10, 21-23 Rifaximin, which also has been used as a therapy for MHE,24-26 has a much better adherence profile. Moreover, rifaximin has been shown to improve driving simulator performance in a placebo-controlled randomized trial.25 We report here the results of a model-based cost-effectiveness of MHE diagnosis and subsequent pharmaceutical treatment (lactulose or rifaximin) to reduce MVAs among cirrhosis patients. The analyses compared four potential strategies for diagnosing and treating MHE with a no-treatment alternative. Because the effectiveness learn more of pharmaceutical treatments with respect to reducing accidents among treated patients has not

been well established, we conducted extensive sensitivity analyses around this key parameter. The aim was to provide a cost-effectiveness platform for MHE diagnosis and treatment from a societal perspective and tailored to individual treatment options available in the U.S. ICT, inhibitory control test; MHE, minimal hepatic encephalopathy; MVA, motor vehicle accident; NPE, neuropsychological exam; OHE, overt hepatic encephalopathy; PHES, Psychometric Hepatic Encephalopathy Score; QOL, quality of life; SPT, standard psychometric

test battery. The cost-effectiveness analysis combined a Markov model of Epacadostat progression from cirrhosis without MHE, to MHE, to OHE, with empirically derived and literature-based estimates of MHE diagnostic tests and treatment parameters and MVA-related parameters. The analysis adopted a societal perspective and included time costs borne by patients, as well as the societal costs associated with MVAs. All future costs and benefits were discounted at a 3% annual rate (0% and 5% in the selleck chemicals sensitivity analyses) in accordance with recommended practice.27 The results are expressed in base-year 2010 dollars. The Markov model followed a simulated cohort of 1,000 cirrhosis patients with compensated liver disease and without OHE, from

entry into treatment (at which point they might or might not have MHE), through the potential development of MHE, and later OHE, at which time they exited the modeled cohort. The model assumed that cirrhosis patients were screened for MHE on a semiannual basis.12, 13 State changes within the Markov model also occurred at 6-month intervals. Annual state-transition probabilities, from non-MHE cirrhosis to MHE, and from MHE to OHE, were derived from a published study.28 Six-month state-transition values were derived from these annual probabilities using the equation: p(6 mo) = 1 − (1 − p(12 mo))0.5. The baseline prevalence of MHE was set to 55%.2-5, 15, 21, 24, 29 The simulated cohort of cirrhosis patients was followed for a total of 5 years.