Display Settings:


Send to:

Choose Destination
See comment in PubMed Commons below
Pharmacoeconomics. 1996 Jul;10(1):79-92.

Decision analysis of Helicobacter pylori eradication therapy using omeprazole with either clarithromycin or amoxicillin.

Author information

  • 1Pharmacoeconomics Research Centre, Universitie of Dundee, Scotland.


In patients with duodenal ulcer, omeprazole plus clarithromycin (OC) has achieved Helicobacter pylori eradication rates of about 80%, compared with 50% for omeprazole plus amoxicillin (OA). The drug acquisition costs for OC are 102.92 pounds sterling (pounds) compared with 38.96 pounds for OA using generic amoxicillin and 51.63 pounds using the proprietary brand 'Amoxil' (costs for 2-week regimens in 1995). The aim of this analysis was to estimate the total healthcare costs to the general practitioner (GP) of eradication therapy using a simple generalised model. Data about current practice in the UK were obtained from 502 respondents in a survey of hospital specialists and GPs. It was assumed that patients would derive no benefit from eradication therapy unless they had a duodenal ulcer, and that all OA patients received generic amoxicillin. The survey confirmed that OA was the commonest eradication therapy prescribed by UK GPs at that time. Three distinct patient groups were identified: patients with proven duodenal ulcer who were already receiving maintenance treatment with a histamine H2 receptor antagonist, and new patients with dyspepsia who were subdivided into those aged above or below 45 years. Patients receiving maintenance treatment for a duodenal ulcer would be prescribed eradication therapy by their GP without further endoscopy. If dyspepsia recurred after eradication therapy, they would be referred to a gastroenterologist, who would perform an endoscopy to confirm the recurrence of ulceration. In this model, the expected total healthcare costs (i.e. the costs of drug acquisition and subsequent treatment when required) following prescription of eradication therapy were lower for OC (157 pounds) than for OA (173 pounds). New patients aged over 45 years would be referred for endoscopy because of the risk that dyspepsia might be the initial presentation of gastric cancer. If duodenal ulceration was found, eradication therapy would be prescribed and, if dyspepsia remained or recurred, the patient would be referred back to the gastroenterologist. In this case, it was considered unlikely that a further endoscopy would be performed. Thus, the healthcare costs associated with failure of eradication in these patients were less than for patients on maintenance treatment, and the expected total healthcare costs were higher for OC (349 pounds) than for OA (335 pounds). Finally, a new patient aged under 45 years with dyspepsia would have eradication therapy prescribed on the basis of a clinical diagnosis of duodenal ulcer plus serological evidence of infection with H. pylori. Continuation or recurrence of dyspepsia would result in referral to a gastroenterologist, who would perform an endoscopy. The total expected healthcare costs were higher for OC (253 pounds) than for OA (251 pounds). The cost effectiveness of OA was sensitive to changes in the default costs (i.e. the average costs from the survey used in the decision analysis), particularly in patients < 45 years old. In these patients, OC would become the cheaper option if amoxicillin were prescribed by brand name instead of in generic form. In this patient group, the outcome was crucially dependent on the accuracy of the clinical diagnosis of duodenal ulcer; if this was at least 60%, then OC would be the cheaper regimen. Overall, the model clearly shows that the higher drug cost of OC is likely to be substantially offset by savings in other healthcare costs. If the direct healthcare costs of OC are higher than OA, then the decision maker must consider the indirect and intangible costs associated with failure of eradication therapy.

[PubMed - indexed for MEDLINE]
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Loading ...
    Write to the Help Desk