opinion. It is prepared by a Gastroenterology Registrar and a Consultant Gastroenterologist at the 35(5): 32-38. Royal Infirmary of Edinburgh, UK. The authors outline recommendations for initial clinical assessment and good history taking and discuss benefits of this (i.e. guiding fluid resuscitation and rapid diagnosis) and make recommendations for immediate management.

Page | 6