Threshold analysis of Helicobacter pylori therapy.

OBJECTIVE

Conflicting recommendations have been made on whom to treat with antibacterials to eradicate Helicobacter pylori. The present analysis aims to explain the basis of such discrepancies.

DESIGN AND SETTING

The decision in favour of or against antibacterial therapy in patients with upper gastrointestinal symptoms is modelled as a threshold analysis. The threshold is the lowest probability for a given diagnosis at which the decision in favour of antibacterial therapy yields a higher expected outcome than the decision against it. A strong indication for antibacterial therapy is shown by a low threshold value.

MAIN OUTCOME MEASURES AND RESULTS

In patients with suspected peptic ulcer disease, both the high success rate of antibacterial therapy and its low cost make it the most favourable treatment option, its threshold being less than 35%. In ulcer patients receiving nonsteroidal anti-inflammatory drugs, the threshold is 73% if based on success rates of different treatment modalities, but only 7% if treatment costs are considered. The relatively poor success rate of antibacterial therapy in non-ulcer dyspepsia raises the diagnostic threshold for antibacterial therapy to 76% if based on therapeutic success rates. The small marginal cost of antibacterial therapy lowers the diagnostic threshold to 16%.

CONCLUSION

If therapeutic success is the primary concern, patients with vague abdominal symptoms should not be given antibacterial therapy, unless H. pylori has been established as a probable cause of their symptoms. If healthcare costs are the driving force for choosing one type of therapy over another, a trial of antibacterials appears indicated irrespective of any firm validation of H. pylori playing a role in the patient's disease.