Abstract
Objectives: To provide Canadian primary care physicians with an evidence-based clinical management tool, including diagnostic and treatment recommendations, for patients who present with uninvestigated dyspepsia.
Recommendations: The management tool has 5 key decision steps addressing the following: (1) evidence that symptoms originate in the upper gastrointestinal tract, (2) presence of alarm features, (3) use of nonsteroidal anti-inflammatory drugs (NSAIDs), (4) dominant reflux symptoms and (5) evidence of Helicobacter pylori infection. All patients over 50 years of age who present with new-onset dyspepsia and patients who present with alarm features should receive prompt investigation, preferably by endoscopy. The management options for patients with uninvestigated dyspepsia who use NSAIDs regularly are: (1) to stop NSAID therapy and assess symptomatic response, (2) to treat with NSAID prophylaxis if NSAID therapy cannot be stopped or (3) to refer for investigation. Gastroesophageal reflux disease can be diagnosed clinically if the patient's dominant symptoms are heartburn or acid regurgitation, or both; these patients should be treated with acid suppressive therapy. The remaining patients should be tested for H. pylori infection, and those with a positive result should be treated with H. pylori-eradication therapy. Those with a negative result should have their symptoms treated with optimal antisecretory therapy or a prokinetic agent.
Validation and evidence: Evidence for resolution of the dyspepsia symptoms was the main outcome measure. Supporting evidence for the 5 steps in the management tool and the recommendations for treatment were graded according to the strength of the evidence and were endorsed by consensus of committee members. If no randomized controlled clinical trials were available, the recommendations were based on the best available evidence.
Literature review: Evidence was obtained from MEDLINE searches for pertinent articles published from 1966 to October 1999. The searches focused on dyspepsia, diagnosis and treatment. Additional articles were retrieved through a manual search of bibliographies and abstracts from international gastroenterology conferences.
Sponsors: Supported by unrestricted educational grants from AstraZeneca Canada Inc., Isotechnika Inc. and Integrated Healthcare Communications Inc.
- used in this article: AC amoxicillin + clarithromycin
- BMT bismuth (subsalicylate) + metronidazole + tetracycline
- COX-2 cyclo-oxygenase 2
- GERD gastroesophageal reflux disease
- H2-RA H2-receptor antagonist
- MC metronidazole + clarithromycin
- NSAID nonsteroidal anti-inflammatory drug
- PPI proton-pump inhibitor
- RBC ranitidine bismuth citrate
- UBT urea breath test
Dyspepsia is a common condition in Canada (prevalence 29%1) that significantly diminishes the quality of life of those affected. Primary care physicians treat most patients with dyspepsia. An estimated 7% of the average Canadian family physician's practice is devoted to the management of dyspepsia, and 23% of these patients are presenting for the first time.2
The term "dyspepsia" describes a heterogeneous group of symptoms with numerous underlying causes. One of the challenges for the primary care physician is to determine optimal treatment for the patient presenting with new-onset or previously uninvestigated dyspeptic symptoms. The clinician must decide whether investigations are needed and must determine optimal drug therapy when treatment is warranted. Selection of the appropriate agent, determination of the duration of treatment and an appropriate follow-up plan are then required.
Recognition of the role of Helicobacter pylori in the pathogenesis of peptic ulcer disease has revolutionized ulcer therapy and has prompted re-evaluation of the clinical approach to dyspepsia. New data have been published concerning treatment of dyspepsia in patients who harbour and those who do not harbour H. pylori. These data have implications for the diagnosis and treatment of uninvestigated dyspepsia in primary care.
Evidence-based medicine combines the best available evidence from the medical literature and clinical expertise to aid decision-making in patient care. The Canadian Dyspepsia (CanDys) Working Group was convened with the mandate to develop an evidence-based management tool for uninvestigated dyspepsia that would be practical and would reflect the realities of the primary care setting. The aim was to provide primary care physicians with recommendations and guidance concerning appropriate investigations, treatments and indications for referral for patients with uninvestigated dyspepsia. The resulting clinical management tool is intended not to regulate practice but, rather, to support clinical decision-making.
Methods
Canadian Dyspepsia Working Group
The 18 members of the working group were selected by the chair (S.v.Z.) because of their expertise in dyspepsia, evidence-based medicine and continuing medical education. Broad representation from across the country was sought. The group is a mixture of university-based and private practice family physicians, gastroenterologists and pharmacists.
All members initially met in Montreal in May 1998 to discuss the aims of the working group, the important elements for inclusion in the management tool and the methods by which the evidence in support of a newly developed management tool would be reviewed. This meeting included several presentations from gastroenterologists and family physicians on important aspects of dyspepsia. It also included case presentations by family physicians, selected from their clinical practices, to ensure that the realities of managing dyspepsia in primary care were addressed adequately. The management tool and associated treatment recommendations, in their final form, are the outcome of several revisions made during 2 additional meetings and through teleconferences.
Review and grading of evidence
Literature review methods for the initial meeting included MEDLINE searches for articles published from 1966 to April 1999 on the following major topics: the definition of dyspepsia, the differential diagnosis and the value of subclassification based on symptoms alone; the prevalence and natural history of dyspepsia; the prevalence of organic disease in patients with uninvestigated dyspepsia; the yield of specific investigations; and evidence in support of various treatments for dyspepsia. Searches relevant to treatment options were conducted for lifestyle modification and for various drug therapies, including antacids, H2-receptor antagonists (H2-RAs), prokinetic agents and proton-pump inhibitors (PPIs). The titles and abstracts from the literature search were reviewed, and all relevant articles were retrieved for further evaluation.
Individual working group members compiled the existing evidence on particular topics and provided overviews and summary conclusions of the topics for the entire committee. High-quality reviews, systematic overviews and meta-analyses were considered as a source of evidence. The original studies on which the conclusions of these reviews were based were retrieved and evaluated by the working group to ensure that committee members agreed with the conclusions.
To ensure incorporation of current information by the end of the document revision process, specific searches for clinical trial data up to January 2000 were conducted for PPIs (lansoprazole, omeprazole and pantoprazole) and H2-RAs (ranitidine, famotidine, cimetidine and nizatidine) as well as the prokinetic cisapride, the cyclo-oxygenase 2 (COX-2)-specific inhibitors celecoxib and rofecoxib, and ranitidine bismuth citrate. Searches included only articles written in English and were limited to data for human subjects. A summary of these searches related to treatment recommendations is given in Table 1 for each mini-management schema. The most relevant articles were retrieved and reviewed.
The bibliographies of key articles, including previously published guidelines, were reviewed manually for relevant references. Additional MEDLINE searches were conducted to address specific issues that arose during development of the project, such as patient's age at endoscopy and gastric cancer rates, the role of NSAIDs in dyspepsia, and the diagnosis and treatment of gastroesophageal reflux disease (GERD). Because some of the most recent data were available only in abstract form, abstracts of the major gastroenterology meetings, such as the American Gastroenterological Association and the World Congress of Gastroenterology, were also reviewed.
The importance of costs and other economic considerations was recognized, and if appropriate economic data were available they were included. However, health economic data are limited in this area, particularly with respect to the Canadian health care system, and this review highlighted the need for more data to help guide management choices.
Relevant studies were graded to provide an indication of the strength of the evidence and the related recommendations. Diagnostic test studies were graded according to the rating used in the Report of the Canadian Hypertension Society Consensus Conference3 (Appendix 1). All other key studies were graded using the levels of evidence and treatment recommendations of the Canadian Task Force on the Periodic Health Examination4 (Appendix 2). Gaps in the existing evidence were identified, and where evidence was lacking, the best available evidence was considered and consensus sought.
Presentation of treatment recommendations in the clinical management tool
For each mini-management schema in the clinical management tool, treatment evidence was sought for 3 PPIs (lansoprazole, omeprazole and pantoprazole), 4 H2-RAs (cimetidine, famotidine, nizatidine and ranitidine), cytoprotective agents (misoprostol only) and prokinetic agents (cisapride only). The treatment recommendations related to these compounds are listed by drug class (PPI for proton-pump inhibitors and H2-RA for H2-receptor antagonists) rather than by specific drug. If trial data are not available for all drugs in each category, footnotes are used to clarify the specific drugs and dosages for which evidence is currently available.
Background on dyspepsia
Terminology
Dyspepsia is a symptom complex rather than a specific disease. Most patients do not present to their physician complaining of "dyspepsia" but, instead, describe symptoms that the clinician interprets as dyspepsia.5 These symptoms are subjective, and the patient's description may be language- and culture-dependent.[6, 7] The variability of the physician's interpretation and of the patient's description has created confusion concerning a unifying definition of dyspepsia.5 The term "dyspepsia" encompasses all relevant symptoms regardless of whether there is a demonstrable cause. However, for the purpose of the management tool, it is important to distinguish the terms dyspepsia, functional dyspepsia (dyspepsia for which investigation has shown no cause) and uninvestigated dyspepsia (dyspepsia for which no cause has yet been sought).
Dyspepsia
The CanDys Working Group agreed on the following working definition of dyspepsia: "Dyspepsia is a symptom complex of epigastric pain or discomfort thought to originate in the upper gastrointestinal tract, and it may include any of the following symptoms: heartburn, acid regurgitation, excessive burping/belching, increased abdominal bloating, nausea, feeling of abnormal or slow digestion, or early satiety." Duration was not specified as part of the definition because patients may present immediately following the onset of symptoms, or they may wait years before consulting their family physician.
Over the years, several definitions of dyspepsia have been proposed.[8–12] In 1991 an international panel of clinical investigators developed a comprehensive classification of functional gastrointestinal disorders, including dyspepsia, commonly referred to as the Rome Criteria.13 The criteria were recently updated at the Rome II consensus conference,14 and the following definition of dyspepsia was recommended: "Dyspepsia refers to pain or discomfort centred in the upper abdomen."
Dyspepsia can be caused by a variety of conditions. In a technical report of the American Gastroenterological Association, 36 original studies were reviewed in which patients with dyspepsia were investigated by means of endoscopy.8 The 3 main organic causes of dyspeptic symptoms were duodenal or gastric ulcer (15% to 25% of cases), reflux esophagitis (5% to 15% of cases) and gastric or esophageal cancer (less than 2% of cases). More than 50% of patients with typical reflux symptoms, with pathological gastroesophageal reflux (e.g., diagnosed by means of 24-hour esophageal pH monitoring), have no macroscopic evidence of esophagitis and are classified as having endoscopy-negative reflux disease. The overall frequency of GERD is 20% to 40%. In Canada GERD is now more common than ulcers. Similarly, up to 60% of patients who present with dyspepsia will have no definite structural or biochemical explanation for their symptoms and are considered to have functional dyspepsia, often referred to in the literature as nonulcer dyspepsia.
Functional dyspepsia
At the Rome II consensus conference14 the following definition for functional dyspepsia was recommended: "Twelve weeks or more (within the last 12 months) of persistent or recurrent dyspepsia and evidence that organic disease likely to explain the symptoms is absent (including at upper endoscopy)." In addition, 2 symptom subgroups for functional dyspepsia were proposed based on the predominant (or most bothersome) single symptom identified by the patient. For ulcer-like dyspepsia, pain centred in the upper abdomen is the most bothersome symptom. For dysmotility-like dyspepsia, an unpleasant or troublesome nonpainful sensation (discomfort) in the upper abdomen is the predominant symptom.
Uninvestigated dyspepsia
The Canadian clinical management tool developed by the CanDys Working Group begins with uninvestigated dyspepsia and is intended for the patient with new-onset or recurrent dyspeptic symptoms in whom no investigations have been conducted and no specific diagnosis for the current symptoms exists. Although the Rome II definition does not include reflux disease within dyspepsia, the CanDys Working Group felt strongly that this does not coincide with the conceptual framework that primary care physicians follow when a patient presents with uninvestigated dyspeptic symptoms. To be practical and reflect the reality of primary care, the consensus was that reflux disease is an integral constituent of uninvestigated dyspepsia. This is a major difference from the Rome II consensus guidelines and the American Gastroenterological Association medical position statement.[14, 15] The difference reflects the needs of the primary care physician rather than those of the specialist and researcher.
Epidemiology and quality of life
Dyspepsia is a common condition. A recent study evaluating the prevalence of dyspepsia in the general population showed that 29% of all Canadians have substantial dyspeptic symptoms.1 The reported prevalence of dyspepsia in Western countries generally ranges from 25% to 50%.[8–12, 16–19] Much of the variability is likely the result of using different definitions, sampling methods or periods of surveillance.[17, 20]
The natural history of dyspepsia is one of persisting or frequently recurring symptoms.[8, 18, 21, 22] Although the overall prevalence of dyspepsia in a given community remains stable, symptoms will resolve in some people and develop in others.[16, 17, 19, 23, 24]
Few dyspepsia treatment trials have incorporated measures of quality of life,25 but 2 recent studies from Canada1 and Britain26 investigated the effect of dyspeptic symptoms on quality of life in the general population. Both studies used the Psychological General Well Being Index as a measure and demonstrated that patients with dyspepsia have a significantly lower quality of life than do healthy subjects in the community. The quality of life of patients with dyspepsia is also lower than that of patients with peptic ulcer disease.27
Clinical management tool for uninvestigated dyspepsia
The clinical management tool maps a series of 5 key decision points with 4 related mini-management schemata (Figs. 1 to 5). The 5 key decision points (boxes A to E in Fig. 1) address the following questions: (A) Are there other possible causes for the symptoms? (B) Is the patient over 50 years of age, or does the patient have any alarm features? (C) Is the patient regularly using NSAIDs (including ASA)? (D) Is the dominant symptom heartburn or acid regurgitation, or both? (E) Is the patient infected with H. pylori?
The key decision points link directly to 4 related mini-management schemata that include treatment recommendations for the following groups: (1) patients using NSAIDs, (2) patients with GERD, (3) patients with positive results of testing for H. pylori and (4) patients with negative results of testing for H. pylori. The treatment recommendations in each mini-management schema are ranked according to strength of evidence, and the schemata provide guidance for management until symptom resolution or referral for investigation.
The remainder of this document summarizes the evidence used to support each of the 5 key decision points related to diagnosis and investigation, followed by evidence for the treatment advice provided in the 4 related mini-management schemata. Each section is followed by a graded recommendation, and the highest level of evidence available for each recommendation is indicated, with relevant references. Therapeutic choices are listed from most preferred to least preferred in each mini-management schema, based on available evidence of best clinical outcome.
Evidence related to the 5 key decision points: diagnosis and investigation
Identification of patients with other possible causes of dyspeptic symptoms or an increased risk of underlying structural abnormalities
Many patients with dyspeptic symptoms consult a physician because they fear serious disease,[28–30] including cancer[28, 31] and heart disease.28 The first priority of the primary care physician is to identify patients who have a higher risk of underlying structural abnormalities or other causes for their dyspeptic symptoms. Careful history-taking is crucial for the identification of these patients, and the first 3 key decision boxes in the clinical management tool outline the related recommendations (Fig. 1). Although the decision boxes are mapped in a linear fashion for clarity of presentation, history-taking is not necessarily conducted in this sequence in clinical practice.
Box A: Other possible causes of symptoms
The clinician must consider the possibility that symptoms suggestive of dyspepsia may not originate from the upper gastrointestinal tract. A thorough history-taking and physical examination should identify patients for whom it is necessary to exclude cardiac, hepatobiliary and other nongastrointestinal origins of the presenting dyspeptic symptoms,9 including possible medication-induced dyspepsia,32 lifestyle or dietary indiscretions. Although it may be difficult to exclude all of these causes on history-taking, it is important to know when to investigate further.
Box B: Older patients and patients with alarm features
Although cancer is a rare cause of dyspeptic symptoms (accounting for less than 2% of cases),8 it is an important consideration. Alarm features and increased age identify patients with a higher risk of an organic cause for their uninvestigated dyspeptic symptoms, including cancer and ulcers.[33, 34]
Retrospective studies indicate that most younger patients with gastric or esophageal cancer present with at least one alarm feature.[34, 35] Unfortunately, if alarm features are present, the cancer is usually advanced, and curative resection is not possible.35 This, however, does not diminish the importance of making the diagnosis.
Data from the United Kingdom suggest that cure rates for gastric cancer may be improved by early detection.[36, 37] Prompt investigation in older patients with dyspeptic symptoms may increase the proportion of early gastric cancers that are detected and that potentially can be cured.37 Four prevalence studies, assessing 5933 patients with dyspepsia, showed that gastric and esophageal cancers are rarely a cause of dyspeptic symptoms in younger patients.[33, 34, 37, 38] In these studies, no cases of gastric or esophageal cancer were reported in any patient under 45 years of age.
The age threshold for a recommendation of prompt endoscopy in patients with uninvestigated dyspepsia should be driven in part by the incidence of gastric and esophageal cancer in the population in which one practises.[8, 39–41] The American Gastroenterological Association's position statement recommended a cutoff age of 45 years for investigations.15 The Canadian Helicobacter pylori Consensus Conference recommended an age threshold of 50 years for endoscopy in patients with new-onset dyspepsia42 but provided no data to support their recommendation. Canadian statistics for the probability of the development of gastric cancer are presented by age in Table 2.43 Both men and women have only a 0.1% probability of receiving a diagnosis of gastric cancer by age 50, but the probability increases with increasing age. Incidence rates of esophageal cancer also increase with increasing age but are lower than those of gastric cancer. The incidence rates of various cancers in Canada and the United States are similar.43 The following esophageal cancer rates per 100 000 have been reported for the United States: 1.1 for people aged 40 to 44 years, 2.7 for those aged 45 to 49, 5.6 for those aged 50 to 54, and 10.2 for those aged 55 to 59.44
Another important aspect of prompt investigation is that normal findings on endoscopy can provide reassurance for both the patient[45–47] and the physician. It was the consensus of the CanDys Working Group that 50 years is a reasonable cutoff age for prompt investigation in patients with uninvestigated dyspepsia.41 A referral for further investigation is always based on clinical judgement. For example, when assessing a patient who presents with dominant heartburn symptoms, the patient with stable symptoms responsive to antacids for many years must be considered differently from the older patient with recent onset of heartburn.
There are no prospective studies evaluating the value of alarm features. There is consensus in the literature that the presence of alarm features warrants prompt investigation. The acronym "VBAD" can serve as a memory cue for these alarm features, which include (persistent) vomiting, evidence of gastrointestinal bleeding or anemia, abdominal mass or unexplained weight loss, and dysphagia.
Diagnostic test: endoscopy or radiography? Barium meal studies are commonly used by Canadian family physicians for investigating dyspepsia.2 However, it has repeatedly been shown that endoscopy is superior to radiography of the upper gastrointestinal tract for the diagnosis of organic causes of dyspepsia.[8, 48–52] Although the sensitivity (proportion of true-positive results) and specificity (proportion of true-negative results) of radiography varied between studies, in a study directly comparing double-contrast barium meal and endoscopy a diagnostic accuracy of 70% was reported for radiography, compared with 96% for endoscopy (p < 0.001).49
The choice of test may vary according to where one practises, and clinical judgement must be applied in each case. The time it takes to get a referral to a gastroenterologist or to have endoscopy of the upper gastrointestinal tract performed may influence the decision as to which test is ordered.2 Furthermore, a normal x-ray film will provide reassurance to the physician and patient that a serious cause for the dyspepsia symptoms is less likely.
Box C: Patients who use NSAIDs
Chronic use of a conventional NSAID, including ASA, is associated with an increased frequency of gastric and duodenal ulcers. Endoscopic studies indicate that the prevalence of peptic ulcers in people who use NSAIDs regularly is between 10% and 30%.[53, 54] A meta-analysis showed that users of NSAIDs have approximately a 3 times greater relative risk for the development of serious adverse gastrointestinal events than nonusers.55 There are differences among NSAIDs in the frequency with which they cause peptic ulcers.[56, 57] Although H. pylori infection is clearly the most common cause of peptic ulcers, NSAIDs are responsible for most H. pylori-negative ulcers.[58–61] It is unclear whether there is synergy between H. pylori and NSAIDs in causing ulcer. Any interaction is likely to be small.[62, 63]
COX-2 inhibitors have recently become available and appear to have a safer gastrointestinal profile than conventional NSAIDs.64 Their development is based on the notion that the COX-1 enzyme predominates in the stomach, producing protective prostaglandins in the gastric mucosa. In contrast, the COX-2 enzyme is preferentially induced by inflammation, leading to pain and swelling. Although in reality the balance between COX-1 and COX-2 production is more complex, there is strong evidence that the COX-2 inhibitors have fewer side effects, such as gastric and duodenal ulcers and possibly dyspepsia, than conventional NSAIDs.[64–70]
Although most physicians believe that both ASA and non-ASA NSAIDs lead to dyspeptic symptoms, the actual data are controversial, especially in younger patients. When ASA use was assessed as a potential risk factor for dyspepsia in subjects less than 65 years of age with uninvestigated dyspepsia, two studies did not show a significant association.[71, 72] One subsequent study did demonstrate a significant association between dyspepsia and ASA intake, but a dose-response curve could not be detected.73 The data showing that non-ASA NSAIDs cause dyspepsia are not consistent.[72, 73] However, a recent meta-analysis, available at present only in abstract form, indicates that NSAIDs are clearly associated with dyspepsia.74 Studies of the prevention of NSAID-related ulcers showed that acid suppression lowers the frequency of ulcers and reduces dyspepsia symptoms in the absence of an active ulcer.[75, 76] The data on COX-2 inhibitors also suggest that patients taking these drugs have a slightly lower frequency of dyspepsia than those taking conventional NSAIDs.[67, 68]
Few trial data are available to assess whether NSAID use is predictive of organic disease in patients with dyspepsia, especially in those aged 50 years or less. One study, involving 109 patients with arthritis who had been taking NSAIDs in therapeutic dosages for at least 4 weeks, showed that upper abdominal complaints were an independent predictor of ulceration.75 The report did not provide data for young versus older patients, although the demographic characteristics of the study participants indicated that 75% were over 45. In a retrospective UK study assessing 8156 patients, no difference was found in the prevalence of endoscopic abnormalities between patients who used ASA or other NSAIDs and nonusers.76 The applicability of these data to patients aged 50 years or less who present with uninvestigated dyspeptic symptoms is unknown. Despite the lack of definitive data, there was consensus among the CanDys Working Group members that NSAIDs, including ASA, should be considered as potential causes of dyspepsia. If there are no alarm features, investigation is not warranted. The first step in management is to stop the NSAID use if possible and to determine whether the patient's condition subsequently improves.
Differential diagnosis of uninvestigated dyspepsia in patients aged 50 years or less with no alarm features
The division of dyspepsia into subgroups, based on clusters of symptoms, was first proposed in 1988.77 The subgroups ulcer-like, reflux-like and dysmotility-like dyspepsia were intuitively attractive because they coincide with beliefs about the cause of symptoms: acid for ulcer-like and reflux-like dyspepsia, and motility abnormalities for dysmotility-like dyspepsia. It was initially considered that subclassification had the potential to guide the choice of treatment.[77, 78] However, the usefulness of subclassification has not been evaluated in light of the knowledge that most ulcer disease is attributable to H. pylori infection or NSAID use. Symptom subgroups have subsequently been found to have a poor predictive value for endoscopic diagnosis.[31, 79, 80] Furthermore, several studies have shown considerable overlap between the subgroups, both for patients with uninvestigated dyspepsia and for those with functional dyspepsia.[17, 29, 31, 38, 80–83] In a recent study assessing the value of clinical judgement in predicting the endoscopic diagnosis in patients with uninvestigated dyspepsia, the overall accuracy of clinical diagnosis was only 57%.84 In addition, one-third of patients with a major endoscopic finding, such as ulcers, were misclassified as having functional dyspepsia.
Box D: Identification of patients with dominant symptom of heartburn or acid regurgitation, or both
Although symptom clusters have generally proven to be poor predictors in the differential diagnosis of dyspepsia, Klauser and colleagues85 found that when heartburn or acid regurgitation are dominant symptoms, they have a high specificity (89% and 95% respectively) for GERD. The presence of heartburn or acid regurgitation as dominant symptoms is now widely recognized as a reliable indicator that a patient has GERD.[8, 13, 86–88] A study conducted in the primary care setting of 3 European countries assessed the clinical features and endoscopic diagnoses in patients presenting with GERD.89 The results demonstrated that primary care physicians were able to diagnose GERD accurately based on dominant symptoms. This finding supports the approach that initial treatment can be started based on symptoms of reflux in primary care.
An important point for the diagnosis of GERD is that most affected patients do not have macroscopic esophagitis.[8, 90–94] These patients are considered to have endoscopy-negative reflux disease. Endoscopy is therefore not a useful diagnostic "gold standard" for GERD, nor is 24-hour pH monitoring.[90–99]
A reliable interpretation of the term "heartburn" is key for the diagnosis of GERD.[99, 100] The term is often misinterpreted by patients and described as "pain or discomfort in the stomach."87 If heartburn is described as a sensation of "a burning feeling rising from the stomach or lower chest toward the neck," patient recognition is higher.87 Furthermore, this definition predicted the response to treatment with antisecretory agents.87
Box E: Helicobacter pylori "test-and-treat" strategy for patients with uninvestigated dyspepsia without alarm features
There is little doubt that the isolation of H. pylori is one of the most important medical discoveries in the last 20 years. Treatment of infection with this organism has revolutionized the management of duodenal and gastric ulcer disease.101 H. pylori infection is associated with 90% to 95% of duodenal ulcers[88, 101] and 60% to 80% of gastric ulcers.[60, 61, 88] A systematic review of the literature disclosed moderate epidemiologic evidence for an association between chronic H. pylori infection and gastric cancer.102 The International Agency for Research on Cancer103 has classified H. pylori as a group I (definite) carcinogen in humans. However, development of gastric cancer is a multifactorial process that includes other factors, such as a diet low in vitamin C or high in salt, and smoking. As mentioned, the lifetime risk for gastric cancer is low in Canada,43 and the incidence has gradually decreased over the last 40 years.
Currently, there is uncertainty as to whether H. pylori plays a role in dyspepsia in the absence of ulcers. Four separate reviews have evaluated whether there is an association between H. pylori and functional dyspepsia,[102, 104–106] and none reached a definitive conclusion. Five recent randomized placebo-controlled trials assessing the value of therapy to eradicate H. pylori in patients with functional dyspepsia produced conflicting results: 3 studies showed no significant reduction of dyspeptic symptoms after 1 year,[107–109] and 2 showed that eradication of the organism did resolve symptoms in 25% to 30% of patients.[110, 111] One explanation for the 2 studies with positive results is that they were conducted in Scotland and Ireland, both countries with a high prevalence of peptic ulcer disease.
The distinction between functional (investigated) and uninvestigated dyspepsia is relevant when a noninvasive H. pylori test-and-treat strategy is considered. Among patients with uninvestigated dyspepsia who harbour H. pylori, there will be a proportion (estimated at 5% to 15%) with peptic ulcer disease; these patients will benefit from eradication of the organism. Perhaps some patients in whom gastric cancer will ultimately develop will also benefit from eradication of H. pylori, although there are no clinical trial data to prove this hypothesis. Despite the lack of direct evidence for prevention of gastric cancer, this is a consideration when a decision regarding treatment of H. pylori infection is made.
The main options for the treatment of younger patients with uninvestigated dyspepsia who have no alarm features include the following: (1) a trial of empiric (antisecretory or prokinetic) therapy (other than H. pylori-eradication therapy), with investigation reserved for symptomatic failures; (2) diagnostic evaluation (preferably endoscopy) for all, with treatment based on the findings of the diagnostic test; (3) noninvasive testing for H. pylori followed by eradication therapy for patients with positive results ("test-and-treat"); and (4) noninvasive testing for H. pylori followed by endoscopy for patients with positive results.15 The last option is based on the assumption that H. pylori infection should be treated only if it has led to macroscopic disease, such as ulcer. Data from randomized clinical trials comparing some of these strategies are just emerging, but it should also be noted that the Canadian Helicobacter pylori Consensus Conference recommended that eradication therapy be offered to all patients with a positive result of testing for H. pylori.112
A prospective randomized study assessing patients with uninvestigated dyspepsia showed that a strategy of empiric H2-blocker therapy was associated with lower patient satisfaction and higher costs than a strategy of prompt endoscopy,113 although at 1 year there were no differences between the 2 groups in overall severity of dyspeptic symptoms or improvement in quality of life. A more recent randomized controlled trial involving patients with persistent dyspeptic symptoms demonstrated that empiric treatment with a PPI resulted in 69% fewer diagnostic endoscopy procedures, lower medical costs and equal effectiveness in the first year compared with a strategy of prompt endoscopy.114
Two recent randomized trials conducted in primary care settings compared a strategy of prompt endoscopy with a test-and-treat strategy. Lassen and associates115 reported that the test-and-treat strategy was highly cost-effective after 1 year of follow-up. There were no differences in severity of dyspeptic symptoms between the strategies, and the number of endoscopy procedures was reduced by 63% in the test-and-treat group. The prompt-endoscopy group, however, reported greater satisfaction with their treatment. Jones and coworkers116 also reported that, at 1 year, the test-and treat strategy resulted in a significant reduction in the number of endoscopy procedures and substantially reduced costs compared with a strategy of prompt endoscopy. In another randomized trial, among patients less than 45 years of age with dyspepsia who harboured H. pylori, empiric anti-H. pylori therapy reduced dyspepsia severity scores and increased the scores for several measures of quality of life more than a strategy of endoscopy with subsequent treatment based on the results of the procedure.117 Results from other studies also support the cost-effectiveness of the H. pylori test-and-treat strategy.[118–120]
Decision analysis models comparing the cost-effectiveness of various strategies suggest that a noninvasive test-and-treat strategy would reduce the endoscopy workload and be cost-effective.[121–123] A recent Canadian analysis model that used Canadian data on health care costs also suggested that empirically based management strategies for patients under 45 years of age with uninvestigated dyspepsia are cost-effective compared with endoscopy or barium investigation.124 Another model suggested that a strategy based on urea breath testing (UBT) in the same patient population is more cost-effective than empiric antisecretory therapy.125 The first randomized controlled Canadian trial of a noninvasive H. pylori test-and-treat strategy in patients with uninvestigated dyspepsia in primary care has just been completed.126 It showed a 14% benefit in favour of H. pylori treatment over placebo (50% v. 36%), thus supporting the test-and-treat strategy.
The feasibility of the various management options in the Canadian primary care setting is influenced by the time it takes for the patient to be seen by a gastroenterologist after a referral has been made, the availability and waiting times for endoscopy and the availability of noninvasive tests for H. pylori. In a recent survey of Canadian family physicians, 70% of the respondents indicated that the estimated mean delay of 5 weeks for a gastrointestinal referral adversely influenced their decision to refer patients.2
Family physicians have serologic testing and UBT available as noninvasive tests. The carbon-14 (radioactive) UBT is available only in a few large urban centres. However, the carbon-13 (nonradioactive, stable) UBT is becoming increasingly available. Unfortunately, for both carbon-13 and carbon-14 UBTs the cost is not as yet reimbursed by provincial health ministries or most insurance companies, despite the recommendation of the Canadian Helicobacter pylori Consensus Conference that it is the preferred noninvasive test for H. pylori.112 Most general practitioners in Canada have access only to serologic testing for the diagnosis of H. pylori infection. There is evidence that a negative H. pylori test result may provide as much reassurance for patients as a normal result of endoscopy.[127, 128] This is discussed further in the section "Testing for H. pylori infection."
In deciding on a noninvasive H. pylori test-and-treat strategy in patients with uninvestigated dyspepsia, it is necessary to consider both the potential benefits and the drawbacks. There are several arguments in favour of a test-and-treat option. First, some patients with functional dyspepsia may benefit; the overall magnitude of benefit remains controversial but is likely under 20%.[106–110] Second, patients with undiagnosed duodenal and gastric ulcers will benefit. The estimated lifetime prevalence of peptic ulcer disease in the general population is 5% to 15%.129 Third, eradication may halt the progression from chronic gastritis to gastric cancer and hence prevent this cancer. Although there are no data from longitudinal studies to support this hypothesis, results from a modelling study130 and a nonrandomized Japanese study131 suggest this may be the case in selected patient populations. Finally, there is a strong desire among many patients to undergo testing and be treated for H. pylori infection, as public awareness of the infection and its consequences is high. This is not in itself a reason to test for H. pylori infection, but symptoms alone are inadequate to exclude this diagnosis in a patient with dyspepsia.
Testing for H. pylori infection: Recent guidelines from the Canadian Helicobacter pylori Consensus Conference include an overview of diagnostic tests for the detection of H. pylori infection.42 Infection can be detected by invasive (endoscopy based) or noninvasive (UBT or serologic testing) diagnostic tests. For serologic testing, both laboratory (serum) and office-based (whole blood) tests are available.20 Because different commercial and noncommercial serologic assays are available, it is important that they be validated locally.42
Comparison of serologic testing and UBT has shown the latter to be consistently superior for the diagnosis of H. pylori infection.[132–138] An important difference between UBT and serologic testing is that the latter cannot be used to determine cure, as the IgG antibodies against H. pylori remain detectable for a long time (6 to 12 months) after eradication of the organism.[139–141]
In the interpretation of a diagnostic test, the positive and negative predictive values are the most important. In contrast to the sensitivity and specificity of the test, the positive and negative predictive values are influenced by the prevalence of the disease one wants to diagnose. In Canada the overall prevalence of H. pylori infection is estimated to be 30% to 40% and increases with increasing age (8% increase per decade).142 The prevalence is higher among patients with a lower socioeconomic status and probably also among immigrants from countries where H. pylori infection is endemic.[143, 144]
Calculation of positive and negative predictive values using the mean sensitivity and specificity values from a meta-analysis of serologic testing136 and Canadian data on the prevalence of H. pylori infection142 shows that serologic testing has a high negative predictive value (90%) in young patients (Table 3). A younger patient is therefore unlikely to be infected with H. pylori if the test result is negative. In contrast, below the age of 50 years, the positive predictive value ranges from 52% to 72%, which indicates that false-positive results do occur. UBT has a high positive predictive value and negative predictive value (both greater than 90%), even in groups with a low prevalence of H. pylori infection. Therefore, it is the preferred test.146 However, UBT is currently not widely available for primary care physicians in Canada. If neither UBT nor serologic testing is available, endoscopy plus gastric biopsy is an alternative method to assess H. pylori status.
Evidence related to the 4 mini-management schemata
Management of patients who use NSAIDs
There is consensus that NSAIDs can cause dyspeptic symptoms.74 The evidence is less compelling for patients who use ASA regularly (i.e., daily). However, since even low-dose ASA can lead to ulcer formation, it seems likely that ASA may sometimes cause dyspepsia. If possible, NSAID use should be stopped and the patient's response monitored closely.58 If the patient must continue with NSAID therapy (including ASA) or if symptoms have not resolved after NSAID therapy is stopped, the patient should be treated empirically or referred for investigation.
Ulcer prevention is an important consideration in patients taking NSAIDs. Randomized placebo-controlled trials of at least 2 months' duration have shown that treatment with misoprostol (600 to 800 μg/d),[147, 148] high doses of famotidine (40 mg twice daily)149 and omeprazole (20 mg once daily)[150, 151] provide effective prophylaxis against NSAID-related peptic ulcer in patients receiving long-term NSAID therapy. In addition to famotidine, other H2-RAs have been studied for the prevention of NSAID-induced ulcers. Randomized placebo-controlled studies have shown that ulcer occurrence with nizatidine treatment (150 mg twice daily) was not statistically different from that with placebo,152 whereas therapy with ranitidine at a standard dosage (150 mg twice daily)[153, 154] and at 300 mg twice daily155 was protective against duodenal ulcer but not against gastric ulcer in long-term NSAID users. A randomized study comparing misoprostol (200 μg 4 times daily) and ranitidine (150 mg twice daily) demonstrated that misoprostol was more effective than ranitidine in preventing NSAID-induced gastric ulcers, but the 2 medications had comparable efficacy in the prevention of NSAID-induced duodenal ulcers.156
Randomized controlled trials that assessed both healing and prevention of ulcers associated with NSAID use have demonstrated the effectiveness of high-dose famotidine therapy (40 mg twice daily) compared with placebo157 and the superiority of omeprazole (20 or 40 mg once daily) compared with misoprostol (200 μg 4 times daily)158 and ranitidine (150 mg twice daily).159 Another recent study showed that lansoprazole (15 or 30 mg/d) was significantly better than ranitidine (150 mg twice daily) in healing NSAID-induced gastric ulcers in patients who continued to take these agents.160 Data on the prevention of NSAID-related ulceration are not yet available for pantoprazole and lansoprazole.
The COX-2 inhibitors celecoxib and rofecoxib have been shown to be much safer than conventional NSAIDs, causing fewer gastric and duodenal ulcers and upper gastrointestinal tract symptoms.[60, 74] To date, these drugs have been studied only in acute pain syndromes and in patients with rheumatoid arthritis or osteoarthritis.[65–69] Intuitively, it seems reasonable to change from a conventional NSAID to a COX-2 inhibitor in patients with dyspeptic symptoms associated with NSAID use since COX-2 inhibitor therapy produces a lower rate of dyspepsia and ulcer-related complications. However, this strategy has not been tested specifically in a patient population with symptoms or complications from NSAID therapy, and the evidence for a change is therefore anecdotal at best. Furthermore, there is experimental evidence in animals that COX-2 inhibitors may delay ulcer healing.[161, 162] There must therefore be reservations about a change from NSAIDs to COX-2 inhibitors in patients with uninvestigated dyspepsia who may have undiagnosed ulceration.
Because none of the NSAID studies included patients aged 50 years or less with uninvestigated dyspepsia, they are not directly applicable to this mini-management schema. Most of the studies reviewed by the group focused on ulcer prevention and ulcer healing. However, only studies evaluating omeprazole also showed benefit for dyspeptic symptoms compared with misoprostol and ranitidine.[158, 159] Although several of the studies were level I randomized controlled trials for healing and prevention of ulcers, only grade C recommendations can be made for dyspepsia. There are also data on high-dose famotidine therapy (40 mg twice daily) showing a preventive effect on ulcer formation. A single study assessing ranitidine (300 mg twice daily) showed prevention of duodenal ulcers but not of gastric ulcers. Clearly, more studies are needed in this area.
Management of patients with gastroesophageal reflux disease
Patients aged 50 years or less with uninvestigated dyspepsia who have dominant symptoms of heartburn or acid regurgitation, or both, should be managed as patients with GERD. The treatment goal is symptom relief. Five treatment possibilities for GERD were assessed: lifestyle modification, antacids, H2-RAs, prokinetics and PPIs.
Although the CanDys Working Group recognizes that lifestyle modifications are commonly recommended for patients with GERD, there are sparse controlled data available. Most studies assessing lifestyle modification involved patients with moderate or severe GERD. Elevation of the head of the bed failed to decrease the frequency of reflux episodes significantly in 2 studies,[163, 164] but it reduced symptoms and improved endoscopic appearance in patients with severe esophagitis in another study.165 The authors of a more recent work found that gastroesophageal reflux in obese patients did not abate after weight reduction.166 Clinical experience suggests that patients with milder symptoms of GERD may derive benefit from lifestyle modification.
In a review of lifestyle modification and medical therapy with antacids in GERD, Kitchin and Castell167 concluded that definitive evidence of efficacy is unavailable because of the lack of well-controlled trials. The review included several studies assessing the efficacy of antacids versus placebo. Three of these studies showed comparable efficacy for antacids and placebo,[168–170] and 2 studies showed that antacids provided benefit compared with placebo.[171, 172]
A recent meta-analysis of 43 randomized studies involving adults with endoscopically proven erosive esophagitis (grade II to IV GERD) showed that more complete esophageal healing and heartburn relief was obtained with PPIs than with H2-RAs.173 In addition, the speed of healing and symptom relief was nearly twice as fast with PPIs. However, these studies were restricted to patients with more severe disease than that of the average patient who presents in primary care with symptoms of milder GERD.174 The meta-analysis included studies using lansoprazole, omeprazole and pantoprazole.
As noted previously, most patients with GERD will not have esophagitis but will have endoscopy-negative reflux disease. PPIs have been reported to be superior as initial treatment in these patients.[92, 93, 175] A 4-week study in a general practice setting demonstrated that omeprazole (20 mg once daily) produced a significantly higher rate of symptom relief than ranitidine (150 mg twice daily) among patients with GERD (61% v. 40%).94 Similarly, in another 4-week randomized study, heartburn was resolved in significantly more patients treated with omeprazole (20 mg once daily) than in those who received cisapride (10 mg four times daily) (65% v. 41%).91 Symptom relief was comparable in patients with or without endoscopic evidence of erosive esophagitis.
The evidence indicates that PPIs are also superior for endoscopy-negative reflux disease. Most data to date are for omeprazole, but studies with lansoprazole and pantoprazole are emerging.[176, 177]
A recent analysis of the efficacy of prokinetic therapy for GERD revealed that cisapride is the only prokinetic agent that can be recommended.178 The analysis included 4 placebo-controlled studies showing that cisapride produced significantly higher healing rates than placebo among patients with mild to moderate esophagitis.[179–182] Six additional studies demonstrated that cisapride and H2-RA therapy produced similar healing rates.[183–188] The combination of cisapride with either an H2-RA[189, 190] or omeprazole191 did not produce an additional benefit in healing after 6 to 8 weeks of therapy. Furthermore, among 11 placebo-controlled studies evaluating resolution of reflux symptoms, cisapride produced greater symptom relief than placebo in 10 studies and comparable relief to placebo in 1 study.178 The symptom relief seen with cisapride (39%) was comparable to that with H2-RAs (32%) in 6 studies.178 Comparison of the results from studies of therapy for endoscopy-negative reflux disease also support the notion that cisapride and H2-RAs produce comparable relief of reflux symptoms.[91–94, 192] However, both treatments produced significantly lower rates of complete symptom relief than omeprazole.[91, 193, 194] Concerns have been raised both in Canada and in the United States about potentially serious cardiac side effects associated with cisapride.[195, 196] The side effects include prolongation of the Q-T interval, ventricular tachycardia and death. Interactions can occur with other medications, such as antiarrhythmics, antidepressants and antihistamines.
There is a progressive increase in the proportion of patients who become symptom-free as the duration of therapy increases, regardless of the type of therapy; however, the proportion of patients who become symptom-free each week is highest during the first 2 to 4 weeks of therapy, and the rate of relief is greatest among patients receiving PPIs.[91, 173] Most patients who will become symptom-free with PPI therapy will have done so by 4 weeks, and those who have not responded to treatment with H2-RAs or cisapride by 4 weeks are less likely to respond to another 4 to 8 weeks of therapy. Thus, reassessment of the patient after 4 weeks of therapy is reasonable.
Although the efficacy of a standard-dose H2-RA and of cisapride is essentially equivalent, the former is preferred because of the potential adverse events associated with cisapride therapy.197
Currently, there is debate as to whether a step-up approach (start with H2-RA and switch to PPI if no response) or a step-down approach (start with PPI and switch to H2-RA if the patient responds) is better.193 Studies in primary care that enrolled patients with reflux symptoms showed that the severity of heartburn and the duration of symptoms were similar in patients who were subsequently identified as having endoscopy-positive GERD (esophagitis) or endoscopy-negative GERD.198 The implication is that neither the severity of disease nor the duration of symptoms is a reliable predictor of treatment response. The data clearly show that PPIs are superior to H2-RAs and cisapride. PPIs provide more rapid and more complete relief of symptoms and, if erosive lesions are present, higher rates of healing of the esophagitis.
In the recommendations, PPIs are listed as first choice over H2-RAs based on efficacy data. Few studies directly compared the various PPIs or H2-RAs with each other. However, because healing rates and data on symptom relief are similar within each drug class, the class is listed rather than individual compounds. The current uncertainty regarding the step-up and step-down approaches is driven mainly by cost considerations. Therefore, the initial approach to management should be determined by the physician's assessment of the patient's clinical needs and the cost of medications.
Management of patients with a positive result of testing for Helicobacter pylori
H. pylori eradication therapy is the standard of care for all patients with duodenal or gastric ulcers who have been shown to harbour the organism.[58, 199–201] Because of the compelling evidence that H. pylori is a true pathogen, there is an increasing consensus worldwide to treat all patients with positive results of testing for H. pylori.42
Helicobacter pylori eradication therapy
In accordance with the recommendations of the Canadian Helicobacter pylori Consensus Conference, first-line eradication therapies for H. pylori are triple therapies of a PPI plus 1000 mg of amoxicillin plus 500 mg of clarithromycin (PPI + AC), or a PPI plus 500 mg of metronidazole plus 250 or 500 mg of clarithromycin (PPI + MC), twice daily for 1 week; or ranitidine bismuth citrate plus either AC or MC.42 These triple therapies achieve eradication rates of about 85% to 90%. For the PPI-based therapies, lansoprazole, omeprazole or pantoprazole can be used.
If the first eradication therapy has failed, the action recommended by the Canadian Helicobacter pylori Consensus Conference is to use a different first-line therapy than that used initially (e.g., switch from PPI + AC to PPI + MC). An alternative therapy is a 14-day quadruple regimen of a PPI (twice daily) plus bismuth (subsalicylate, 2 tablets 4 times daily) plus metronidazole (250 mg 4 times daily) plus tetracycline (500 mg 4 times daily) (PPI + BMT).42
Management of patients with a negative result of testing for Helicobacter pylori
Most clinical trials of drug therapy for dyspepsia have assessed patients with functional dyspepsia. These patients have been investigated (usually by endoscopy) before enrolment, and no organic cause of the dyspepsia symptoms has been detected. It is uncertain whether the evidence from these trials can be extrapolated to the patient who presents in primary care with uninvestigated dyspepsia and subsequently is found to be H. pylori negative.
A systematic review evaluating drug treatment of functional (nonulcer) dyspepsia showed that many studies had methodologic weaknesses, including small sample, short duration and use of unvalidated outcome measures.6 Interpretation of these studies is also hampered by the high placebo response rates, ranging from 13% to 73%, with values typically ranging from about 25% to 55%.
The CanDys Working Group evaluated 4 treatment possibilities for patients who present with uninvestigated dyspepsia and are subsequently found to be H. pylori negative: antacids, H2-RAs, prokinetics and PPIs.
Four randomized placebo-controlled studies evaluated the efficacy of antacids in the treatment of functional dyspepsia, and all 4 failed to demonstrate superiority of antacids over placebo.[202–205]
In a recent meta-analysis that identified 19 eligible studies, the efficacy of H2-RAs was compared with placebo in patients with functional dyspepsia.206 The odds ratios for the following outcomes were 2.3 (95% confidence interval [CI] 1.6-3.3) for lessening of epigastric pain, 1.8 (95% CI 1.2-2.8) for complete relief of epigastric pain and 1.5 (95% CI 0.9-2.3) for global assessment of improvement. The results for epigastric pain were statistically significant, whereas the 95% CIs for the global assessment included 1.0, which indicates a nonsignificant result. However, owing to the differences in outcome measures in the studies, data from many eligible studies could not be pooled for the statistical analysis. The cautious interpretation of these findings is that there is a modest benefit from H2-RAs, but the authors are guarded in their conclusion.
A meta-analysis of randomized controlled trials assessing cisapride for the treatment of functional dyspepsia included 19 studies.207 Global assessment of symptoms by the physician or patient was used as the outcome measure in the analysis and was rated on a 4-point scale (no change, or mild, good or excellent response). The odds ratios in favour of cisapride were 2.8 (95% CI 1.5-5.1) for the 12 studies in which the response could be categorized as excellent and 3.3 (95% CI 2.1-5.2) for the 14 studies in which the response could be categorized as good or excellent. These results suggest that there is a modest benefit of cisapride in patients with functional dyspepsia. However, caution is needed in the interpretation because several of the studies had methodologic shortcomings, including small samples.
Most of the trials of PPIs in this area have been conducted with omeprazole. The recent omeprazole studies had large samples and used better-validated outcome measures than earlier trials.6 Two placebo-controlled trials of functional dyspepsia showed that omeprazole was significantly better than placebo in providing complete resolution of dyspeptic symptoms.[208, 209] The larger of the 2 studies involved 1262 patients with functional dyspepsia.209 After 4 weeks of treatment, complete relief of epigastric pain or discomfort was observed in 38% of the patients who received standard-dose omeprazole (20 mg once daily) (p = 0.002) and 36% of the patients who received low-dose omeprazole (10 mg once daily) (p = 0.02), as compared with 28% of the patients who were treated with placebo. Patients were classified in dyspepsia subgroups according to the most bothersome symptoms. Omeprazole was superior to placebo for complete symptom relief in patients with ulcer-like dyspepsia (40% v. 27%) (p < 0.05) and reflux-like dyspepsia (54% v. 23%) (p < 0.05) but not in those with dysmotility-like dyspepsia (32% v. 31%).
A randomized placebo-controlled study involving 269 patients with functional dyspepsia treated with lansoprazole (15 mg once daily) showed superior symptom resolution rates after 2 weeks of treatment compared with placebo (62% v. 44%).210 There was no difference in symptom resolution rates between lansoprazole and placebo in the subgroup of patients who were H. pylori negative, but the study did not have enough power to assess this population properly.
Three large randomized trials of omeprazole have been conducted involving patients with dyspepsia in general practice. One trial compared omeprazole (10 mg once daily) with antacid-alginate liquid (10 mL 4 times daily) for 4 weeks.211 The second trial compared omeprazole (10 mg once daily, increasing to 20 mg and 40 mg once daily as required) with antacid-alginate-ranitidine therapy (10 mL of antacid-alginate 4 times daily, stepping up to 150 mg of ranitidine twice daily and 150 mg of ranitidine 4 times daily as required) in uninvestigated dyspepsia.212 The final study had 3 treatment arms and compared omeprazole (20 mg once daily) with cimetidine (400 mg twice daily) and placebo in patients with ulcer-like or reflux-like dyspepsia.213 Omeprazole was found to be significantly superior to every other treatment strategy assessed in these trials. Another study in general practice showed that superior symptom relief was provided by lansoprazole (compared with ranitidine) in patients who presented with ulcer-like and reflux-like dyspeptic symptoms.214 This study largely included patients with documented GERD or peptic ulcer disease (approximately 70%) and did not provide information about the patients with uninvestigated dyspepsia or analysis according to H. pylori status.
Although most studies in this area have used omeprazole, it was the consensus of the CanDys Working Group that other PPIs (lansoprazole[210, 214] and pantoprazole [no data]) would likely show comparable efficacy. For this reason, PPIs are listed together for the management schema.
Conclusion
The CanDys Working Group's clinical management tool consists of 5 key steps in the evaluation of patients with uninvestigated dyspepsia. The tool also includes 4 mini-management schemata. The tool is practical, easy to use, explicit and concise, and it reflects the realities of the primary care setting. We believe that adoption of this tool will optimize the treatment of patients with dyspepsia, improve quality of care and be cost-effective.
Recommendation
Exclude other possible causes of the dyspeptic symptoms with thorough history-taking and physical examination. Consider cardiac and hepatobiliary sources as well as medication-induced symptoms, possible dietary indiscretions, lifestyle or other causes (grade C recommendation, consensus).
Recommendations
Prompt investigation is recommended for patients over 50 years of age with uninvestigated dyspepsia and for any patient presenting with alarm features. Alarm features include persistent vomiting, evidence of gastrointestinal bleeding or anemia, presence of an abdominal mass or unexplained weight loss, and dysphagia (grade B recommendation, level III evidence[33, 34, 37, 43]).
Endoscopy is the recommended method of investigation for patients with uninvestigated dyspepsia who are over 50 years of age or who have alarm features (grade A recommendation, level II evidence8).
Recommendation
Patients with uninvestigated dyspepsia who are regular users of NSAIDs (including ASA) should be identified, and if there are no alarm features, they can be managed without initial endoscopy (grade C recommendation, consensus).
Recommendations
Patients aged 50 years or less with uninvestigated dyspepsia and dominant symptoms of heartburn or acid regurgitation, or both, should be diagnosed as having GERD and be treated accordingly (grade B recommendation, level II-2 evidence[85, 87, 90–94]).
Rather than using the term "heartburn," describing the sensation of "a burning feeling rising from your stomach or lower chest toward your neck" increases the diagnostic accuracy for GERD (grade B recommendation, level II-2 evidence87).
Recommendation
A test-and-treat strategy for uninvestigated dyspepsia in younger patients (aged 50 years or less) who have no alarm features is recommended (grade B recommendation, level I evidence[115–117]).
Recommendation
Noninvasive methods are recommended for the detection of H. pylori in patients aged 50 years or less with uninvestigated dyspepsia who have no alarm features. UBT is the preferred test, and a locally validated serologic test is considered a second option (grade B recommendation, level II-2 evidence[132, 134–138]).
Recommendations
If possible, NSAID use should be stopped and the patient's response monitored (grade C recommendation, level III evidence). If NSAIDs cannot be stopped the choice is to treat or investigate.
Treatment recommendations for patients aged 50 years or less who present with uninvestigated dyspepsia, who have no alarm features and who need to use NSAIDs (including ASA) are as follows:
(a) PPI (grade C recommendation, consensus).
(b) Cytoprotective agent (grade C recommendation, consensus).
(c) High-dose H2-RA therapy (grade C recommendation, consensus).
(d) Consider switch to COX-2 inhibitor (grade C recommendation, consensus)?
Recommendations
The effectiveness of lifestyle modifications and antacids for the treatment of GERD is not proven. Patients with mild GERD symptoms may derive benefit from these treatments (grade C recommendation, consensus).
Treatment recommendations for patients with a dominant symptom of heartburn or acid regurgitation, or both, are as follows:
(a) PPI (grade A recommendation, level I evidence[91, 94]).
(b) H2-RA (grade A recommendation, level I evidence178).
(c) Prokinetic agent (grade A recommendation, level I evidence178).
Patients should be reassessed after 4 weeks of therapy (grade C recommendation, consensus).
Recommendations
Eradication therapies recommended for patients with uninvestigated dyspepsia who are found to be H. pylori positive are as follows:
(a) PPI + AC or MC, or ranitidine bismuth citrate + AC or MC (grade A recommendation, level I evidence42).
(b) Alternative first-line therapy (grade A recommendation, level I evidence42)
(c) PPI + BMT (grade A recommendation, level I evidence42).
Recommendations
There is good evidence that antacids are ineffective for functional dyspepsia, and they are not recommended for the treatment of uninvestigated dyspepsia in patients subsequently found to be H. pylori negative (grade E recommendation, level I evidence[202–205]).
Treatment recommendations for patients who present with uninvestigated dyspepsia and who subsequently have negative results of testing for H. pylori are as follows:
(a) PPI (grade B recommendation, level I evidence[208, 209]).
(b) H2-RA (grade B recommendation, level I evidence206).
(c) Prokinetic agent (grade B recommendation, level I evidence207).
The following people were members of the CanDys Working Group and contributed to this work: Dr. Bernard Marlow, Department of Family and Community Medicine, University of Toronto, Toronto, Ont.; William Bartle, University of Toronto, Pharmacy Department, Sunnybrook Health Science Centre, Toronto, Ont.; Dr. Brian Craig, Department of Family Medicine, Dalhousie University, Department of Family Medicine, Atlantic Health Sciences Corporation, Saint John Regional Hospital Facility, Saint John, NB; Dr. Jean-Guy Émond, Département de médecine générale, Centre hospitalier universitaire de Québec, Pavillon Enfant-Jésus, Quebec City, Que.; Dr. Krishnasamy Govender, Department of General Practice, Regina Health District, Regina, Sask.; Dr. Richard McCammon, Physician Manager (retired), Salvation Army Grace General Hospital, Winnipeg, Man.; Dr. Wayne Putnam, Department of Family Medicine, Dalhousie University, Family Medicine Centre, Queen Elizabeth II Health Sciences Centre, Halifax, NS; and Dr. Robert Woodland, Department of Family Practice, Health Care Corporation, St. John's, Nfld.
This project was funded by unrestricted educational grants from AstraZeneca Canada Inc., Isotechnika Inc. and Integrated Healthcare Communications Inc.
Dr. van Zanten is the recipient of a Nova Scotia Clinical Scholar Award.
Competing interests: Receipt of consultation fees from one or more pharmaceutical companies (S.v.Z., N.C., D.A., A.B.); employment with AstraZeneca Canada Inc. (P.S.); receipt of honoraria from Integrated Healthcare Communications Inc. for participation in the development and writing of the article (S.v.Z., N.F., N.C., D.A., A.B., M.B., A.T., F.B., P.B., D.F.); receipt of funding from one or more pharmaceutical companies for research related to material contained in the article (S.v.Z., N.C., A.B.); speaker fees from one or more companies (S.v.Z., N.F., N.C., D.A., A.B., M.B., F.B., D.F.); and travel assistance from one or more companies to attend meetings related to the project or to material in the article (S.v.Z., N.F., N.C., D.A., A.B., M.B., F.B., P.B., D.F.).
Footnotes
-
Information about the authors appears at the end of the article.
This article has been peer reviewed.
Reprint requests to: Dr. Sander J.O. Veldhuyzen van Zanten, Rm. 928, South Wing, Centennial Building, Queen Elizabeth II Health Sciences Centre - Victoria General Site, Halifax NS B3H 2Y9; fax 902 473-4406; [email protected]
About the authors: Dr. van Zanten is with the Department of Medicine, Queen Elizabeth II Health Sciences Centre, Dalhousie University, Halifax, NS; Dr. Flook is with the Department of Family Medicine, University of Alberta, and the Misericordia Community Health Centre, Edmonton, Alta.; Dr. Chiba is with the Division of Gastroenterology, McMaster University, Hamilton, Ont., and the Surrey Gl Clinic/Research, Guelph, Ont.; Dr. Armstrong is with the Department of Medicine, McMaster University, and the Division of Gastroenterology, Hamilton Health Sciences Corporation, McMaster Site, Hamilton, Ont.; Dr. Barkun is with the Department of Medicine, McGill University, and the Division of Gastroenterology, McGill University Health Centre, Montreal, Que.; Dr. Bradette is with the Department of Medicine, Université Laval, and the Department of Gastroenterology, Centre hospitalier universitaire de Québec, Pavillon Hôtel-Dieu de Québec, Quebec City, Que.; Dr. Thomson is with the Department of Medicine, University of Alberta, Edmonton, Alta.; Dr. Bursey is with the Health Sciences Centre, Memorial University of Newfoundland, St. John's, Nfld.; Dr. Blackshaw is with the Surrey Memorial Hospital, Surrey, BC; Ms. Frail is with the College of Pharmacy, Dalhousie University, Halifax, NS; and Mr. Sinclair is with AstraZeneca Canada Inc., Mississauga, Ont.
References
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.
- 11.
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.
- 47.↵
- 48.↵
- 49.↵
- 50.
- 51.
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.
- 82.
- 83.↵
- 84.↵
- 85.↵
- 86.↵
- 87.↵
- 88.↵
- 89.↵
- 90.↵
- 91.↵
- 92.↵
- 93.↵
- 94.↵
- 95.
- 96.
- 97.
- 98.
- 99.↵
- 100.↵
- 101.↵
- 102.↵
- 103.↵
- 104.↵
- 105.
- 106.↵
- 107.↵
- 108.
- 109.↵
- 110.↵
- 111.↵
- 112.↵
- 113.↵
- 114.↵
- 115.↵
- 116.↵
- 117.↵
- 118.↵
- 119.
- 120.↵
- 121.↵
- 122.
- 123.↵
- 124.↵
- 125.↵
- 126.↵
- 127.↵
- 128.↵
- 129.↵
- 130.↵
- 131.↵
- 132.↵
- 133.
- 134.↵
- 135.
- 136.↵
- 137.
- 138.↵
- 139.↵
- 140.
- 141.↵
- 142.↵
- 143.↵
- 144.↵
- 145.
- 146.↵
- 147.↵
- 148.↵
- 149.↵
- 150.↵
- 151.↵
- 152.↵
- 153.↵
- 154.↵
- 155.↵
- 156.↵
- 157.↵
- 158.↵
- 159.↵
- 160.↵
- 161.↵
- 162.↵
- 163.↵
- 164.↵
- 165.↵
- 166.↵
- 167.↵
- 168.↵
- 169.
- 170.↵
- 171.↵
- 172.↵
- 173.↵
- 174.↵
- 175.↵
- 176.↵
- 177.↵
- 178.↵
- 179.↵
- 180.
- 181.
- 182.↵
- 183.↵
- 184.
- 185.
- 186.
- 187.
- 188.↵
- 189.↵
- 190.↵
- 191.↵
- 192.↵
- 193.↵
- 194.↵
- 195.↵
- 196.↵
- 197.↵
- 198.↵
- 199.↵
- 200.
- 201.↵
- 202.↵
- 203.
- 204.
- 205.↵
- 206.↵
- 207.↵
- 208.↵
- 209.↵
- 210.↵
- 211.↵
- 212.↵
- 213.↵
- 214.↵