G Goodall

ED2 LONG-TERM COST-UTILITY ANALYSIS OF INSULIN ASPART (NOVORAPID®) VERSUS HUMAN SOLUBLE INSULIN INTYPE 2 DIABETES PATIENTS IN THE GERMAN SETTING

pies until the first 30-day gap. Propensity-score weighted logistic regression and proportional hazard models were used to adjust for baseline demographics, copay and pharmacy utilization variables. RESULTS: At baseline, across the SPAA and 2PAA groups (N = 1,530), mean age was 62 years, 49% were female, 10% utilized coronary vasodilators, 28% utilized anti-diabetics; mean number of other baseline medications was 7. These characteristics varied among all cohorts. Patients receiving SPAA were nearly twice as likely to achieve adherence, and approximately 20% less likely to discontinue therapy at all doses; compared to the European dose equivalents, the adjusted odds ratio of achieving adherence was 1.83 (95% confidence interval (CI) 1.60–2.10, P < 0.0001) and the discontinuation hazard ratio was 0.83 (CI 0.74–0.93, P = 0.0012). CONCLUSION: Single-pill amlodipine/ atorvastatin was associated with greater adherence and less discontinuation vs. 2-pill amlodipine and atorvastatin, at low, high, and European doses of both medications.

ED1 COST-EFFECTIVENESS OF INSULIN DETEMIR VERSUS NPH FOR TYPE 1 DIABETES PATIENTS IN A GERMAN SETTING. A MODELING EVALUATION BASED UPON RESULTS FROM A META-ANALYSIS

pies until the first 30-day gap. Propensity-score weighted logistic regression and proportional hazard models were used to adjust for baseline demographics, copay and pharmacy utilization variables. RESULTS: At baseline, across the SPAA and 2PAA groups (N = 1,530), mean age was 62 years, 49% were female, 10% utilized coronary vasodilators, 28% utilized anti-diabetics; mean number of other baseline medications was 7. These characteristics varied among all cohorts. Patients receiving SPAA were nearly twice as likely to achieve adherence, and approximately 20% less likely to discontinue therapy at all doses; compared to the European dose equivalents, the adjusted odds ratio of achieving adherence was 1.83 (95% confidence interval (CI) 1.60–2.10, P < 0.0001) and the discontinuation hazard ratio was 0.83 (CI 0.74–0.93, P = 0.0012). CONCLUSION: Single-pill amlodipine/ atorvastatin was associated with greater adherence and less discontinuation vs. 2-pill amlodipine and atorvastatin, at low, high, and European doses of both medications.

PDB32 EVALUATING THE LONG-TERM CLINICAL AND ECONOMIC IMPLICATIONS OF CONVERTING TYPE 2 DIABETES PATIENTS TO INSULIN DETEMIR (±ORAL HYPOGLYCEMIC AGENTS) FROM INSULIN GLARGINE BASED REGIMENS IN GERMANY; DATA FROMTHE PREDICTIVE STUDY

hypoglycaemia reduction. METHODS: A discrete event life time simulation with microvascular complications incorporated via the DCCT (Diabetes Control and Complications Trial) study and cardiovascular events modelled using the Framingham equations was adapted to include the combined effects of HbA1c and hypoglycaemia reduction using published meta-regression results from 11 randomised clinical trials. Direct costs and quality of life (EQ5D) were derived from published sources and the HODaR database respectively; costs and benefits were discounted annually at 3.5%. The model was adapted to the profile of T1DM patients switched from NPH to glargine identified via the THIN database. Analysis was conducted on a total of 383 patients with data for the 12 month period prior to, and post switch; using primary outcome measure of adjusted Hba1c change. As hypoglycaemia was not directly collected from the THIN database a sensitivity analysis was performed taking into account HbA1c benefit only. RESULTS: The median age of patients switched from NPH to glargine was 34 years with mean duration of T1DM of 11.4 years. Baseline HbA1c was 8.71% and patients switching to glargine showed a reduction in HbA1c of 0.195% (p = 0.0045) between switch and 12-months post initiation. In a simulated cohort of 10,000 the discounted incremental cost effectiveness ratio (ICER) was 3,665 per quality adjusted life year gained (QALY). In sensitivity analysis using HbA1c benefit only the ICER was 9,411. CONCLUSION: Based on real life observational data, switching to glargine is cost-effective when compared to NPH; with a corresponding ICER well within accepted thresholds, even in sensitivity analysis using HbA1c effect only.

PDB34 LONG-TERM COST-EFFECTIVENESS ANALYSIS OF A MODERN INSULIN IN PATIENTS WITH POORLY CONTROLLED TYPE 2 DIABETES IN THE GERMAN SETTING; DATA FROMTHE PREDICTIVE STUDY

and body weight (BMI 0.52 kg.m-2). The aim of this analysis was to estimate the long-term clinical and cost implications associated with therapy conversion from insulin glargine to detemir in type 2 diabetes patients in Germany. METHODS: A previously published and validated diabetes model (CORE Diabetes Model) was used to make long-term projections of clinical and cost outcomes based on patient characteristics (age 62.3 years, duration of diabetes 7 years, HbA1c 8.30%, 50.4% male) and treatment effects from the German part of PREDICTIVE. The model was used to estimate life-expectancy, quality-adjusted life expectancy and to account direct medical costs (pharmacy, patient management and complication costs). Costs were derived from published sources and expressed in 2006 Euros. Future costs and clinical benefits were discounted at 5% annually. RESULTS: Therapy conversion from insulin glargine to insulin detemir was projected to improve life expectancy by approximately 0.13 years (7.08 0.13 versus 6.95 0.12 years) and quality-adjusted life expectancy by 0.29 quality-adjusted life years (QALYs) (4.53 0.09 versus 4.24 0.08 QALYs). Direct costs associated with insulin detemir treatment were projected to be lower over patient lifetimes than with glargine (€ 54,807 1,788 versus € 55,839 1,749 per patient, difference € 1,032). Cost savings were driven by lower complication costs (due to HbA1c improvements) associated with insulin detemir. CONCLUSION: Modeling the long-term implications of therapy conversion from insulin glargine to detemir based on data from German patients in PREDICTIVE indicates that insulin detemir is associated with benefits in terms of life expectancy, qualityadjusted life expectancy and complication rates, as well as reducing costs from a third-party health care payer perspective in Germany.

PDB37 MEALTIME INSULIN ASPART REDUCES THE LONG-TERM COST OF COMPLICATIONS COMPARED TO HUMAN INSULIN AS PART OF BASAL-BOLUS THERAPY IN POLISH TYPE 2 DIABETES PATIENTS

Cost analysis: Costs during the first year of treatment after switching to either BOT (glargine plus glimepirid and metformine) or CT (premixed insulin 30/70) were analysed. Costs for antidiabetic agents, injection devices (pens, syringes) and monitoring devices (test strips, lancets) were included. Other medical supplies were assumed to be the same or not relevant from the SHI perspective. Insulin utilization was extrapolated up to one year by linear regression based on LAPTOP trial data. The cost of drugs, devices and supplies (retail prices) were taken from official standard sources. Univariate sensitivity analyses were performed for all cost parameters (variation 20%). DMMSimulation: Baseline values for the model simulation were as follows: mean age of the population 60 9.0 years, mean duration of diabetes 9.0 7.0 years and mean HbA1c value 8.8 0.9%. The response rate (HbA1c 7%) for BOT was 49% and for CT 39%. Mean HbA1c values for responders were 6.46% and 6.55% respectively. RESULTS: Annual savings with the BOT regimen add up to 236.35€. Insulin utilization and insulin prices had the highest impact on overall costs. The relative risk reduction (RRR) for micro-vascular events after 10 simulation years varied between 14% (end stage renal disease) and 2% (retinopathy) in favour of BOT versus CT. CONCLUSION: The combined results of these two analyses show that BOT in addition to its short (HbA1c) and long term (microvascular events) medical benefits is associated with lower costs from the SHI perspective.

PDB20 COST-EFFECTIVENESS OF INSULIN DETEMIR VERSUS NPH FOR TYPE 1 DIABETES PATIENTS IN AN ITALIAN SETTING. A META-ANALYSIS

COST-EFFECTIVENESS OF INSULIN DETEMIRVERSUS NPH FORTYPE 1 DIABETES PATIENTS IN AN ITALIAN SETTING. A META-ANALYSIS Aristides M, Kotchie R, Nielsen S,Aagren M,ValentineWJ, Goodall G IMS Health, London, UK, Novo Nordisk A/S,Virum, Denmark, IMS Health, Basel, Switzerland OBJECTIVES: A meta-analyses of the results from three clinical trials in type 1 diabetes patients showed that insulin detemir (IDet) based basal-bolus treatment compared to neutral protamine Hagedorn (NPH) insulin based basal-bolus therapy led to improved HbA1c (0.13% points lower), a decrease in hypoglycaemic events (by 4%) and lower body mass index (BMI) (0.21 kg.m-2). METHODS: A published, validated, peerreviewed computer simulation model of diabetes (the CORE Diabetes Model) was used to project short-term results obtained from the fixed-effects (weighted average) meta-analysis to longterm clinical and cost outcomes (including life expectancy, quality adjusted life expectancy, incidence of complications, and direct medical costs) for IDet versus NPH in type 1 diabetes patients, when used in combination with either insulin aspart (IAsp) or human soluble insulin (HSI) as the bolus component of therapy. Probabilities of complications were derived from landmark clinical and epidemiological studies and the costs of treating complications in Italy were retrieved from published sources. Total direct costs (complications + treatment costs) for each arm were projected over patient lifetimes from an Italian Heath Service perspective. Both costs and clinical outcomes were discounted at 3% annually. RESULTS: Improved glycaemic control, decreased hypoglycaemic events and lower BMI with IDet-based basal/bolus therapy led to fewer diabetes-related complications, an increase in quality-adjusted life expectancy of 0.185 qualityadjusted life years (QALYs) (7.71 0.09 versus 7.53 0.09 QALYs), increased total lifetime costs/patient of €5,680 (€104,234 2354 versus €98,554 2124), and an incremental cost-effectiveness ratio of €30,704 per QALY gained. Results were stable under variation in a range of reasonable assumptions. CONCLUSION: The increased cost of therapy for IDet versus NPH is partly offset by reductions in the treatment costs of complications and, given the associated clinical benefits, leads to a cost-effectiveness ratio which falls within a range generally considered to represent excellent value for money (<€50,000/ QALY gained).

PDB22 EVALUATING THE LONG-TERM COST-EFFECTIVENESS OF INSULIN ASPART (NOVORAPID®) VERSUS HUMAN SOLUBLE INSULIN IN PATIENTS WITHTYPE 2 DIABETES IN SPAIN AND ITALY

COST-EFFECTIVENESS OF INSULIN DETEMIRVERSUS NPH FORTYPE 1 DIABETES PATIENTS IN AN ITALIAN SETTING. A META-ANALYSIS Aristides M, Kotchie R, Nielsen S,Aagren M,ValentineWJ, Goodall G IMS Health, London, UK, Novo Nordisk A/S,Virum, Denmark, IMS Health, Basel, Switzerland OBJECTIVES: A meta-analyses of the results from three clinical trials in type 1 diabetes patients showed that insulin detemir (IDet) based basal-bolus treatment compared to neutral protamine Hagedorn (NPH) insulin based basal-bolus therapy led to improved HbA1c (0.13% points lower), a decrease in hypoglycaemic events (by 4%) and lower body mass index (BMI) (0.21 kg.m-2). METHODS: A published, validated, peerreviewed computer simulation model of diabetes (the CORE Diabetes Model) was used to project short-term results obtained from the fixed-effects (weighted average) meta-analysis to longterm clinical and cost outcomes (including life expectancy, quality adjusted life expectancy, incidence of complications, and direct medical costs) for IDet versus NPH in type 1 diabetes patients, when used in combination with either insulin aspart (IAsp) or human soluble insulin (HSI) as the bolus component of therapy. Probabilities of complications were derived from landmark clinical and epidemiological studies and the costs of treating complications in Italy were retrieved from published sources. Total direct costs (complications + treatment costs) for each arm were projected over patient lifetimes from an Italian Heath Service perspective. Both costs and clinical outcomes were discounted at 3% annually. RESULTS: Improved glycaemic control, decreased hypoglycaemic events and lower BMI with IDet-based basal/bolus therapy led to fewer diabetes-related complications, an increase in quality-adjusted life expectancy of 0.185 qualityadjusted life years (QALYs) (7.71 0.09 versus 7.53 0.09 QALYs), increased total lifetime costs/patient of €5,680 (€104,234 2354 versus €98,554 2124), and an incremental cost-effectiveness ratio of €30,704 per QALY gained. Results were stable under variation in a range of reasonable assumptions. CONCLUSION: The increased cost of therapy for IDet versus NPH is partly offset by reductions in the treatment costs of complications and, given the associated clinical benefits, leads to a cost-effectiveness ratio which falls within a range generally considered to represent excellent value for money (<€50,000/ QALY gained).

PDB25 COST-EFFECTIVENESS OF DETEMIR VERSUS NPH FOR TYPE 1 DIABETES PATIENTS TREATED WITH BASAL-BOLUS THERAPY IN A BELGIUM SETTING, A MODELING EVALUATION BASED ON RESULTS FROM A META-ANALYSIS OFTHREE CLINICAL TRIALS

COST-EFFECTIVENESS OF BIPHASIC INSULIN ASPART 30 VERSUS HUMAN PREMIX INSULIN FORTYPE 2 DIABETES PATIENTS IN A POLISH SETTING Aristides M, Kotchie R, Nielsen S,Townsend C,ValentineWJ, Scheijbeler H IMS Health, London, UK, Novo Nordisk A/S,Virum, Denmark, IMS Health, Basel, Switzerland OBJECTIVES: The aim of this analysis was to project the longterm clinical and economic outcomes associated with therapy conversion from human premix insulin to Biphasic Insulin Aspart 30 (BIAsp 30) in type 2 diabetes patients in a Polish setting. METHODS: A previously published and validated computer simulation model for diabetes was used to make long-term projections of clinical and cost outcomes based on patient characteristics and treatment effects from a sub-analysis of the PRESENT study (1219 patients). The study analyzed the impact of converting patients receiving human premix insulin (with or without conventional oral medication) to treatment with BIAsp 30 whilst maintaining existing oral therapy. Probabilities of complications were derived from landmark clinical and epidemiological studies and the costs of treating complications in Poland were retrieved from published sources. Total direct costs (complications + treatment costs) were projected over patient lifetimes with both costs and clinical outcomes discounted at 5% per annum. RESULTS: Improved glycemic control (HbA1c reduction of 1.82%) and decreased hypoglycemic events associated with BIAsp 30 were projected to lead to fewer diabetesrelated complications and an increase in quality-adjusted life expectancy of 0.280 quality-adjusted life years (QALYs) (3.338 0.075 versus 3.058 0.072 QALYs). The reduction in predicted diabetes-related complications resulted in a net saving in direct medical costs of PLN 7,790 (PLN 28,746 1,097 versus 36,536 1,379). CONCLUSION: This modeling study indicated that the increased cost of therapy for BIAsp 30 versus human premix insulin will be offset by reductions in the cost of diabetes-related complications leading to a net saving in direct costs. Moreover, BIAsp 30 was associated with improved life expectancy and quality-adjusted life expectancy, making it a dominant treatment option compared to human premix insulin.