Helicobacter pylori eradication in clinical practice: retreatment rates and costs of competing regimens

OBJECTIVE: To measure outcomes for eradication regimens for Helicobacter pylori infection in routine clinical practice. DESIGN: Retrospective analysis of an integrated medical and pharmacy claims database identified patients treated from June 1, 1995 through May 31, 1996, and followed the patients' claims until December 31, 1996. SETTING: The database represented multiple health plans throughout the US. PATIENTS: Patients were > or = 16 years old, continuously enrolled from April 1, 1995, to December 31, 1996, and met clinical (diagnostic or procedural) criteria. INTERVENTION: Patient cohorts were treated with bismuth-based triple (n = 98), proton-pump inhibitor (PPI)-based triple (n = 180), or PPI-based dual (n = 337) regimens. OUTCOME MEASURES: Retreatment; monthly postregimen medical expense, controlling for preregimen expense; and drug cost per successfully treated patient. Cox regression (retreatment analysis) and ANCOVA (postregimen expense analysis) adjusted for age, gender, diagnostic/procedural criteria met by patient, and specialty physician use. RESULTS: Retreatment rates were higher (p < 0.05) for PPI-based dual than bismuth-based or PPI-based triple-therapy cohorts. Retreatment rates for bismuth- and PPI-based triple-therapy cohorts were not significantly different. Total and follow-up (excluding the first 2 wk of treatment) expenses were higher for retreated than nonretreated patients (p < 0.01). Total expenses were higher for the PPI-dual cohort (p < 0.05) than for triple cohorts. Drug costs per successfully treated patient were $30 for bismuth-based, $172 for PPI-based triple, and $208 for PPI-based dual-therapy regimens. CONCLUSIONS: PPI-based dual-therapy regimens are not cost-effective in H. pylori treatment. Further study should compare more costly (PPI-based) versus less costly (bismuth-based) triple-therapy regimens.