Efficient Identification and Evaluation of Effective Helicobacter pylori Therapies

Annually hundreds if not thousands of patients fail empiric H pylori eradication therapy. Failure occurs in part because routine post treatment testing, which provides an early warning of increasing antibiotic resistance, is not universally done and physicians are generally unaware that cure rates with legacy triple therapy (a proton-pump inhibitor, amoxicillin, and clarithromycin) in most places has fallen below 80%. We propose first, institution of routine post eradication testing and second, abandonment of the “better than another therapy” approach to separating acceptable from unacceptable therapies. Instead, we propose using results-based outcomes (ie, >95% cure rates). H pylori should be evaluated as other infectious diseases (ie, few would compare a new antibiotic for pneumonia with the previous best choice whose effectiveness was now impaired because of resistance). Randomized comparisons should be restricted to studies designed to improve (eg, simplify or reduce costs) high cure rate therapies while maintaining efficacy. We also discuss potential ethical issues such as those including known or suspected low cure rate therapies. Legacy therapies cannot be identified as “approved” or “recommended” even if both statements were true. Instead patients and ethics boards must receive “full disclosures” both before and during studies that include all that might affect a patient’s decision to enter or to continue. Finally, we provides advice regarding trial design using “best shot” pilot studies to efficiently identify tentative effective regimens which are then confirmed in randomized trials all the while minimizing patient risks and drug exposure.

Abbreviations used in this paper: CI, confidence interval, PPI, proton pump inhibitor, RCT, randomized controlled trial