Cost-Utility Analysis of Herpes Zoster Vaccine in Thailand

A808

VA L U E I N H E A LT H 1 9 ( 2 0 1 6 ) A 8 0 7 – A 9 1 8

Cost Utility And Qaly Studies CU1 Cost-Utility Analysis of Herpes Zoster Vaccine in Thailand Taychakhoonavudh S1, Suranant N1, Issarasenee K1, Laoprasertsuk C1, Jiamton S2, Chanyachailert P2, Leeyaphan C2 1Faculty of Pharmaceutical Sciences, Chulalongkorn University, Bangkok, Thailand, 2Department of Dermatology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand

Herpes zoster (HZ), a skin disease caused by varicella zoster virus is a health problem worldwide, including Thailand. HZ vaccine has been available in the market since 2006 and proved to significantly reduce the incidence of HZ. Previous researches in other countries have suggested that using HZ vaccine is cost-effective in an elderly over 60 years old. However, no research has been done on the economic evaluation of HZ vaccine in Thailand.Objectives: The objective of this study was to conduct a cost-utility analysis of HZ vaccine in Thailand in 60 years and older population.  Methods: This study was conducted as an economic evaluation, employing decision analytic model technique (Markov model). The cost and outcome of HZ vaccine comparing with no vaccine in an elderly population aged 60 years and older from the payer perspective was estimated. Costs were composed of cost of HZ treatment, cost of PHN treatment, cost of HZ vaccine, and cost of vaccine administration. Outcomes were measured in quality-adjusted life years (QALYs) obtained from published literatures. Both costs and outcomes were discounted at 3% annual rate. Cost-effectiveness was measured using incremental cost-effectiveness ratio (ICER).  Results: Use of the HZ vaccine increased costs and QALYs per person by US$142 (4,951.23 Thai Baht) and 0.0035 QALYs, respectively. The incremental costeffectiveness ratio (ICER) for HZ vaccine comparing with no vaccine was US$40,972 (1,427,345.86 Thai Baht) per QALY. Comparing to Thailand’s Willingness to pay, HZ vaccine for elderly aged 60 or more did not seem to be cost-effective. Results from sensitivity analysis revealed that if the HZ incidence increased or the vaccine price was cheaper, the use of HZ vaccine would be cost-effective.  Conclusions: Our analysis demonstrates that the use of HZ vaccine in elderly aged 60 years and older does not seem to be cost-effective when compared to Thailand’s Willingness to pay. CU2 Forecasting Lifetime Health Outcomes and Costs of Treatment for Non-Alcoholic Fatty Liver Disease Chongmelaxme B1, Phisalprapa P2, Sawangjit R3, Dilokthornsakul P1, Chaiyakunapruk N4 University, Phitsanulok, Thailand, 2Mahidol University, Bangkok, Thailand, 3Mahasarakham University, Muang, Thailand, 4Monash University Malaysia, Selangor, Malaysia

gression of MS, allowing us to calculate the impact of MS on life expectancy, QALYs and total lifetime costs in Australia. LE and QALYs were substantially decreased in comparison to general population norms due to relapses and progressive disability associated with MS, with correspondingly huge lifetime costs. This model will form the basis for future cost-effectiveness analyses of interventions that reduce relapses and progression of disability resulting from MS. CU4 Development of An Alcohol Policy Model That Predicts Life Years, Qalys, and Health Care Costs Accounting for Alcohol use Disorder Identification Test Lewsey J, Leelahavarong P, Briggs A University of Glasgow, Glasgow, UK

Objectives: To develop an alcohol policy model that predicts life years (LYs), quality adjusted life years (QALYs) and healthcare costs using the Alcohol Use Disorder Identification Test (AUDIT) screening tool and other risk factors. Further, to transfer the developed model into a Thai setting.  Methods: The Scottish Health Surveys from 1995-2012 were linked to Scottish morbidity records and death records for the period 1981-2013. Parametric survival analysis was used to estimate the hazard risks of first alcohol-related and non-alcohol related hospitalisation and death. For men and women separately, multivariable regression analyses were applied for modelling utility score, risks of subsequent hospitalisation and annual health care costs within the follow-up period. The risk profiles were modelled using the covariates of age, socio-economic status, health condition, alcohol drinking, smoking, body mass index and physical activity. A health-state transition model with annual cycle length was developed using outputs from analyses to predict LYs, QALYs and lifetime health care costs.  Results: The sample size of the cohort was 46,230. The modelling showed that increasing AUDIT score and cigarettes per day were associated with an increased risk of first alcohol-attributable hospitalisation. Predicted outcomes for a male aged 30 years with high risk drinking levels (AUDIT >  7) were worse than for low risk drinking (AUDIT ≤  7), with approximately 5 and 7 years LY and QALY lost, respectively. The corresponding results for females show for high risk drinking (AUDIT >  4) compared to low risk drinking (AUDIT ≤  4), approximately 10 and 12 years LY and QALY lost, respectively.  Conclusions: The developed policy model framework could be used for further economic evaluation of alcohol interventions and other health behaviour interventions, and the model can be transferred to the Thai and other country settings.

1Naresuan

Objectives: Prevalence of non-alcoholic fatty liver disease (NAFLD) has been increasing over the last decades due to lifestyle change and industrialization. However, the knowledge of long-term health outcomes and costs for NAFLD is limited. This study aims to forecast lifetime health outcomes and costs of treatment for NAFLD.  Methods: A Markov model was developed to mimic NAFLD disease progression. A total of ten treatment options (metformin, metadoxine, pentoxifylline, pioglitazone, ursodeoxycholic acid, vitamin E, weight reduction program, pioglitazone/losartan, pioglitazone/metformin, and vitamin E/vitamin C) and no treatment were included in this study. The number of cases prevented from advanced fibrosis, cirrhosis, hepatocellular carcinoma (HCC), and life expectancy were forecasted. Lifetime cost in US dollar unit of each treatment option was also estimated. Probabilistic sensitivity analysis was performed to determine robustness of our findings and presented as 95% credible interval.  Results: Pentoxifylline resulted in the highest health benefit. It could increase the percentages of cases prevented from advanced fibrosis [23.78% (95% CrI: 6.45% to 41.26%)], cirrhosis [3.45% (95% CrI: -0.88% to 6.39%)], HCC [0.56% (95% CrI: -0.04% to 1.07%)], and it increases life expectancy [0.09 years (95% CrI: 0.03 to 0.15)] compared to no treatment. The average additional lifetime cost of pentoxifylline was $866.37 (95% CrI: $508.52 to $1,175.16). Weight reduction program and pioglitazone/metformin were the second and third best options, respectively. Health benefit and cost of ursodeoxycholic acid were worse than those of no treatment.  Conclusions: Pentoxifylline shown the highest health benefit for NAFLD treatment. However, the average lifetime cost was moderately high. Our findings could be an important evidence supported, especially for clinicians and policy makers, for making decisions on NAFLD treatment and considering on coverage and reimbursement of treatment options. CU3 Life Expectancy, Quality-Adjusted Life Years, and Total Lifetime Costs for Australian People with Multiple Sclerosis Palmer AJ, Taylor B, Van der Mei I, Si L, Ahmad H University of Tasmania, Hobart, Australia

Objectives: Multiple Sclerosis (MS), a long term neurodegenerative disease that has a substantial effect on morbidity and mortality, as well as being a major economic burden to patients and society. Our aim was to quantify life expectancy (LE), quality adjusted life years (QALYs) and total lifetime costs for a typical cohort of Australian people with MS.  Methods: A 4-state Markov model simulated progression from mild disability to moderate and severe disability and death for a cohort of 35 year old women starting with mild disability over a lifetime horizon. Risks of death were calculated from Australian life tables, adjusted by increasing disability severity. State-dependent relapse probabilities and their associated disutilities were accounted for. The probabilities of MS progression and relapses were estimated from Auslong and TASMSL MS epidemiological databases. Direct and indirect societal costs (2016 Australian dollars (AUD)) and health state utility values for each state were derived from the Australian MS Longitudinal Study. Probabilistic sensitivity analysis sampling from key input parameter distributions was performed. Costs were discounted at 5% annually.  Results: Mean (standard deviation) LE from age 35 years was 43.5 (0.3) years. This was 6.7 years less than for the Australian general population. QALYs were 29.5 (0.6) years, a loss of 12.3 years compared with the general population. Discounted total lifetime costs were AUD 940,434 (31,952) per person.  Conclusions: We have developed a health economics model of the pro-

Health Care Expenditure and Pricing Studies HC1 Impact of Adoption of A New Drug on the Trends in Utilization Patterns Changes and Pharmaceutical Expenditure- An Example of Urate-Lowering Therapy Peng Y1, Hsu C2 Medical University, Kaohsiung, Taiwan, 2Kaohsiung Chang Gung Memorial Hospital, Kaohsiung, Taiwan

1Kaohsiung

Objectives: The introduction of new technology in the health care market affects therapeutic choice and pharmaceutical expenditures at both hospital and healthcare system levels. This research aims to assess the trends in utilization and expenditures of urate-lowering agents (ULAs) and to determine how a new drug, febuxostat adoption impacts the prescribing patterns in a large medical center in Taiwan.  Methods: A cross-sectional study was conducted to evaluate patients with ULAs from 2011 through 2015 in inpatient and outpatient settings. ULAs, including allopurinol (A), benzbromazone (B), sulfinpyrazone (S), and febuxostat (F, available in March, 2013; cost 16-fold of A and B, 6-fold of S) were assessed with prescription volume and cost. We compared prescription cost with calendar month unit and prescribing patterns between incident and prevalent users for determining the contribution of diffusion of febuxostat. Descriptive analysis and Cochran– Armitage tests were performed to examine monthly changes in choice of ULAs over the study period.  Results: A total 31,994 patients ever treated with one of ULAs. The average monthly pharmaceutical cost increased 88% compared to the time prior to febuxostat introduction (1/2011-2/2013 vs 3/2013-12/2015= US$ 17,941 vs US$ 33,854), and continued to grow with average increasing rate of 3.2% (±10.3%) (Trend test, P< .0001). Of incident users, initiation rate revealed an increasing trend in B and F users, but a decreasing trend in A and S users. Among the prevalent users, the most common switching patterns were: A switched to F or B (29.74%), B or S to F (15.8%), S to B (8.7%) and B to A (5.8%).  Conclusions: ULAs expenditure has continued grown after diffusion of febuxostat in the study setting. Although febuxostat is a safer alternative to allopurinol but more costly than existing ULAs, further comparative effectiveness and cost-effectiveness analysis of febuxostat are imperative to guide appropriate health resources allocation. HC2 Multimorbidity and Health Care Service Utilization in the Australian Workforce: Findings From the National Health Survey Wang L1, Palmer AJ1, Cocker F2, Sanderson K1 Institute for Medical Research, Hobart, Australia, 2Monash Centre for Occupational and Environmental Health (MonCOEH), Melbourne, Australia

1Menzies

Objectives: To understand the patterns of health care utilization in employees with multimorbidity (more than one chronic condition in one individual), we: (1) characterized diseases in terms of comorbidity status (alone and coexisting with other chronic diseases); and (2) determined the associations between multimorbidity and disease-specific healthcare utilization (HSU) among these employees.  Methods: Data were derived from the cross-sectional 2011-12 Australian National Health Survey (NHS) which included the self-reported data from 10363 employees aged 15 years and over. HSU in the 12-months prior to the face-to-face NHS