Pharmacological treatment for gastro‐oesophageal reflux in children
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
To assess the effects of pharmacological treatment for GER in children.
Background
Description of the condition
Gastro‐oesophageal reflux (GER) is a common disorder, characterised by the regurgitation of gastric contents into the oesophagus. Gastro‐oesophageal reflux disease (GERD) is defined as 'GER associated with troublesome symptoms or complications' (Sherman 2009), although the authors caution that this definition is complicated by unreliable reporting of symptoms in children under eight years of age. Gastrointestinal sequelae include oesophagitis, haematemesis, oesophageal stricture formation, and Barrett's oesophagitis. Extra‐intestinal sequelae can include acute life‐threatening events and apnoea, chronic otitis media, sinusitis, secondary anaemia, and chronic respiratory disease (chronic wheezing/coughing or aspiration), as well as failure to thrive.
GER is a very common presentation; both in a primary care or secondary care setting. GER can affect approximately 50% of infants less than three months old (Nelson 1997). The natural history of GER is generally that it improves with age;less than 5% of children with vomiting or regurgitation in infancy continue to have symptoms after the age of 14 months (Martin 2002). This is due to a combination of growth in length of the oesophagus, a more upright posture, increased tone of the lower oesophageal sphincter, and a more solid diet. A more recent study of 210 children with GER in infancy diagnosed by Rome II criteria and followed up for 24 months, showed that 88% were symptom‐free by 12 months (Campanozzi 2009). However, presence of severe oesophagitis has historically been shown to predict the need for surgical reconstruction (Hyams 1988).
Children with certain predisposing conditions are more prone to severe GERD. These include those with neurological impairment (e.g. cerebral palsy), those with repaired oesophageal atresia or congenital diaphragmatic hernia, and those with chronic lung disease.
Diagnosis of GER is usually made based on the symptoms alone, avoiding the need for expensive and possibly harmful investigations. Where the diagnosis cannot be made on clinical grounds, further investigations include 24 hour oesophageal pH monitoring, upper gastrointestinal endoscopy, oesophageal manometry, scintigraphy or sonography. All have been shown to correlate poorly with symptomatology and do not accurately predict the degree of improvement with treatment (Augood 2003).
Description of the intervention
The main aims of treatment of GER are to alleviate symptoms, promote normal growth and prevent complications.
Pharmacological treatments include:
Altering the viscosity of feeds
Alginates (e.g. Gaviscon Infant)
Gaviscon Infant contains sodium and magnesium alginate, and mannitol; it acts as a feed thickener and prevents reflux by increasing the viscosity of feeds. This is differentiated from other Gaviscon preparations, which can also contain sodium bicarbonate or potassium bicarbonate, as well as alginate that, in the presence of gastric acid, form a gel in which carbon dioxide (derived from the breakdown of bicarbonate) is trapped. This 'foam raft' floats on top of the gastric contents and is designed to neutralise gastric acid (providing symptomatic relief), thickens the feed (to reduce reflux) and reduces oesophageal irritation (Mandell 2000).
Caution should be applied when using alginates that contain aluminium due to risk of accumulation in the body , and in children with vomiting or diarrhoea, or children at risk of intestinal obstruction (Gaviscon Product Information 2008). In children whose feeds are already thickened (e.g. Infamil AR; SMA Staydown), coexistent Gaviscon Infant could potentially cause intestinal obstruction (Keady 2007). Gaviscon Infant contains 0.92 mmol Na+ per dose, which should be considered if a child’s sodium intake needs to be monitored with caution (e.g. renal impairment, congestive cardiac failure, preterm infants, or children with diarrhoea and vomiting) (BNF for Children 2008).
Altering the gastric pH
These medications improve symptoms not by reducing reflux but by reducing the acidity of refluxate, in theory reducing oesophageal irritation and providing symptomatic relief.
Antacids
Magnesium hydroxide and aluminium hydroxide (MHAH) is an antacid, commercially available as Maalox, which works by reducing gastric pH. However, as with alginates, long‐term use of aluminium should be avoided in infants and children with chronic renal failure, due to the risk of accumulation.
H2 antagonists
There are several studies that have suggested that H2 antagonists are efficacious in children. Ranitidine is well tolerated and has a low incidence of side‐effects (common side‐effects include fatigue, dizziness or diarrhoea) (Cucchiara 1993). Ranitidine is the most commonly‐used H2 antagonist to reduce acidity of gastro‐oesophageal refluxate. Cimetidine is rarely used clinically as concerns exist about its effects on the cytochrome P450 system and, thus, the potential to causemultiple drug interactions, as well as interfering with Vitamin D metabolism and endocrine function. Famotidine is a recently developed H2 antagonist not commonly used in children.
Proton Pump Inhibitors (PPIs)
PPIs such as omeprazole and lansoprazole belong to a group of drugs that irreversibly inactivate the parietal cell membrane transporter H+/K+ ATPase. This increases the pH of gastric contents and decreases total volume of gastric secretion, thus facilitating emptying.
There are five PPIs approved by the US Food and Drug Administration (FDA) in adults: Omeprazole (since 1988), lansoprazole, pantoprazole, rabeprazole and esomeprazole (the pure S‐isomer of omeprazole). Omeprazole is licensed in children over one year of age in the UK, and lansoprazole is only recommended by the British National Formulary for children when treatment with the available formulations of omeprazole is unsuitable (BNF for Children 2008).
All are metabolised by the cytochrome P450 system within 60 minutes in adults and all are relatively safe, with few reported side‐effects. PPIs are also safe in children with renal impairment but hepatic metabolism of PPIs may be impaired.
The efficacy of PPIs may also be affected by immature parietal cells that are less responsive, and a hypochlorhydria (low production of gastric acid) in the first 20 months. The gastric pH, however, does provide some protection as there is evidence that potentiating the hypochlorhydria in neonates further with omeprazole can result in bacterial overgrowth (Nelis 1985). Consequent increases in respiratory infections have been identified in critically ill patients but in infants and children who are otherwise well, no clear ill‐effects have been demonstrated from this overgrowth .
Altering the motility of the gut (prokinetics)
The use of prokinetics is considered when GER fails to improve with conservative measures. There are several classes of drugs designed to increase gastrointestinal motility.
Metoclopramide
Metoclopramide has alpha‐sympathomimetic activity and blocks dopamine and serotonin receptors. Several adverse effects have been associated with metoclopramide: this is seen in 11 to 34% of children receiving this drug. Adverse effects can include drowsiness or restlessness, or the rarer extrapyramidal reaction (neck pain, rigidity, trismus and oculogyric crisis), which may be more likely with higher doses (Cucchiara 2000). Metoclopramide has recently been the subject of an FDA 'black box' warning (FDA 2009).
Domperidone
Domperidone is a dopamine‐receptor (D‐2) blocker that has relatively fewer side‐effects but case reports of extrapyramidal side‐effects exist (Franckx 1984; Shafrir 1985). Domperidone acts to increase motility and gastric emptying and decreases post‐prandial reflux time. It is commonly used in clinical practice, either as part of empirical medical therapy of GERD or if delayed gastric emptying has been demonstrated on a barium swallow or milk scan.
Cisapride
Cisapride is a gastro‐oesophageal prokinetic agent which stimulates motility in the lower oesophagus, stomach and small intestine by increasing acetylcholine release in the myenteric plexus, controlling smooth muscle. At its peak, cisapride had been prescribed to over 36 million children worldwide (Vandenplas 1999) and was recommended by the European Society for Paediatric Gastroenterology and Nutrition. However, concerns about the effect of cisapride in prolonging the QT interval led to its removal from general paediatric use (Committee on Safety of Medicines 2000). A Cochrane review found that there was no clear evidence that cisapride reduces symptoms of GER (Augood 2003). There was, however, evidence of substantial publication bias favouring studies showing a positive effect of cisapride. The only study to compare cisapride with another treatment (Gaviscon, with or without Carobel) failed to show superior efficacy (Greally 1992). Given the known risks of toxicity and its suspension of manufacture, further trials of cisapride are unlikely.
As cisapride has been the subject of a separate Cochrane review and is now no longer manufactured, we do not propose to review the literature regarding cisapride.
Conservative options
These include reassurance of parents and positioning of the baby to reduce GER through the effect of gravity on gastric contents. This can include elevating the head of the cot or basket that the baby is placed in to sleep, or keeping the baby sitting upright after a feed.
Altering the feed's consistency can be achieved with feed thickeners (e.g. Carobel) and may reduce the reflux of gastric contents because of their increased viscosity. Some feeds are manufactured with a thickening agent added, e.g. SMA Staydown; Infamil AR. Weaning also has a similar effect by increasing the viscosity of gastric contents and GER is known to improve with weaning.
Changes in milk can also improve GER. Some evidence exists that using a partially hydrolysed formula (e.g. Peptijunior) or completely hydrolysed formula (e.g. Neocate) may ameliorate GERresulting from food protein intolerance. Hill and Hoskings found that in infants with GER refractory to medications (prokinetics and H2 antagonists), symptoms of irritability improved on an elemental amino‐acid based formula; and worsened when re‐challenged with low‐allergen soy formula or a partially hydrolysed cow's milk‐based formula (Hill 1996).
Surgical options
These exist to limit GERD. The commonest strategy is a Nissen's fundoplication involving a 360º wrap (Hassall 2005). The aim of this procedure is to augment the lower oesophageal sphincter by a number of factors, such as creating a high pressure zone at the distal oesophagus. Further details are given elsewhere in the literature and are beyond the scope of this review. However, if underlying dysmotility exists, this will persist even after surgery and retching will continue as a prominent feature.
Conservative and surgical strategies will not be addressed by this Cochrane review. Surgery relates to only a small minority of children with GER and would divert from the main focus of this review which is to assess pharmacological treatments.
Why it is important to do this review
Gastro‐oesophageal reflux in children is a common condition often presenting to general practitioners (GPs) and paediatricians. No systematic review has yet assessed the medical evidence for pharmacological treatments which are commonly prescribed. This systematic review aims to critically appraise the existing paediatric literature using the Levels of Evidence adopted by the Oxford Centre for Evidence‐based Medicine.
Objectives
To assess the effects of pharmacological treatment for GER in children.
Methods
Criteria for considering studies for this review
Types of studies
All randomised controlled trials assessing pharmacological treatments for GER will be considered and evaluated. We will exclude studies in which participants have coexisting conditions such as TracheoEsophageal Fistula (TEF), cerebral palsy, or asthma, that predispose to GERD in order to avoid heterogeneity between participants. Any exclusion of randomised studies will be justified in the analysis, with each individual exclusion outlined.
Types of participants
All children (0 to 16 years) with 'GER associated with troublesome symptoms or complications'. Consideration will be given to patient selection and the potential for selection bias. This would include assessing the strategy of recruitment, discussion of the process of randomisation (this should be independent of and remote to the investigators) and blinding (up to and after the point of treatment allocation).
Types of interventions
All currently available medical treatments for GER in children will be included in the review (e.g. alginates, antacids, H2 antagonists, PPIs, prokinetics).
We will consider all randomised controlled trials comparing the medication in question versus placebo or comparisons against other medications. There will be no restrictions on dose, frequency, or duration.
We will look to produce comparisons of all active treatments versus placebo with respect to their treatment class: i.e. Gaviscon versus placebo, Proton Pump Inhibitors (omeprazole, lansoprazole, pantoprazole, rabeprazole and esomeprazole) versus placebo, H2 antagonists (ranitidine, famotidine, cimetidine) versus placebo, prokinetics (metoclopramide, domperidone, erythromycin, bethanechol) versus placebo, and sucralfate versus placebo.
Types of outcome measures
We will include all reported outcomes that are likely to be meaningful to clinicians (such as GPs and paediatricians) making a medical decision about treating GER. Specifically, we will consider clinical symptoms and investigative tools.
Clinical symptoms include:
-
Number of vomiting episodes, back arching, regurgitation, failure to thrive, feeding difficulties, or abdominal pain in infants.
-
Heartburn, epigastric pain or regurgitation symptoms in older children.
Regurgitation is defined according to the Montreal criteria as occurring when relaxation of the lower esophageal sphincter (LES) allows the retrograde movement of gastric contents into the oesophagus and beyond, and can include the ejection of refluxate from the mouth. Regurgitation is distinguished from vomiting physiologically by the absence of: 1) a central nervous system emetic reflex, 2) retrograde upper intestinal contractions, 3) nausea, and 4) retching. Regurgitation is generally characterized as effortless and non‐projectile, although it may be forceful in infants (Sherman 2009).
Investigative tools include:
-
24‐hour pH probe and/or impedance studies:
-
reflux index on pH probe = percentage of time that oesophageal pH < 4;
-
number of reflux episodes.
-
-
Macroscopic appearance of oesophagus on endoscopy.
Consensus exists that there are insufficient data to recommend histology as a tool to diagnose or exclude GERD in children.However, histology is useful to rule out other conditions such as eosinophilic oesophagitis, Barrett's oesophagus, Crohn's disease, infection, and graft‐versus‐host disease (Sherman 2009).
Primary outcomes
Primary outcomes will be concerning improvement in clinical symptoms. These are usually assessed through questionnaires of parents and carers and include number of vomiting episodes (continuous data), episodes of back arching (continuous data), number of episodes of regurgitation (continuous data), failure to thrive (binary outcome), feeding difficulties (binary outcome), or abdominal pain in infants (continuous data). In older children the number of episodes of heartburn, epigastric pain or regurgitation (continuous data) are again assessed through questionnaires completed by patients, parents and health professionals. These include, for example, the GER‐Q questionnaire completed daily by parents and health professionals to provide quantitative data through validated symptom scores. Also included are any serious reported side‐effects of individual medical treatments. (These are currently classified as either a Serious Suspected Adverse Reaction (SSAR), or a Suspected Unexpected Serious Adverse reactions (SUSARs), as defined by the Medicines Health Regulation Authority: "All adverse events judged either by the investigator or sponsor as having a reasonable suspected causal relationship to an Investigational Medicinal Product".)
Secondary outcomes
Secondary outcomes will be improvement in the reflux index (continuous data) or number of reflux episodes on 24‐hour pH probe (continuous data), results of impedance studies (continuous data), or improvement of oesophagitis on endoscopy (visual appearance ‐ binary outcome). Different grading scales currently exist for classifying macroscopic appearances of the oesophagus, currently no single grading scale has been demonstrated to show superior validity to existing alternatives. The number of children within a study population who failed to improve and required fundoplication will also be a secondary outcome (binary outcome).
Search methods for identification of studies
Electronic searches
We shall search for relevant published trials in:
-
The Cochrane Upper Gastrointestinal and Pancreatic Disease Group's Specialised Register and the Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library (last update) Appendix 1;
-
MEDLINE via OVIDSP (January 1950 to May 2010) Appendix 2;
-
EMBASE via OVIDSP (January 1974 to March 2010 ) Appendix 3;
-
Science Citation Index via ISI web of science.
We shall search for ongoing trials in: The meta Register of Controlled Trials (www.controlled‐trials.com), which includes the UK NHS National Research Register.
Search strategy to locate RCTs
Search terms 1 to 29, as given in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2008).
We will interrogate PubMed, MEDLINE and EMBASE for all articles with combinations of the key words "(gastro‐oesophageal or gastroesophageal or gastro‐esophageal or reflux or oesophagitis NOT eosinophilic oesophagitis), and (child$ or infant) and (drug$ or therapy or treatment)".
We will develop this search strategy with the Review Group's Trials Search Co‐ordinator.
This is a draft search strategy and will be adapted to include additional search terms where necessary and will be modified for the other databases listed.
Searching other resources
Reference lists from trials selected by electronic searching will be scanned to identify further relevant trials. Published abstracts from conference proceedings from the United European Gastroenterology Week (published in Gut) and Digestive Disease Week (published in Gastroenterology) will be handsearched. We will handsearch reviews discovered in this search (published in the past five years) for any relevant paediatric studies of medical treatment of GER.
Adverse outcomes
We shall not do a separate search for adverse events.
Language
We will not restrict our search by language and will translate papers as necessary.
Data collection and analysis
We will use Review Manager 5 for data analysis assuming a fixed‐effect model. For continuous measurements, summarised using means and standard deviations, we plan to use weighted mean differences to pool results from studies where a common measurement scale has been used. Where different measurement scales have been employed standardised mean differences will be pooled. For binary outcomes we will compute and summarise rate ratios. We will present 95% confidence intervals for individual studies and summary effects.
Where statistical analyses are not possible (or inappropriate) a descriptive summary will be provided. We will look at all studies, performing a subgroup analysis of those employing an intention‐to‐treat (ITT) analysis where such information exists.
Selection of studies
Two authors (MT and RMB) will check the titles and abstracts identified from the searches. If it is clear that the study does not refer to a randomised controlled trial of a pharmacological treatment of GER in children or infants, it will be excluded. All the authors will assess the full text version of each remaining study to determine whether it meets the pre‐defined selection criteria. Any differences of opinion will be resolved through discussion within the review team. We shall list those studies excluded in the 'Characteristics of excluded studies' table.
Data extraction and management
Two authors (MT and RMB) will independently extract the data using a specially designed data extraction form, then check and enter the data into RevMan. A third team member (NAA) will resolve any differences in opinion.
Assessment of risk of bias in included studies
We shall describe each study in a ‘Risk of bias’ table and address the following issues which may be associated with biased estimates of treatment effect: sequence generation, allocation sequence concealment, blinding, incomplete outcome data, selective outcome reporting and other potential sources of bias (Higgins 2008). We shall comment specifically on:
-
the method of generation of the randomisation sequence;
-
the method of allocation concealment ‐ it will be considered 'adequate' if the assignment could not be foreseen;
-
who was blinded and not blinded (participants, clinicians, outcome assessors) if this is appropriate;
-
how many participants were lost to follow up in each arm, and whether reasons for losses were adequately reported;
-
whether all participants were analysed in the groups to which they were originally randomised (intention to treat principle).
In addition we may report on:
-
the baseline assessment of the participants for age, sex, and duration of symptoms;
-
whether outcome measures were described and their assessment was standardised;
-
the use and appropriateness of statistical analyses, where tabulated data cannot be extracted from the original publication.
We shall record information on all these components in a ‘Risk of bias’ table. We shall also summarise the general quality of all the studies in the section on 'Risk of bias in the included studies'. If there are sufficient trials we will use a funnel plot to investigate reporting (publication) bias. We will perform a sensitivity analysis by excluding those studies where we have identified a high risk of bias.
Measures of treatment effect
For studies of a single pharmacological agent (e.g. omeprazole) versus placebo, if sufficient trials are available and their populations are clinically similar, we will carry out meta‐analyses of primary and secondary end‐points.
We may not expect to see time‐to‐event outcome data, but if these measures are provided in a trial report, they will be analysed using the log‐rank test.
For meta‐analyses of dichotomous outcomes (e.g. healing/no healing of oesophagitis on endoscopy), relative risks (RR) or odds ratios (OR) will be calculated with 95% confidence intervals (CI) and combined for meta‐analysis with RevMan software (RevMan). Data will be combined for the same duration of follow‐up rounded to the nearest month.
Continuous data (e.g. symptom scores) will be combined for meta‐analysis. We will use mean and standard deviations to derive a mean difference (MD) with 95% CI using a fixed‐effect model.
Unit of analysis issues
We will seek input from the Cochrane Upper Gastrointestinal and Pancreatic Diseases Group editorial base for analysis issues involving any included trials with multiple treatment groups, and cluster randomised designs. We may assess cross‐over trials and will assess only the first stage of therapy prior to cross‐over.
Dealing with missing data
We shall contact trial authors or sponsors of studies less than 10 years old to provide missing data, or clarification, where there is uncertainty about the specifics of a trial that are pertinent to analysis.
Assessment of heterogeneity
Studies will be screened to assess clinical heterogeneity and planned subgroup analyses considered if appropriate. We will examine the forest plot and formally test for heterogeneity using the Q‐test, reporting on the extent of any heterogeneity using I2 statistic. If severe heterogeneity exists (I2 statistic > 50%) data extraction and entry will be rechecked. We will seek to identify the source of heterogeneity and perform a sensitivity analysis.
Assessment of reporting biases
We shall assess for the presence of reporting bias by the use of a funnel plot when adequate data are available for individual pharmacological agents (Higgins 2008). If our analysis contains sufficient trials to make visual inspection of the plot meaningful (there is no standard for this and we will seek statistical advice), and if the presence of asymmetry in the inverted funnel suggests a systematic difference between large and small trials in their estimates of treatment effect, we may discuss this further in the Discussion section.
Data synthesis
All individual agents will be assessed separately. We will not combine data, for example, on different proton pump inhibitors.
Subgroup analysis and investigation of heterogeneity
Subgroup analysis will be undertaken in two groups: infants less than one year of age and children between 1 year and 16 years old. These subgroups have different GER characteristics, e.g. infants with symptomatic GER have different symptoms to older children (who are generally on a more solid diet and are upright). Some sections of the review will assess treatments such as milk thickeners (i.e. alginates) which would mainly be used in the infant population. This would mean that studies using the same end‐points in the two subgroups can be compared, making a meta‐analysis more meaningful as the studies available for analysis would be less heterogeneous.
If substantial heterogeneity (I2 > 50%) exists between studies for the primary outcome, we shall explore the reasons for heterogeneity; such as severity of reflux, demographic differences (age, gender, ethnicity) and comorbidity. If it is inappropriate to pool the data because of clinical or statistical heterogeneity a systematic review without meta‐analysis will be performed or a meta‐analysis excluding outlying studies will be performed. We will liaise closely with the Review Group Co‐ordinators at the Cochrane Upper Gastrointestinal and Pancreatic Diseases Group.
Sensitivity analysis
This is mentioned above with respect to potential bias and heterogeneity.
