Family Practice Vol. 21, No. 3, 238-243
Family Practice Vol. 21, No. 3 © Oxford University Press 2004, all rights reserved.
Empirical treatment followed by a test-and-treat strategy is more cost-effective in comparison with prompt endoscopy or radiography in patients with dyspeptic symptoms: a randomized trial in a primary care setting
Department of Gastroenterology, UMC St Radboud, a Medical Diagnostic Centre, b Department of Microbiology, c Department of Radiology and d Department of Internal Medicine, Canisius Wilhelmina Hospital, Nijmegen, The Netherlands
Correspondence to RJF Laheij, Department of Gastroenterology, PO Box 9101, 6500 HB Nijmegen, The Netherlands; E-mail: R.Laheij{at}mdl.umcn.nl
Received 20 May 2003; Revised 5 August 2003; Accepted 7 January 2004.
Laheij RJF, Hermsen JTh, Jansen JBMJ, Horrevorts AM, Rongen RJ, van Rossum LGM, Witteman E and de Koning RW. Empirical treatment followed by a test-and-treat strategy is more cost-effective in comparison with prompt endoscopy or radiography in patients with dyspeptic symptoms: a randomized trial in a primary care setting. Family Practice 2004; 21: 238–243.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Objective. Management of patients with dyspepsia remains controversial. No consensus has yet been reached concerning diagnostic and medical strategies. We conducted a randomized trial to assess the effectiveness of three management strategies for patients with uninvestigated persistent dyspeptic symptoms.
Methods. A total of 199 patients presenting in primary care with dyspeptic symptoms (age 18–65 years, no alarming symptoms) were randomized to either empirical treatment with omeprazole and, in the case of symptomatic relapse, serological Helicobacter pylori infection testing plus eradication therapy (treat-and-test group), prompt upper gastrointestinal endoscopy (endoscopy group) or prompt upper gastrointestinal radiography (radiography group) followed by directed medical treatment. Symptoms, patients' satisfaction and use of resources were recorded during 6 months of follow-up.
Results. Sixty-nine patients were assigned to the treat-and-test group, 64 to the radiography group and 66 to the endoscopy group. The median age was 44 years; 104 patients were male and 37% were H.pylori infected. A total of 170 patients (85%) returned the 6 months questionnaire. The numbers of patients with complete symptom relief in the treat-and-test group, endoscopy group and radiography group were 21, 16 and 15, respectively, at 3 months (P = 0.59), and 23, 13 and 12, respectively, at 6 months (P = 0.05). Twenty-two patients in the treat-and-test group underwent endoscopy or radiography. Two patients in the endoscopy group and four patients in the radiography group underwent more than one diagnostic test. The average medical cost per patient for the treat-and-test group was 
276, for the endoscopy group 426 and for the radiography group 321, respectively.
Conclusion. Empirical treatment followed by a test-and-eradicate strategy resulted in fewer diagnostic tests, more symptom relief and lower medical costs compared with prompt upper gastrointestinal radiography or endoscopy in the management of uninvestigated patients with persistent dyspeptic symptoms.
Keywords. Dyspepsia, endoscopy, radiography, randomized controlled trial, treat-and-test.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Although several cost-effectiveness studies have been conducted, appropriate management of patients with dyspepsia still remains controversial. Due to the high prevalence of dyspepsia in the primary care setting and the cost of endoscopy, it is hardly feasible to investigate every patient with dyspeptic symptoms by endoscopy or radiography; empirical treatment or Helicobacter pylori-guided management strategies may be initiated.1–4 Several years ago, we used decision analysis to determine which treatment strategy results in more appropriate management of patients with persistent dyspepsia.5 In this treat-and-test strategy, we suggested that patients should be evaluated based on symptom response to the proton pump inhibitor omeprazole, recommending endoscopy for only a selective subgroup of patients. We used empirical treatment with omeprazole to discriminate patients with non-acid-related disorders from those with acid-related disorders. However, most patients with acid-related disorders such as reflux oesophagitis and peptic ulcers will have a symptomatic relapse after a short period of therapy with gastric acid inhibition. Therefore, treatment failures received another 8 weeks medication. Furthermore, because H.pylori treatment is essential to avoid relapse in patients with peptic ulcers, a test-and-treat strategy will follow.
Subsequently, we evaluated this treatment strategy in 84 patients who were randomly assigned to prompt endoscopy and the treat-and-test management strategy.6 The results of this study showed that the treat-and-test strategy cost less and was as equally effective as prompt endoscopy followed by directed medical treatment. The difference in cost between the management strategies was most sensitive to the cost of endoscopy.7 Because of this finding, we conducted a new study in which we again compared the treat-and-test strategy with prompt endoscopy and, additionally, we included prompt radiography as an alternative diagnostic approach. Although this mode of investigation is considered suboptimal by gastroenterologists, and most current guidelines do not include radiography in the algorithmic evaluation of dyspepsia, many patients undergo upper gastrointestinal radiographic examination.8
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Patient population
The trial was conducted between September 1998 and July 2000. Patients were selected by 48 GPs in the south-eastern part of The Netherlands. Eligible patients were those aged 18 years or older, with persistent dyspeptic symptoms according to the Rome II criteria and of sufficient severity, as judged by the GP, to justify diagnostic upper gastrointestinal endoscopy. Empirical treatment for dyspeptic symptoms before recruitment was allowed, but the use of proton pump inhibitors was an exclusion criterion. Other exclusion criteria were signs or suspicion of malignancy (food transit complaints, weight loss, anaemia, vomiting of blood), treatment with non-steroidal anti-inflammatory drugs (NSAIDs), previous gastrointestinal surgery, pregnancy or lactation, chronic alcoholism or drug abuse, or lack of motivation. The ethics committees of the University Hospital Nijmegen and the Canisus Wilhelmina Hospital approved the protocol. All the participants gave written informed consent before taking part in the study.
Management strategies
Patients who were eligible were given a patient number in consecutive order according to a computer-generated randomization list (central randomization system) and allocated to three different groups. The treat-and-test group started with omeprazole, 20 mg once daily for 2 weeks. Patients without symptom improvement after this period were referred for upper gastrointestinal endoscopy. Patients whose symptoms improved continued with another 2 weeks of omeprazole treatment. If there was relapse after this period, patients were treated with omeprazole 20 mg once daily, but now for a period of 8 weeks. Patients who presented with a second relapse within the study period were treated according to a H.pylori test-and-eradicate strategy. If they were seropositive, an attempt was made to eradicate the infection with a triple or quadruple therapy regimen. The second group underwent prompt upper gastrointestinal endoscopy followed by directed therapy, and the third group underwent prompt upper gastrointestinal radiography followed by directed treatment. In patients who underwent prompt upper gastrointestinal endoscopy or radiography, H.pylori testing and eradication therapy was performed only on indication of the GP.
Assessments
A global physical examination was performed and each patient's general medical history was recorded. At baseline, H.pylori infection was assessed serologically in all patients. Serological screening for H.pylori infection was performed with the Pyloriset which was locally validated (Pyloriset test kit, Imphos, The Netherlands). The Pyloriset is a commercially available enzyme-linked immunosorbent assay (ELISA) that measures IgG antibodies to H.pylori infection.9 Patients and GPs were not told the test result. The following symptoms were evaluated by questionnaire at baseline, and at 1, 3 and 6 months during follow-up: upper abdominal pain during the day and/or night, pyrosis and regurgitation, bloating and nausea. For each symptom, severity was assessed on a 4-point Likert scale as follows: 1, mild (can be ignored); 2, moderate (cannot be ignored but does not affect lifestyle); 3, severe (affects lifestyle); and 4, very severe (markedly affects lifestyle). Patients were asked to report symptoms over the past month. Furthermore, medical-related quality of life was assessed at the beginning of the study period, and at 1, 3 and 6 months using the Short Form 36. Finally, patient satisfaction was assessed on a visual analogue scale (0 = worst imaginable satisfaction, 10 = best imaginable satisfaction). Questionnaires were completed irrespective of doctors visits.
Economic evaluation
The financial figures used in our study were based on the principles of the societal viewpoint. Real costs were calculated for medical aspects. All financial figures were based on price lists for the year 2000. Costs for upper gastrointestinal endoscopy and radiography were based on the real costs at the University Hospital Nijmegen and Canisius Wilhelmina Hospital Nijmegen. Charges for laboratory tests were used as estimates for the costs of histological assessment of the biopsies and H.pylori serology (Table 1). Guidelines for cost analysis in health care research were used to determine the costs for consulting a GP and for a visit to the out-patient clinic. Retail prices formed the basis for calculating the costs of medication per day.
|
Sample size
The calculated sample size was based on data from Bytzer et al.10 According to results of this study, 20% of the patients reported complete symptom relief after prompt endoscopy. We estimated that an additional 15% of patients in the treat-and-test strategy would report symptom relief. We expected no difference in effectiveness between prompt endosocpy and radiography. To detect this difference, 270 patients overall needed to be studied. This sample size had a power of 80% to detect a true difference between the management strategies after 6 months, with a significance level of 5%. Taking into account loss to follow-up, the intention was to include 100 patients in each group. However, only 199 patients in total were included. Although we could not find any objective reasons for it, more and more GPs changed their management strategy to primary prescription of proton pump inhibitors during the study period. Therefore, we had to stop the study prematurely, because at the end of the study almost all prospective patients were already using proton pump inhibitors. A sample size of 199 patients has a power of 68% to detect a true difference at a 5% significance level. All analyses were performed using SAS software (SAS institute, Inc. Cary, NC).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
A total of 199 patients presenting with persistent dyspeptic symptoms of sufficient severity to justify diagnostic upper gastrointestinal endoscopy were randomly assigned to three different management strategies: 69 patients underwent empirical treatment followed by a test-and-eradicate strategy in the case of relapse; 66 patients underwent prompt endoscopy; and 64 patients underwent prompt radiography. Overall, seven patients were excluded from the analysis because no information about the primary outcome measures was available (Fig. 1). Sixty patients (90%) from the treat-and-test group, 58 from the endoscopy group (92%) and 52 from the radiography group (84%) completed the questionnaire at 6 months. The demographic characteristics of the randomized study groups were comparable (Table 2). Overall, 37% of the patients were H.pylori seropositive at baseline.
|
|
During the study period, 22 patients in the treat-and-test group underwent four upper gastrointestinal endoscopies and 22 radiographs (Table 3). Most patients from the treat-and-test group who underwent endoscopy or radiography had normal findings (82%), three patients had a peptic ulcer (14%) and two patients had reflux oesophagitis (9%) (Table 4). There was no difference between the groups in visits to GPs or specialists and, on average, a patient made three visits to one or more physicians, predominantly GPs. Overall, 7309 daily defined doses of medication were prescribed in the treat-and-test group, 8417 in the endoscopy group and 8651 in the radiography group. Seventy-five percent of all prescriptions in the treat-and-test group were proton pump inhibitors, on average 82 doses per patient. In the endoscopy group,
50% of the prescriptions were proton pump inhibitors, on average 68 doses per patient. In the radiography group, <50% of the prescriptions were proton pump inhibitors and 25% of the prescriptions were H2-receptor antagonists, on average 59 and 35 doses per patient, respectively. Overall daily defined doses of prescribed proton pump inhibitor and H2-receptor antagonist differed significantly between the management strategies.
|
|
At 6 months, 20 patients in the treat-and-test group, 22 in the endoscopy group and 18 in the radiography group reported use of medication, which was not a statistically significant difference (Table 5). During follow-up, nine patients in the treat-and-test group, 16 in the endoscopy group and seven in the radiography group underwent H.pylori treatment. Overall, 26 patients reported complete symptom relief at 1 month, 52 at 3 months and 48 at 6 months. The numbers of patients with complete symptom relief in the empirical group, endoscopy group and radiography group were 21, 16 and 15, respectively, at 3 months (P = 0.59), and 23, 13 and 12, respectively, at 6 months (P = 0.05). We did not find a significant difference in satisfaction between the three patient groups. We also could not find a significant difference in medical-related quality of life scores between the three treatment strategies. The average medical cost for the treat-and-test group was 276 euro [interquartile range (IQR) 149.0–337.12)], whereas for the endoscopy and radiography group the costs were 426 (IQR 229.8–504.7) and 321 euro (IQR 185.5–356.5), respectively (P < 0.05). The difference in medical costs between the three strategies was predominantly the result of the differences in the number of endoscopies and radiographs (P < 0.05) and H.pylori diagnostics (P < 0.05) performed.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Empirical treatment with a proton pump inhibitor followed by a test-and-eradicate strategy was more effective in comparison with prompt endoscopy or radiography in previously uninvestigated patients with persistent dyspeptic symptoms. We found a 16% difference in number of patients with complete symptom relief between the treat-and-test group (40%) in comparison with prompt endoscopy (24%) or radiography (24%), 6 months after randomization. Patients in the treat-and-test group were prescribed as much medication as in the other groups, but underwent fewer diagnostic procedures in comparison with prompt endoscopy or radiography followed by directed treatment. As a consequence of the difference in the number of diagnostic procedures between the treatment strategies, on average, patients in the empirical group cost less. As the costs of upper gastrointestinal endoscopy used in our analysis were very low compared with those reported in other publications, the difference in cost-effectiveness might be even greater when performing an economic evaluation in different clinical settings.7 At 6 months, equal numbers of patients of the three treatment strategies were still using medication. There was no difference in patient satisfaction between the management strategies. We did not find a difference in cost-effectiveness between prompt endoscopy and radiography.
Empirical treatment has been evaluated in several studies.3,6,10–13 The results from most studies showed no differences in outcome between empirical treatment and prompt endoscopy. There are several reasons why no differences between these management strategies were found. First of all, management of patients with dyspeptic symptoms is complex. In order to investigate management strategies, overall protocols need to be compared instead of a single aspect of one specific management strategy. For example, a single empirical proton pump inhibitor treatment is not sufficient to prevent symptomatic relapses in most patients with persistent symptoms. Secondly, in some studies, less effective empirical therapy has been used.10,11 In acid-related dyspepsia, an H2-receptor antagonist or prokinetic therapy is less effective in symptom relief than proton pump inhibitor therapy. Moreover, empirical treatment with H2-receptor antagonists or prokinetics will lead to more symptomatic relapses compared with proton pump inhibitors. Thirdly, the most obvious reason for the variance in results between our study and others was introducing the H.pylori test-and-eradicate strategy after the empirical proton pump inhibitor treatment in the management strategy. A substantial proportion of the patients will have a symptomatic relapse after empirical treatment because of H.pylori. Treatment of H.pylori infection will lead to symptom relief in patients with peptic ulcers and in some with functional dyspeptic symptoms.14
As the treat-and-test strategy is more cost-effective than prompt endoscopy or radiography, we need to reassess the value of upper gastrointestinal endoscopy or radiography for the evaluation of patients with persistent dyspeptic symptoms. The majority of patients seem to experience symptom relief as a result of proton pump inhibitor treatment, and there appears to be no necessity for directed treatment in most patients with persistent dyspeptic symptoms. Our findings showed that in 67% of all dyspeptic patients, including those with reflux oesophagitis, normal findings or irrelevant inflammations will be found. Therefore, diagnosis is not essential for most patients, with the exception of those with H.pylori infection-related disorders and those with a malignancy. In patients with H.pylori infection-related disorders, such as peptic ulcers or relevant inflammations, serious gastritis or duodentitis, the bacteria need to be eradicated to control dyspeptic symptoms and prevent relapses. Screening and treating H.pylori infection is achievable without upper gastrointestinal endoscopy. The only patients who could possibly benefit from prompt endoscopy or radiography are those with a malignancy or pre-maligant disorder. Selecting these patients is difficult; overall, only a small percentage will have a malignancy. Alarm symptoms are considered to be very useful predictors of malignancies in dyspeptic symptoms.15
To exclude serious pathology in patients with persistent dyspeptic symptoms, the GP has two options, upper gastrointestinal endoscopy or radiography. Some gastroenterologists consider radiography as a diagnostic test obsolete in comparison with upper gastrointestinal endoscopy.16 Indeed, upper gastrointestinal endoscopy is more sensitive than radiography to investigate the upper gastrointestinal tract. In spite of this, radiography is still used often by GPs to exclude serious pathology or to reassure patients, which might be at least as important. The procedure is considered less of a burden for the patient, especially in those with poor medical condition and older patients. On the other hand, if something suspicious is found by radiography, additional upper gastrointestinal endoscopy will often be necessary. In our study, this did not lead to substantial additional cost, as prompt radiography was at least as cost-effective as prompt endoscopy. During endoscopy, other procedures often are routinely performed such as histology and H.pylori bacteriology, making this procedure on average more expensive than radiography.
Empirical treatment followed by a H.pylori test-and-eradicate strategy increases appropriate use of endoscopy and radiography facilities.4,6 However, most patients who underwent endoscopy or radiography after empirical treatment had normal findings. The treat-and-test strategy did not result in a more accurate selection of patients with functional disorders. Our follow-up was only 6 months and perhaps not long enough to estimate the final numbers of patients who need a diagnostic examination. It is still unclear whether empirical treatment followed by a test-and-eradicate strategy reduces or merely delays the number of endoscopies over time. However, in earlier studies, we observed that most patients are managed within the first 6 months of the follow-up period.6
In conclusion, empirical treatment with a proton pump inhibitor followed by H.pylori test-and-treat strategy in patients with dyspeptic symptoms resulted in fewer diagnostic tests with more symptom relief and equal patient satisfaction compared with prompt upper gastrointestinal endoscopy or radiography in the management of patients with dyspepsia. Prompt radiography can be considered as a cost-effective alternative diagnostic approach to prompt upper gastrointestinal endoscopy for patients with uninvestigated persistent dyspeptic symptoms in primary care.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 Sobala GM, Crabtree JE, Pentith JA et al. Screening dyspepsia by serology to Helicobacter pylori. Lancet 1991; 338: 94–96.[CrossRef][Web of Science][Medline]
2 Patel P, Khulusi S, Mendall MA et al. Prospective screening of dyspeptic patients by Helicobacter pylori serology. Lancet 1995; 346: 1315–1318.[CrossRef][Web of Science][Medline]
3 Delaney BC, Wilson S, Roalfe A et al. Cost effectiveness of initial endoscopy for dyspepsia in patients over age 50 years: a randomised controlled trial in primary care. Lancet 2000; 356: 1965–1969.[CrossRef][Web of Science][Medline]
4 Lassen AT, Pedersen FM, Bytzer P, Schaffalitzky de Muckadell OB. Helicobacter pylori test and eradicate versus prompt endoscopy for management of dyspeptic patients: a randomised trial. Lancet 2000; 356: 455–460.[CrossRef][Web of Science][Medline]
5 Laheij RJF, Verbeek ALM, Severens JL, van de Lisdonk E, Jansen JBMJ. Management in general practice of patients with persistent dyspepsia. J Clin Gastroenterol 1997; 25: 563–567.[CrossRef][Web of Science][Medline]
6 Laheij RJF, Verbeek ALM, Severens JL, van de Lisdonk E, Jansen JBMJ. Randomised controlled trial of omeprazole or endoscopy in patients with persistent dyspepsia: a cost-effectiveness analysis. Aliment Pharmacol Ther 1998; 12: 1249–1256.[CrossRef][Web of Science][Medline]
7 Laheij RJF, Severens JL. Cost of endoscopy in economic evaluation. Gastroenterology 1997; 113: 2023–2024.[Medline]
8 Nandurkar S, Locke G III, Van Dyke C, Murray J. Reflux disease in the community: a study of healthcare utilization. Gastroenterology 2002; 122(Suppl): P115.
9 Laheij RJF, Verbeek ALM, Straatman H, Jansen JBMJ. Evaluation of commercially available Helicobacter pylori serology kits: a review. J Clin Microbiol 1998; 36: 2803–2809.
10 Bytzer P, Hansen JM, Schaffalitzky de Muckadell OB. Empirical H2-blocker therapy or prompt endoscopy in management of dyspepsia. Lancet 1994; 343: 811–816.[CrossRef][Web of Science][Medline]
11 Duggan A, Elliot C, Tolley K et al. Randomised controlled trial of four dyspepsia management strategies in primary care with 12 months follow-up. Gastroenterology 2000; 118: A438.
12 Brignoli R, Watkins P, Halter F. The omega-project–a comparison of two diagnostic strategies for risk- and cost-oriented management of dyspepsia. Eur J Gastro Hepatol 1997; 9: 337–343.[Medline]
13 Delaney BC, Innes MA, Deeks J et al. Initial management strategies for dyspepsia (Cochrane Review). The Cochrane Libary 2000; issue 2. Oxford: Update Software Ltd.
14 Laheij RJF, van Rossum LGM, Verbeek ALM, Jansen JBMJ. Helicobacter pylori infection treatment of non-ulcer dyspepsia, an analysis of meta-analyses. J Clin Gastroenterol 2003; 36: 315–320.[Medline]
15 Numans ME, van der Graaf Y, de Wit N, de Melker RA. How useful is selection by using alarm symptoms in requesting gastroscopy? An evaluation of diagnostic determinants for gastro-oesophageal malignancy. Gastroenterology 2000; 118: A266.
16 AGA Technical Review: evaluation of dyspepsia. Gastroenterology 1998; 114: 582–595.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  A. K. Musana, S. H. Yale, and K. A. Lang Managing Dyspepsia in a Primary Care Setting Clin. Med. Res., December 1, 2006; 4(4): 337 - 342. [Full Text] [PDF]
|
|
|
|
 |
|
 Other articles noted Evid. Based Med., November 1, 2004; 9(6): 191 - 192. [Full Text] [PDF]
|
|
|
|
 |
|
 23 Apr 2004 to 23 Jul 2004 Evid. Based Nurs., October 1, 2004; 7(4): e4 - e4. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||