PHS49 Cost of Hypoglycaemia in Patients With Diabetes in Poland

A526

VALUE IN HEALTH 15 (2012) A277–A575

Japan. In this study, we conducted an economic evaluation of a universal infant hepatitis B vaccination compared with the current selective strategy of vaccinating high-risk infants. METHODS: A cost-effectiveness analysis was conducted using a Markov model. The state transition process was defined as a series of 9 states including susceptible, immune, acute hepatitis, fulminant hepatitis, asymptomatic carrier, chronic hepatitis, liver cirrhosis, hepatocellular carcinoma and death. All data on cost, epidemiology and utility were derived from official statistics, published literatures and expert opinion. The incremental cost-effectiveness ratio (ICER) per quality-adjusted life year gained and life-year gained were calculated at a 3% discount rate. The analysis was performed from payer’s perspective and time horizon was set to 100 years. One-way sensitivity analysis was undertaken. All parameters were projected to the population of 2009 birth cohort, 1.078 million. RESULTS: By introducing the universal hepatitis B vaccination, 1,507 patients and 197 deaths caused by HBV related liver diseases were expected to be prevented. Cost of medical care was saved 770 million yen, but the cost of vaccination was 1.77 billion yen. Estimated ICERs were 18,300,515 yen/QALY and 20,093,520 yen/LYG. When the cost of vaccination was 5,600 yen(1,868 yen per dose), ICER was 5,000,000 yen/QALY. CONCLUSIONS: At a cost of 5,600 yen(1,868 yen per dose) of vaccination, universal hepatitis B policy was considered to be cost- effective in Japan. In hepatitis B, regional differences of genotype were well-known and affected the sequel of hepatitis B related diseases. Because most of the epidemiological data employed in this study were derived from foreign literatures, we try to replace then into estimated local data. This study was partially supported by the grant of Japanese Ministry of Health, Labor and Welfare (Research on Applying Health Technology). PHS45 COST EFFECTIVENESS ANALYSIS OF A SPECIALIST SERVICE AND ADRENALINE INJECTORS IN ANAPHYLAXIS Armstrong N, van Mastrigt G, Wolff R, Kleijnen J Kleijnen Systematic Reviews Ltd., York, UK

OBJECTIVES: Anaphylaxis is a severe life threatening acute event that can have a number of triggers. Specialist services are believed to be important in preventing recurrence and current lack of referral might mean high recurrence rates and unnecessary cost. Also, lack of timely and correct use of adrenaline injectors might lead to significant excess mortality. The study objective was therefore to assess the cost effectiveness of specialist service versus standard care with or without prescription of adrenaline injectors. METHODS: A Markov model validated by clinical experts was constructed, which modeled anaphylaxis according to trigger, either food, drug, insect or idiopathic. Anaphylaxis mortality was modeled as a function of time to die and time for emergency response. Probabilistic sensitivity analysis on key parameters was performed. RESULTS: Standard care with injectors was dominated by specialist service with or without injectors. Specialist service with no injectors would be cost effective if the threshold for a Quality Adjusted Life Year was greater than about £740 and with injectors would be cost effective if the threshold was greater than £1800. These results were robust to all sensitivity analyses except at relatively extreme values of a small number of parameters. CONCLUSIONS: This is the first study to address the cost effectiveness of specialist service or adrenaline injectors in anaphylaxis. The results showed that specialist service with adrenaline injectors was cost effective at a threshold of £20,000 per Quality Adjusted Life Year. More well designed prospective studies on the effectiveness of specialist services are needed to confirm these findings. PHS46 COST-EFFECTIVENESS OF A SECONDARY PREVENTION PROGRAM IN PATIENTS WITH MYOCARDIAL INFARCTION: RESULTS FROM A RANDOMISED CONTROLLED TRIAL (PROACTIVE HEART) Turkstra E1, Hawkes AL2, Oldenburg BF3, Scuffham PA1 1 Griffith University, Meadowbrook, Queensland, Australia, 2Cancer Council Queensland, Fortitude Valley, Queensland, Australia, 3Monash University, Melbourne, Victoria, Australia

OBJECTIVES: Participation in coronary heart disease (CHD) secondary prevention programs is low, therefore an innovative program is needed to meet this treatment gap. As a secondary aim within a large trial, the current study evaluated costeffectiveness of a telephone-delivered secondary prevention program for myocardial infarction patients compared to usual care. METHODS: A total of 430 adult myocardial infarction patients were randomised to a six-month CHD secondary prevention ‘health coaching’ intervention or usual care condition. Primary outcome variables were health-related quality of life (SF-36) and physical activity (Active Australia Survey). Data were collected at baseline, 6 months (post-intervention) and 12 months (6 months post-intervention). A secondary cost-effectiveness analysis was conducted. Health utility (SF-6D) and health care utilisation data were collected using self-reported (GP, specialist, other health professionals, health services, and medication) and claims data (hospitalisation rates). Multiple imputation techniques were applied to adjust for missing data. Intervention effects are presented as mean difference (95% CI), p value. RESULTS: Improvements in health status (SF-6D) were observed in both groups, with no significant difference between the groups at 6 [0.012 (-0.016, 0.041), p⫽0.372] or 12 months [0.011 (-0.028, 0.051) p⫽0.738]. Patients in the health coaching group were significantly more often admitted to hospital due to causes not related to cardiovascular disease (p⫽0.042). The overall cost for the health coaching group was higher ($10,574 vs. $8,534, p⫽0.021), mainly due to higher hospitalisation costs ($6,841 vs. $4,984, p⫽0.036). The incremental cost-effectiveness ratio was $85,423 per QALY. CONCLUSIONS: ProActive Heart is not a cost-effective intervention compared to usual care. There was no intervention effect on SF-6D at 6 or 12 months and it resulted in significantly increased costs. This higher cost may in the future eventuate in cost savings,

as patients are better monitored and health problems may be identified at an earlier stage resulting in better health outcomes. PHS47 COST-UTILITY ANALYSIS OF ROTIGOTINE IN RESTLESS LEGS SYNDROME FROM THE NHS SCOTLAND PERSPECTIVE Benhaddi H1, Lundie G2 1 UCB Pharma, Brussels, Belgium, 2UBC Pharma, Slough, UK

OBJECTIVES: To assess the cost-effectiveness of rotigotine transdermal patch relative to placebo and oral dopamine agonists (ropinirole and pramipexole) in the treatment of moderate-to-severe idiopathic restless legs syndrome (RLS) from the perspective of National Health Service (NHS) Scotland. METHODS: A cost-utility analysis was conducted using a decision tree in order to measure the cost per quality-adjusted life-year (QALY) gained resulting from the treatment of moderateto-severe idiopathic RLS in 2008. Clinical, safety and quality of life data were extracted from the literature. The decision analytic model was run for a 1-year time horizon in the base-case analysis. The model included direct medical costs, augmentation rate and the most common adverse events. Sensitivity analyses assessed the effect of varying the time horizon to 9 months, 2 and 5 years. An annual discount rate of 3.5% was applied to costs and benefits for analysis beyond 1 year. RESULTS: Over a 1-year period, treating patients with rotigotine would result in a total cost of £1658 per patient as compared with £606, £1152 and £1055 for placebo, ropinirole and pramipexole respectively. Treatment with rotigotine resulted in 0.842 QALYs gained compared with 0.691 for placebo, 0.753 for ropinirole and 0.780 for pramipexole. The cost per QALY gained for rotigotine was £6990 versus placebo, £5725 versus ropinirole and £9729 versus pramipexole. Probabilistic sensitivity analysis has shown that the probability of rotigotine being cost-effective compared with placebo, ropinirole and pramipexole is greater than 0.90 at a willingness to pay threshold of £13,000. At a threshold of £20,000, the probability of rotigotine being more cost-effective is close to 1 versus placebo and ropinirole and 0.93 versus pramipexole. CONCLUSIONS: Rotigotine transdermal patch was shown to be a cost-effective treatment in patients with moderate-to-severe restless legs syndrome from the payer’s perspective (NHS Scotland). PHS48 COST-EFFECTIVENESS ANALYSIS OF TOLVAPTAN FOR HYPONATREAMIA IN SOUTH KOREA Kang HJ1, Choi J1, Lee MY2, Lee EK1 1 Sungkyunkwan University, Suwon, South Korea, 2Sungkyunkwan University, Suwon, Gyeonggido, South Korea

OBJECTIVES: To evaluate the cost-effectiveness of tolvaptan for hypervolemic or euvolemic hyponatremia in South Korea. METHODS: A decision tree was constructed to assess the clinical and economic impact over 30 days from restricted societal perspective. A competitor was supportive care. The result was presented as the incremental cost per QALY gained. We supposed that patients would move to 3 different serum sodium levels; normonatremia (⬎135mEq/L), mild hyponatremia (130-135mEq/L), marked hyponatremia(⬍130 mEq/L) after 4 days treatment. Each level had three states; discharge, continuing hospitalization and death. According to serum sodium level, Patients had difference their length of hospital stay. Uncertainty was explored with deterministic sensitivity analysis. RESULTS: The analysis showed that cost of tolvaptan was 1,358,370 KRW and supportive care was 1,396,092 KRW. The use of tolvaptan reduced 37,722 KRW for supportive care. QALYs were 0.047203, 0.042146 (tolvaptan, supportive care, respectively). Tolvaptan had 0.00506 QALYs higher. Cost-effectiveness analysis represented tolvaptan was dominant. The deterministic sensitivity analysis for uncertain parameters demonstrated that this analysis results were robust. CONCLUSIONS: Tolvaptan was cost-effective for hyponatremia in comparison with supportive care. The first medicine as oral vasopressin receptor antagonist, tolvaptan could increase patients’ quality of life for hyponatremia in Korean population. PHS49 COST OF HYPOGLYCAEMIA IN PATIENTS WITH DIABETES IN POLAND Widz K1, Czech M2, Hermanowski T3 1 Medical University of Warsaw, Warsaw, Poland, 2Novo Nordisk Pharma Sp z.o.o., Warsaw, Poland, 3Department of Pharmacoeconomics, Medical University of Warsaw, Warsaw, Poland

OBJECTIVES: Estimation of indirect and selected direct costs of hypoglycaemia in patients with type 1 and type 2 diabetes in Poland. METHODS: The study was conducted at 4 Polish diabetes centres- 2 urban and 2 suburban. Anonymous questionnaire comprising 35 questions was used in direct interviews. Data were analysed in a population of 180 patients with diagnosed diabetes who experienced severe (requiring third party support) or/ and non-severe hypoglycaemia during last year. Indirect costs were estimated using a human capital approach based on lost Gross Domestic Product (GDP) and lost gross earnings. Additional estimations of direct costs, which could be attributed to hypoglycaemic event, were based on used medical resources and their unit prices. RESULTS: There were on average 0.16 episodes of severe and 4.66 episodes of non- severe hypoglycaemia per patient in the recall period (1month). An average total monthly cost of severe hypoglycaemia was 699.77 EUR, with hospitalisation being a main cost driver, and 40 EUR in the case of a non-severe episode. 23% of the studied population was professionally active. A total average time lost was equal to 3.85 hours, which in Polish conditions gave an absenteeism cost per month of 16.27 EUR (GDP lost) and 16.99 EUR (lost earnings) per person per month. 52.38% of employed patients reported reduced productivity while at work (presenteism). Its value was estimated at the level of 14.88 EUR and 15.53 EUR per patient per episode according to GDP and earnings lost respectively. Indirect costs related to sick-leaves and hospitalisations in workingage population were not reported in the study. CONCLUSIONS: Hypoglycaemic

VALUE IN HEALTH 15 (2012) A277–A575

episodes are associated with economic consequences for patients and employers. Hypoglycaemia has a negative impact on work productivity and patients’ presence at work. PHS50 HEALTH CARE RESOURCE USE AND COST IN ALZHEIMER’S DISEASE IN THREE EUROPEAN COUNTRIES – BASELINE FINDINGS OF THE GERAS STUDY Reed C1, Argimon J2, Belger MA1, Bruno G3, Dodel R4, Happich M5, Haro JM6, Jones RW7, Vellas B8, Wimo A9 Eli Lilly and Company, Windlesham, Surrey, UK, 2Servei Catala de la Salut, Barcelona, Spain, 3 University of Rome, Rome, Italy, 4Philipps University, Marburg, Hessen, Germany, 5Eli Lilly and Company, Bad Homburg, Germany, 6Parc Santari Sant Joan de Déu, CIBERSAM, Barcelona, Spain, 7The Rice Centre, Royal United Hospital, Bath, UK, 8Toulouse University Hospital, Toulouse, France, 9Karolinska Institutet, Stockholm, Sweden

1

OBJECTIVES: To describe baseline health care resource use and costs associated with Alzheimer’s disease (AD) for patients and caregivers in three European countries from the GERAS study. METHODS: A prospective, multi-centre, non-interventional cohort study in France, Germany and the UK, enrolment October 2010-October 2011. Patients presenting within the normal course of care who were ⬎55 years, diagnosed with probable AD (NINCDS-ADRDA), not institutionalised and with an informal caregiver were categorised according to Mini-Mental State Examination (MMSE) score as mild (26-21), moderate (20-15) or severe (14 or less) AD. Patient and caregiver resource use in the previous month was collected using the Resource Use in Dementia (RUD) instrument. 2010 unit costs per country were applied. RESULTS: The study cohort included 1497 AD patients (38% mild, 32% moderate and 31% severe AD) and their caregivers (n⫽419 France, n⫽552 Germany and n⫽526 UK) with a mean (SD) patient age of 79.4 (6.8) France, 75.2 (7.6) Germany, 78.5 (7.8) UK and 62% females in France, 50% Germany, 54% UK. One or more nights in hospital were recorded for 4% of patients in France, 7% Germany, 1% UK. One or more outpatient visits (including GP visits) were recorded for 84% of patients in France, 86% Germany, 73% UK. Total mean (SD) patient health care costs were €407 (970) in France, €570 (1618) in Germany, €197 (312) in the UK. Overall mean (SD) caregiver time spent on patients was highest in the UK 208.2 (189.8) hours per month [170.1 (184.8) France and 180.0 (198.4) Germany]. CONCLUSIONS: Health care resource use in the month prior to baseline was lower than expected. The UK had the fewest patients with nights spent in hospital and outpatient visits but total caregiver time was highest. Understanding differences in resource utilisation patterns between countries is necessary when examining cost of illness in AD. PHS51 HOSPITAL STAYS OF MULTIPLE SCLEROSIS PATIENTS IN GERMANY – REASONS, FREQUENCIES, DURATION AND IMPACT ON DRUG THERAPY Seiffert A1, Dippel FW2, Sommer G2, Holz B3, Trottmann M4 1 Gesundheitsforen Leipzig GmbH, Leipzig, Germany, 2Sanofi Deutschland GmbH, Berlin, Germany, 3Genzyme GmbH, Neu Isenburg, Germany, 4Gesundheitsforen Leipzig GmbH, München, Germany

OBJECTIVES: This study provides insights into hospital stays of German statutorily insured MS patients, compared to all insurants, not affected by the disease. METHODS: Information was extracted from a large longitudinal database of German health claims data. Roughly 3.000 MS patients were identified by ICD-10 diagnoses and information on drug prescriptions. The observational period lasts from 2006 until 2010. All hospitalizations of MS patients were included in the analysis. RESULTS: In 2009, MS patients had 0,57 hospitalisations on average (SD 1,30) with a mean length of stay of 6,7 days (SD 8,0). Hospital stays of all insured patients were less frequent (0,20; SD 0,71) and had an average duration of 6,9 days (SD 11,9). Roughly 46,4% of the hospital stays of MS patients were due to MS. Among those, 78% were coded with the DRG B68D [treatment of MS and cerebral ataxia without complications or comorbidities (CCs)], while 19% were coded as treatment of MS with CCs (DRG B68A). The following diagnostic procedures were observed quite often visual evoked potential (12%), somatosensory evoked potential (11,6%), and MRT (8%), as were procedures such as 6-003.f0 [Natalizumab, parental 300 – 600 mg] related to escalating drug therapy (6%). Many hospital admissions were observed shortly before changes in drug therapy: 41% of patients with a transition from basic to escalating treatment regime were hospitalized within 6 months preceding the therapy change. 11% of those patients were admitted in the month of treatment transition. Changes of medication within the basic therapy seem to be similarly affected by admissions (28% and 10% respectively). CONCLUSIONS: MS patients are hospitalized on average more than twice as often as all insurants. Almost half of all hospital stays among MS patients are related to the treatment of MS, which frequently leads to a change in drug therapy. PHS52 ESTIMATING THE RESOURCES NEEDED TO IMPLEMENT A POPULATION-BASED COLORECTAL CANCER SCREENING PROGRAM THROUGH DISCRETE EVENT SIMULATION Comas M1, Arrospide A2, Mendivil J3, Mar J2, Hernandez C3, Buron A1, Andreu M3, Castells X1 1 IMIM (Institut de Recerca Hospital del Mar) Parc de Salut MAR; CIBER de Epidemiología y Salud Pública, Barcelona, Spain, 2Hospital Alto Deba, Mondragon, Spain, 3IMIM (Institut de Recerca Hospital del Mar) Parc de Salut MAR, Barcelona, Spain

OBJECTIVES: Recent guidelines recommend colorectal cancer screening of average-risk population. Our objective was to estimate the resources needed (costs and colonoscopies) to implement a population-based colorectal cancer screening program. METHODS: A discrete-event simulation model was built to represent a colorectal cancer screening program for a target population of 100,000 women and men aged 50 to 69 years. The conceptual model for the screening process was based on the European Guidelines, which recommend biennial screening with

A527

immunochemical faecal occult blood test (iFOBT) and colonoscopy for positives of iFOBT. Follow-up after adenoma removal differed according to findings of each colonoscopy classified by risk of adenomas. Parameters for initial screening were estimated from the areas corresponding to Hospital del Mar within the Colorectal Cancer Screening Program of Barcelona. Parameters for successive screenings were obtained from the literature. A 20-year horizon was simulated. The model included the population ageing. Costs included, lettering, personnel, iFOBT and colonoscopy. RESULTS: Annual cost of screening a target population of 100,000 people varied from 899 thousand euro the first simulated year to 1,322 thousand euro the 20th year. The number of screening colonoscopies varied from 1,659 the first year to 1,790 the 20th year. Colonoscopies for following-up screening findings should be added to this numbers, starting in the second year: from 245 to 1,371 follow-up colonoscopies the 20thyear. CONCLUSIONS: Implementing a population-based colorectal cancer screening program following the European guidelines represents a cost of around 1 million euro by year the first five years for a target population of 100,000 inhabitants. The cost increases during the following 15 years until 1322 thousand euro. An important consequence of screening is the increase in the demand of confirmatory and follow-up colonoscopies related to the findings generated by the program. The demand of follow-up colonoscopies is directly related to the intensity of follow-up through colonoscopy. PHS53 USE OF HEALTH CARE ADMINISTRATIVE DATABASES TO ESTIMATE THE BURDEN OF MULTIPLE SCLEROSIS: A POPULATION-BASED STUDY Furneri G1, Ciampichini R1, Scalone L2, Cortesi PA2, Fornari C2, Madotto F2, Chiodini V2, Mantovani LG3, Cesana G2 1 Charta Foundation, Milan, Italy, 2University of Milano - Bicocca, Monza, Italy, 3Federico II University of Naples, Naples , Italy

OBJECTIVES: To assess the economic burden of multiple sclerosis (MS) from a large population-based study. METHODS: Lombardy Region includes 9.9 million individuals. Its MS population was identified through a data warehouse (DENALI), matching with probabilistic record linkage demographic, clinical and economic data of different Healthcare Administrative databases. All individuals who during the year 2006 had an hospital discharge with a IDC-9 CM code 340.XX, and/or an exemption from co-payment health care costs specific for MS (code 46.34), were selected and followed up until December 31st 2009 or death (if occurring before the last day of observation). We evaluated demographic characteristics and total costs, (hospitalizations, drugs and outpatient examinations/visits) from the National Health Service’s perspective. RESULTS: A total of 1,020 eligible subjects were identified according to the inclusion criteria. The study population (35.1% male) had a mean age of 40.5 years at the index date (median: 39.0 years). Total health care costs were €7,325, €4,753, €4,901, €5,081 per patient-year at first, second, third and fourth year of observation, respectively. During the first year hospitalizations were the costs driver (60.0% of total costs), with drugs and outpatient claims contributing to 30.4% and 9.6% of total costs, respectively. Over the following years (2, 3, 4) drug expenditure was the most relevant cost, accounting for 59%-68% of total costs. Among pharmaceutical costs, ATC-class L drugs (antineoplastic and immuno-modulating agents) contributed for 83-90%, followed by ATC-class B drugs (blood system drugs; range: 4-7%). Only 384 of the 1,020 patients (27.8%) received MS specific treatment (interferons and/or glatiramer acetate) during the first year of follow-up. CONCLUSIONS: Administrative database analysis is an efficient tool to track medical costs in MS. Diagnosis in the hospital setting is the main reason of costs at disease onset, with drugs becoming a very relevant cost with disease progression. PHS54 USE OF HEALTH CARE ADMINISTRATIVE DATABASES TO ESTIMATE THE HEALTH CARE BURDEN OF BONE OR BONE MARROW METASTATIC DEASES IN BREAST CANCER PATIENTS: A POPULATION-BASED STUDY Mantovani LG1, Scalone L2, Furneri G3, Ciampichini R3, Cortesi PA2, Fornari C2, Madotto F2, Chiodini V2, Cesana G2 Federico II University of Naples, Naples , Italy, 2University of Milano - Bicocca, Monza, Italy, 3 Charta Foundation, Milan, Italy

1

OBJECTIVES: To assess the burden of disease of bone or bone marrow metastases (B/BMM) in breast cancer (BC) patients from a large population-based. METHODS: Lombardy Region includes 9.9 million individuals. Eligible patients were identified through a data warehouse (DENALI), which matches with a probabilistic linkage demographic, clinical and economic data of different Healthcare Administrative databases. After excluding patients with diagnosis of BC during the 2000-2003 period, all female who during the year 2004 had a first hospital discharge with a IDC-9-CM code 174.XX, and diagnosis B/BMM (ICD-9-CM 198.5) were identified and followed up to 5 years or death (if occurring before the last day of observation). We evaluated demographic characteristics and total costs, (hospitalizations, drugs and outpatient examinations/visits) from the National Health Service’s perspective. RESULTS: A total of 10,435 eligible subjects were identified, 5.2% with B/BMM. Compared with patients without metastatic disease, patients with B/BMM were older (66.6 vs 62.1 years), showed higher mortality (34.5 vs 3.7 deaths/100 patient/ years), and increased per patient health care cost (Euro/patient-year) in the first (21,014 vs 11,931), second (10,576 vs 4,190), third (9,394 vs 3,305), fourth (8,132 vs 3,364) and fifth (8,485 vs 3,360) year after diagnosis. In 2004, hospitalizations were the driver of total cost in both B/BMM (72.5%) and non-B/BMM (70.2%) groups, followed by drugs (21.0% and 14.9%) and outpatient claims (6.5% and 14.9%). Over the following years (2005-2009) drug expenditure on total costs increased in both B/BMM (41.5%) and non-B/BMM (40.7%) groups. All differences presented above, are statistically significant at the conventional level of significance. CONCLUSIONS: Administrative database analysis is an efficient tool to track medical costs in pa-