Objective: To compare healthcare costs between clopidogrel and prasugrel over 30-day and 365-day periods after discharge from the hospital or emergency room (ER) in patients treated with prasugrel who were hospitalized or had an ER visit for an acute coronary syndrome (ACS) event.
Methods: This retrospective observational study was based on claims from January 2009-July 2012 in the Truven Health Analytics MarketScan database. Clopidogrel patients were propensity-score matched 1:1 to prasugrel-treated patients. Lin's frequentist cost history method for censored data and Bayesian zero-inflated gamma regression models were used to analyze healthcare costs.
Results: The clopidogrel/prasugrel matched-cohort included 10,963 well-matched pairs of patients. Lin's frequentist analysis showed that outpatient visit costs were significantly lower for clopidogrel than prasugrel after 30 days of follow-up. At 30 days, Bayesian data analysis showed strong evidence that clopidogrel was superior to prasugrel for all-cause and ACS-related hospitalization costs and showed very strong evidence that clopidogrel was superior to prasugrel for all-cause and ACS-related outpatient visit costs. At 365 days, Bayesian data analysis showed strong evidence that clopidogrel was superior to prasugrel for all-cause outpatient visit costs and very strong evidence that clopidogrel was superior to prasugrel for ACS-related outpatient visit costs. Point estimates of the all-cause and ACS-related ER visit costs at 30 days and 365 days were similar, but statistical results were inconclusive because of the large variability in this outcome variable.
Conclusion: Based on retrospective observational data in a real-world setting, all-cause and ACS-related hospitalization and outpatient visit costs were lower for clopidogrel than prasugrel over 30 days after discharge from a hospitalization or ER visit associated with ACS in patients treated with prasugrel. At 365 days the difference in all-cause and ACS-related outpatient costs remained, but there was little evidence of a difference in either all-cause or ACS-related hospitalization costs.
Keywords: Acute coronary syndrome; Bayes factor; Clopidogrel; Healthcare economics; Healthcare utilization; Hospitalization; Prasugrel.