- Email: [email protected]
PCV16 DIFFERENCES IN RESOURCE UTILIZATION IN POST-MI PATIENTS WITH AND WITHOUT HEART FAILURE
322 PCAS was 3.3%, the overall adverse event rate was 3.1% and the 30-day mortality rate was 1.2%. A chi-square analysis of the adverse event rates revealed no clear relationship between the study group and proportion of adverse events among all 18 studies (¥2 = 14.6, p > 0.10). Multiple sensitivity analyses were conducted; however, no relationship between the study group and proportion of adverse events resulted (¥2 = 11.2, p > 0.10). CONCLUSION: Although practitioners await stronger evidence, this study demonstrates the relatively low rates of adverse events in PCAS relative to CAS alone and the warranted use of PCAS. PCV14 A COMPARISON OF THE RISK OF ADVERSE THROMBOEMBOLIC AND BLEEDING EVENTS BETWEEN SUBJECTS TREATED AND NOT TREATED WITH WARFARIN
Riedel A1, Hauch O2, Harley C1, Nelson M1, Wygant G2, Reynolds MW2 1 Ingenix, Eden Prairie, MN, USA; 2AstraZeneca, L.P, Wilmington, DE, USA OBJECTIVE: This study was conducted to assess the risk of thromboembolic and bleeding events among atrial fibrillation patients treated with warfarin. METHODS: Using claims data from a large commercial health plan, patients with chronic atrial fibrillation were identified based on medical claims with diagnosis codes 427.31 and 427.32 from 1998 through 1999. Patients with valvular disease were excluded. Cox proportional hazards analysis was used to compare risk of venous, arterial, intracranial, and total thromboembolic events between warfarin exposed and unexposed subjects. Risk of bleeding events was also compared. RESULTS: A total of 6764 subjects were retained for analysis; of these 3541 (52.4%) were exposed to warfarin during the follow-up period. Among thromboembolic events, treated subjects were significantly less likely to experience arterial events compared to non-treated subjects (HR: 0.710, CI: 0.540–0.934). No differences in the risk of venous or intracranial events were found, nor in the risk of thromboembolic events overall. Use of warfarin significantly increased the risk of minor bleeding events (HR: 3.600, CI: 2.537–5.109), and all bleeding events (HR: 1.502, CI: 1.289–1.749). CONCLUSIONS: A large number of atrial fibrillation patients are not being treated with warfarin. After adjusting for baseline characteristics, the risk of thromboembolic events in this population was not significantly different between those exposed and unexposed to warfarin. There was a significant increase in the risk for minor and total bleeding events among patients treated with warfarin. Both observational studies and decision analytic models have indicated that the outcomes of warfarin therapy is highly dependent of how it is managed. The present study seems to indicate that there is a significant gap between the performance of warfarin in reducing the risk of thromboembolic events as shown in tightly controlled clinical trials or coagulation clinics versus what is achievable in general practice. CARDIOVASCULAR DISEASE (including Obesity) CARDIOVASCULAR DISEASE (including Obesity)—Cost Studies PCV15 ECONOMIC EVALUATION OF GLYCOPROTEIN IIB/IIIA ANTAGONISTS IN DIABETIC PATIENTS WITH ACUTE CORONARY SYNDROME UNDERGOING PERCUTANEOUS CORONARY INTERVENTIONS WITH STENTING
Mittmann N1, Brown A2, Seung SJ3, Noorani H2, Mensinkai S2, Cohen E4, Risebrough N3, Oh P5, Tang Z2
Abstracts 1 HOPE Research Centre, Toronto, ON, Canada; 2Canadian Coordinating Office For Health Technology Assessment (CCOHTA), Ottowa, ON, Canada; 3HOPE Research Centre, Toronto, ON, Canada; 4Sunnybrook & Women’s College Health Sciences Centre, Toronto, ON, Canada; 5Toronto Rehabilitation Institute (Rumsey Centre), Toronto, ON, Canada OBJECTIVES: Diabetic patients have a higher risk of coronary artery disease. Percutaneous coronary intervention (PCI) with stenting has become the standard of care to repair coronary vessel blockage. However, stenting increases the risk of thrombus formation at the implantation site. Platelet glycoprotein IIb/IIIa antagonist drugs reduce this risk. This study examines whether abciximab and eptifibatide are cost-effective as adjunct therapies in diabetic patients undergoing PCI with stenting. METHODS: Included were diabetic patients undergoing elective or urgent PCI with stenting. Clinical outcome data was extracted from the published EPISTENT and ESPRIT trials. Abciximab (N = 335) and eptifibatide (N = 419), with stenting, were compared to a treatment group of stent-only patients. Short-term (one year) and long-term (survival-Markov model) decision analytic models (DATA 4.0) were constructed from a Canadian provincial health system perspective. Results are reported in 2001 Canadian dollars and presented on a per person basis. A 5% discount rate was used. Probabilistic sensitivity analysis (SA) was done using Crystal Ball. RESULTS: For abciximab + stent, incremental costs were higher (+$81) but clinical outcomes were better (-18.5% major adverse cardiac events [MACE] and -3% mortality) relative to the stent-only group. The incremental cost-effectiveness ratio for abciximab was $438 per MACE avoided and $2700 per death avoided. Abciximab + stent patients had 0.22 more adjusted life years (LYs) than the stent-only group so the incremental ratio was $368 per adjusted LYs gained. When compared to the stent-only group, eptifibatide + stent was dominant in terms of costs (Incremental = $166), MACE rate (Incremental = 7.1%) and mortality (Incremental = 2%) over the short-term. There was a 0.22 LY increase for eptifibatide. SA indicated that drug acquisition and procedure costs were the major cost drivers. Results were robust. CONCLUSIONS: Eptifibatide and abciximab (+stenting) are considered cost-effective in the treatment of diabetic patients undergoing PCI.
PCV16 DIFFERENCES IN RESOURCE UTILIZATION IN POST-MI PATIENTS WITH AND WITHOUT HEART FAILURE
Arondekar BV1, Basu A2, Manning WG2, Walton SM1 1 University of Illinois at Chicago, Chicago, IL, USA; 2University of Chicago, Chicago, IL, USA OBJECTIVES: The economic consequences of heart failure (HF) in post-myocardial infarction (MI) patients can be severe. Recently there has been an increase in the availability of drugs aimed at treating or delaying the onset of post-MI HF. The objective of this study was to estimate differences in costs, number of hospitalizations and outpatient visits in post-MI patients with and without HF. METHODS: Claims data for patients hospitalized with a principal diagnosis of MI between 1998 and 2000 were used. Patients with a diagnosis of HF or MI in the sixmonths preceding the initial MI were excluded. Data on 13,682 patients for a period of 3-years following discharge for initial MI was available. The follow-up period was divided into 18 twomonth intervals. Mean costs were analyzed using a two-part model (logistic and generalized linear model (GLM)), and outpatient visits and hospitalizations were analyzed using a GLM model with log-link. Clustering of observations within patients was adjusted for, and bootstrapping was used to obtain standard errors. Age, gender, type of MI, insurance type, and comorbidi-
323
Abstracts ties were included as covariates. RESULTS: Following discharge from initial hospitalization, the mean cost was $3755.28, the mean number of outpatient visits was 4.14, and the mean number of hospitalizations was 0.11 per two-month period, for patients with post-MI HF. If these patients had not developed HF, the estimated savings per 2-month interval was $1716, and the estimated reduction in the number of visits and hospitalizations per 2-month interval was 30% and 50% respectively. CONCLUSIONS: The development of HF in post-MI patients results in a significant increase in costs, outpatient visits and hospitalizations as compared to post-MI patients without HF. The cost of newer drugs that aim to delay or prevent the onset of post-MI HF may be offset by savings in costs and resource utilization attributable to HF in post-MI patients. PCV17 A COST-UTILITY ANALYSIS OF LOSARTAN VERSUS ATENOLOL IN THE TREATMENT OF MILD TO MODERATE HYPERTENSION
Anis AH1, Sun H1, Singh S2, Woolcott JC3, Nosyk B2 University of British Columbia, Vancouver, BC, Canada; 2Canadian HIV Trials Network, Vancouver, BC, Canada; 3BC Centre for Disease Control, Vancouver, BC, Canada OBJECTIVES: The Losartan Intervention for End Point Reduction in Hypertension (LIFE) study demonstrated a 13% relative risk reduction in the primary composite endpoint (myocardial infarction, stroke or death) and a 25% relative risk reduction in stroke for patients treated with losartan compared to patients treated with a first-line antihypertensive agent (atenolol). Incorporating the results found in the LIFE study, an incremental costeffectiveness analysis was performed to determine the economic feasibility of making losartan a first line anti-hypertensive agent in the prevention of cardiovascular morbidity and mortality. METHODS: A Markov State Transition model was developed to extrapolate the outcomes observed during the 4-year LIFE trial to the patients’ lifetime. A comprehensive model was created by drawing information from a number of sources within the literature, including other longitudinal cohort studies on patients in the defined health states. Considering a societal perspective, fully allocated costs were calculated using the St. Paul’s Hospital Cost Model. QALY estimates for each of the given health states were obtained from a variety of sources within the literature. Along with several univariate sensitivity analyses, extensive probabilistic sensitivity analysis was performed in order to examine the impact of a broad range of variation in our model parameters. RESULTS: The incremental cost effectiveness ratio of losartan versus atenolol was CDN$1337 (CDN$1 = US$0.78) per QALY gained. This cost-effectiveness ratio was robust to extensive sensitivity analysis, demonstrating a 95% probability that the cost-effectiveness ratio would be less than CDN$20,000 per QALY gained. CONCLUSIONS: Losartan appears to be an exciting, cost-effective alternative to traditional first-line therapies for hypertension. Results were robust to univariate and probabilistic sensitivity analysis and are well within the accepted ranges of cost-effectiveness ratios deemed to be efficient and cost effective. 1
PCV18 COST-EFFECTIVENESS ANALYSIS OF ROSUVASTATIN COMPARED TO ATORVASTATIN, FLUVASTATIN, LOVASTATIN, PRAVASTATIN, AND SIMVASTATIN
Benner JS1, Smith TW1, Klingman D1, Tierce JC1, Mullins CD2, Pethick N3, O’Donnell J3 1 ValueMedics Research, LLC, Arlington, VA, USA; 2University of Maryland, Baltimore, MD, USA; 3Astrazeneca LP, Wilmington, DE, USA
OBJECTIVE: To determine the cost-effectiveness of rosuvastatin compared to five other statins, from a managed care payer perspective. METHODS: A decision-analytic model compared the 1-year costs and effectiveness of the following six statins, titrated over the specified dose ranges in patients with elevated lowdensity lipoprotein cholesterol (LDL): atorvastatin 10–80 mg, fluvastatin 40–80 mg, generic lovastatin 20–80 mg, pravastatin 20–40 mg, rosuvastatin 10–40 mg, and simvastatin 20–80 mg. Effectiveness measures included percent change in LDL, highdensity lipoprotein cholesterol (HDL), and number of patients achieving National Cholesterol Education Program (NCEP) Second Adult Treatment Panel goal (per 10,000 treated). Effectiveness estimates came from two 52-week, comparative clinical trials and dose equivalence tables. Drug, physician and laboratory resource use were estimated using current NCEP guidelines, then multiplied by Medicare reimbursement rates for services, and wholesale acquisition costs for drugs. Probabilistic sensitivity analyses employed the net health benefits framework and cost-effectiveness acceptability curves were constructed. RESULTS: In the base-case analysis, rosuvastatin dominated atorvastatin, lovastatin, pravastatin and simvastatin. Fluvastatin was least costly but also least effective; rosuvastatin was most effective and had the second lowest cost. Compared with fluvastatin, the incremental LDL-C reduction, HDL-C increase, and percent of patients to goal with rosuvastatin were 16%, 3%, and 27%, respectively. Incremental costs per additional 1% reduction in LDL-C, a 1% increase in HDL-C, and patient to goal were $6, $33, and $353, respectively. Results were robust to probabilistic analyses: in each of 1000 simulated populations of 1000 patients, rosuvastatin conferred more net health benefits than atorvastatin, lovastatin, pravastatin, or simvastatin. Rosuvastatin was the optimal choice among all statins when willingness to pay for an additional 1% decrease in LDL, 1% increase in HDL, and patient to goal exceeded $6, $37, and $353, respectively. CONCLUSIONS: Rosuvastatin is less costly and more effective than atorvastatin, lovastatin, pravastatin and simvastatin; and highly cost-effective compared with fluvastatin. PCV19 GENDER AND STATIN THERAPY COST AND EFFICACY
Larsen TL, Smith DG University of Michigan, Ann Arbor, MI, USA OBJECTIVE: Coronary heart disease (CHD) is among the leading causes of premature death in women. Cholesterol lowering therapies are beneficial in primary and secondary prevention of CHD. This study calculates the cost and efficacy of treating women with various HMGCoA reductase inhibitors (statins). METHODS: Our analysis uses data from the Atorvastatin Comparative Cholesterol Efficacy and Safety Study (ACCESS), a randomized, active-controlled, clinical trial. In January 1999, 3273 patients completed the trial which had assigned them to one of five study medications: atorvastatin 10–80 mg/day; fluvastatin 20–80 mg/day; lovastatin 20–40 mg/day; pravastatin 20–40 mg/day: or simvastatin 10–40 mg/day. We examined both costs of treatment to NCEP targets and efficacy in achieving targets for men and women. Cost analyses employed ordinary least squares regression and efficacy analyses employed logistic regression. RESULTS: In this study, we found women to be $47 less costly to treat with lipid-lowering therapies than men (95% CI: $US10-84), controlling for medication, age, weight, race, blood pressure, baseline cholesterol level, number of risk factors, diabetes, hypertension, and smoking status. We found no difference in the likelihood of achieving NCEP targets (odds ratio 0.984, 95% CI: 0.776–1.248), controlling for the same factors as in the cost analysis. CONCLU-
Riedel A1, Hauch O2, Harley C1, Nelson M1, Wygant G2, Reynolds MW2 1 Ingenix, Eden Prairie, MN, USA; 2AstraZeneca, L.P, Wilmington, DE, USA OBJECTIVE: This study was conducted to assess the risk of thromboembolic and bleeding events among atrial fibrillation patients treated with warfarin. METHODS: Using claims data from a large commercial health plan, patients with chronic atrial fibrillation were identified based on medical claims with diagnosis codes 427.31 and 427.32 from 1998 through 1999. Patients with valvular disease were excluded. Cox proportional hazards analysis was used to compare risk of venous, arterial, intracranial, and total thromboembolic events between warfarin exposed and unexposed subjects. Risk of bleeding events was also compared. RESULTS: A total of 6764 subjects were retained for analysis; of these 3541 (52.4%) were exposed to warfarin during the follow-up period. Among thromboembolic events, treated subjects were significantly less likely to experience arterial events compared to non-treated subjects (HR: 0.710, CI: 0.540–0.934). No differences in the risk of venous or intracranial events were found, nor in the risk of thromboembolic events overall. Use of warfarin significantly increased the risk of minor bleeding events (HR: 3.600, CI: 2.537–5.109), and all bleeding events (HR: 1.502, CI: 1.289–1.749). CONCLUSIONS: A large number of atrial fibrillation patients are not being treated with warfarin. After adjusting for baseline characteristics, the risk of thromboembolic events in this population was not significantly different between those exposed and unexposed to warfarin. There was a significant increase in the risk for minor and total bleeding events among patients treated with warfarin. Both observational studies and decision analytic models have indicated that the outcomes of warfarin therapy is highly dependent of how it is managed. The present study seems to indicate that there is a significant gap between the performance of warfarin in reducing the risk of thromboembolic events as shown in tightly controlled clinical trials or coagulation clinics versus what is achievable in general practice. CARDIOVASCULAR DISEASE (including Obesity) CARDIOVASCULAR DISEASE (including Obesity)—Cost Studies PCV15 ECONOMIC EVALUATION OF GLYCOPROTEIN IIB/IIIA ANTAGONISTS IN DIABETIC PATIENTS WITH ACUTE CORONARY SYNDROME UNDERGOING PERCUTANEOUS CORONARY INTERVENTIONS WITH STENTING
Mittmann N1, Brown A2, Seung SJ3, Noorani H2, Mensinkai S2, Cohen E4, Risebrough N3, Oh P5, Tang Z2
Abstracts 1 HOPE Research Centre, Toronto, ON, Canada; 2Canadian Coordinating Office For Health Technology Assessment (CCOHTA), Ottowa, ON, Canada; 3HOPE Research Centre, Toronto, ON, Canada; 4Sunnybrook & Women’s College Health Sciences Centre, Toronto, ON, Canada; 5Toronto Rehabilitation Institute (Rumsey Centre), Toronto, ON, Canada OBJECTIVES: Diabetic patients have a higher risk of coronary artery disease. Percutaneous coronary intervention (PCI) with stenting has become the standard of care to repair coronary vessel blockage. However, stenting increases the risk of thrombus formation at the implantation site. Platelet glycoprotein IIb/IIIa antagonist drugs reduce this risk. This study examines whether abciximab and eptifibatide are cost-effective as adjunct therapies in diabetic patients undergoing PCI with stenting. METHODS: Included were diabetic patients undergoing elective or urgent PCI with stenting. Clinical outcome data was extracted from the published EPISTENT and ESPRIT trials. Abciximab (N = 335) and eptifibatide (N = 419), with stenting, were compared to a treatment group of stent-only patients. Short-term (one year) and long-term (survival-Markov model) decision analytic models (DATA 4.0) were constructed from a Canadian provincial health system perspective. Results are reported in 2001 Canadian dollars and presented on a per person basis. A 5% discount rate was used. Probabilistic sensitivity analysis (SA) was done using Crystal Ball. RESULTS: For abciximab + stent, incremental costs were higher (+$81) but clinical outcomes were better (-18.5% major adverse cardiac events [MACE] and -3% mortality) relative to the stent-only group. The incremental cost-effectiveness ratio for abciximab was $438 per MACE avoided and $2700 per death avoided. Abciximab + stent patients had 0.22 more adjusted life years (LYs) than the stent-only group so the incremental ratio was $368 per adjusted LYs gained. When compared to the stent-only group, eptifibatide + stent was dominant in terms of costs (Incremental = $166), MACE rate (Incremental = 7.1%) and mortality (Incremental = 2%) over the short-term. There was a 0.22 LY increase for eptifibatide. SA indicated that drug acquisition and procedure costs were the major cost drivers. Results were robust. CONCLUSIONS: Eptifibatide and abciximab (+stenting) are considered cost-effective in the treatment of diabetic patients undergoing PCI.
PCV16 DIFFERENCES IN RESOURCE UTILIZATION IN POST-MI PATIENTS WITH AND WITHOUT HEART FAILURE
Arondekar BV1, Basu A2, Manning WG2, Walton SM1 1 University of Illinois at Chicago, Chicago, IL, USA; 2University of Chicago, Chicago, IL, USA OBJECTIVES: The economic consequences of heart failure (HF) in post-myocardial infarction (MI) patients can be severe. Recently there has been an increase in the availability of drugs aimed at treating or delaying the onset of post-MI HF. The objective of this study was to estimate differences in costs, number of hospitalizations and outpatient visits in post-MI patients with and without HF. METHODS: Claims data for patients hospitalized with a principal diagnosis of MI between 1998 and 2000 were used. Patients with a diagnosis of HF or MI in the sixmonths preceding the initial MI were excluded. Data on 13,682 patients for a period of 3-years following discharge for initial MI was available. The follow-up period was divided into 18 twomonth intervals. Mean costs were analyzed using a two-part model (logistic and generalized linear model (GLM)), and outpatient visits and hospitalizations were analyzed using a GLM model with log-link. Clustering of observations within patients was adjusted for, and bootstrapping was used to obtain standard errors. Age, gender, type of MI, insurance type, and comorbidi-
323
Abstracts ties were included as covariates. RESULTS: Following discharge from initial hospitalization, the mean cost was $3755.28, the mean number of outpatient visits was 4.14, and the mean number of hospitalizations was 0.11 per two-month period, for patients with post-MI HF. If these patients had not developed HF, the estimated savings per 2-month interval was $1716, and the estimated reduction in the number of visits and hospitalizations per 2-month interval was 30% and 50% respectively. CONCLUSIONS: The development of HF in post-MI patients results in a significant increase in costs, outpatient visits and hospitalizations as compared to post-MI patients without HF. The cost of newer drugs that aim to delay or prevent the onset of post-MI HF may be offset by savings in costs and resource utilization attributable to HF in post-MI patients. PCV17 A COST-UTILITY ANALYSIS OF LOSARTAN VERSUS ATENOLOL IN THE TREATMENT OF MILD TO MODERATE HYPERTENSION
Anis AH1, Sun H1, Singh S2, Woolcott JC3, Nosyk B2 University of British Columbia, Vancouver, BC, Canada; 2Canadian HIV Trials Network, Vancouver, BC, Canada; 3BC Centre for Disease Control, Vancouver, BC, Canada OBJECTIVES: The Losartan Intervention for End Point Reduction in Hypertension (LIFE) study demonstrated a 13% relative risk reduction in the primary composite endpoint (myocardial infarction, stroke or death) and a 25% relative risk reduction in stroke for patients treated with losartan compared to patients treated with a first-line antihypertensive agent (atenolol). Incorporating the results found in the LIFE study, an incremental costeffectiveness analysis was performed to determine the economic feasibility of making losartan a first line anti-hypertensive agent in the prevention of cardiovascular morbidity and mortality. METHODS: A Markov State Transition model was developed to extrapolate the outcomes observed during the 4-year LIFE trial to the patients’ lifetime. A comprehensive model was created by drawing information from a number of sources within the literature, including other longitudinal cohort studies on patients in the defined health states. Considering a societal perspective, fully allocated costs were calculated using the St. Paul’s Hospital Cost Model. QALY estimates for each of the given health states were obtained from a variety of sources within the literature. Along with several univariate sensitivity analyses, extensive probabilistic sensitivity analysis was performed in order to examine the impact of a broad range of variation in our model parameters. RESULTS: The incremental cost effectiveness ratio of losartan versus atenolol was CDN$1337 (CDN$1 = US$0.78) per QALY gained. This cost-effectiveness ratio was robust to extensive sensitivity analysis, demonstrating a 95% probability that the cost-effectiveness ratio would be less than CDN$20,000 per QALY gained. CONCLUSIONS: Losartan appears to be an exciting, cost-effective alternative to traditional first-line therapies for hypertension. Results were robust to univariate and probabilistic sensitivity analysis and are well within the accepted ranges of cost-effectiveness ratios deemed to be efficient and cost effective. 1
PCV18 COST-EFFECTIVENESS ANALYSIS OF ROSUVASTATIN COMPARED TO ATORVASTATIN, FLUVASTATIN, LOVASTATIN, PRAVASTATIN, AND SIMVASTATIN
Benner JS1, Smith TW1, Klingman D1, Tierce JC1, Mullins CD2, Pethick N3, O’Donnell J3 1 ValueMedics Research, LLC, Arlington, VA, USA; 2University of Maryland, Baltimore, MD, USA; 3Astrazeneca LP, Wilmington, DE, USA
OBJECTIVE: To determine the cost-effectiveness of rosuvastatin compared to five other statins, from a managed care payer perspective. METHODS: A decision-analytic model compared the 1-year costs and effectiveness of the following six statins, titrated over the specified dose ranges in patients with elevated lowdensity lipoprotein cholesterol (LDL): atorvastatin 10–80 mg, fluvastatin 40–80 mg, generic lovastatin 20–80 mg, pravastatin 20–40 mg, rosuvastatin 10–40 mg, and simvastatin 20–80 mg. Effectiveness measures included percent change in LDL, highdensity lipoprotein cholesterol (HDL), and number of patients achieving National Cholesterol Education Program (NCEP) Second Adult Treatment Panel goal (per 10,000 treated). Effectiveness estimates came from two 52-week, comparative clinical trials and dose equivalence tables. Drug, physician and laboratory resource use were estimated using current NCEP guidelines, then multiplied by Medicare reimbursement rates for services, and wholesale acquisition costs for drugs. Probabilistic sensitivity analyses employed the net health benefits framework and cost-effectiveness acceptability curves were constructed. RESULTS: In the base-case analysis, rosuvastatin dominated atorvastatin, lovastatin, pravastatin and simvastatin. Fluvastatin was least costly but also least effective; rosuvastatin was most effective and had the second lowest cost. Compared with fluvastatin, the incremental LDL-C reduction, HDL-C increase, and percent of patients to goal with rosuvastatin were 16%, 3%, and 27%, respectively. Incremental costs per additional 1% reduction in LDL-C, a 1% increase in HDL-C, and patient to goal were $6, $33, and $353, respectively. Results were robust to probabilistic analyses: in each of 1000 simulated populations of 1000 patients, rosuvastatin conferred more net health benefits than atorvastatin, lovastatin, pravastatin, or simvastatin. Rosuvastatin was the optimal choice among all statins when willingness to pay for an additional 1% decrease in LDL, 1% increase in HDL, and patient to goal exceeded $6, $37, and $353, respectively. CONCLUSIONS: Rosuvastatin is less costly and more effective than atorvastatin, lovastatin, pravastatin and simvastatin; and highly cost-effective compared with fluvastatin. PCV19 GENDER AND STATIN THERAPY COST AND EFFICACY
Larsen TL, Smith DG University of Michigan, Ann Arbor, MI, USA OBJECTIVE: Coronary heart disease (CHD) is among the leading causes of premature death in women. Cholesterol lowering therapies are beneficial in primary and secondary prevention of CHD. This study calculates the cost and efficacy of treating women with various HMGCoA reductase inhibitors (statins). METHODS: Our analysis uses data from the Atorvastatin Comparative Cholesterol Efficacy and Safety Study (ACCESS), a randomized, active-controlled, clinical trial. In January 1999, 3273 patients completed the trial which had assigned them to one of five study medications: atorvastatin 10–80 mg/day; fluvastatin 20–80 mg/day; lovastatin 20–40 mg/day; pravastatin 20–40 mg/day: or simvastatin 10–40 mg/day. We examined both costs of treatment to NCEP targets and efficacy in achieving targets for men and women. Cost analyses employed ordinary least squares regression and efficacy analyses employed logistic regression. RESULTS: In this study, we found women to be $47 less costly to treat with lipid-lowering therapies than men (95% CI: $US10-84), controlling for medication, age, weight, race, blood pressure, baseline cholesterol level, number of risk factors, diabetes, hypertension, and smoking status. We found no difference in the likelihood of achieving NCEP targets (odds ratio 0.984, 95% CI: 0.776–1.248), controlling for the same factors as in the cost analysis. CONCLU-