This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn.

The use of simvastatin to treat dyslipidaemia, and to reduce the risk of myocardial infarction in patients with diabetes or with cardiovascular disease (CVD).

Primary and secondary prevention.

Cost-effectiveness analysis.

The study population comprised patients with diabetes but without CVD, and patients with CVD but without diabetes.

The setting was unclear, but it appears to have been the community. The economic study was carried out in several European countries.

The effectiveness and resource use data were derived from studies published in 1998 and 1999. The price year was 1998.

The effectiveness evidence was derived from published studies and from the authors' assumptions.

The Cardiovascular Disease Life Expectancy Model (Markov cycles) was used to estimate the annual probability of fatal and nonfatal CVD events. A cohort of patients with or without CVD was entered into the model at a specific age and with specified levels of risk factors. Each patient could die of coronary disease, cerebrovascular disease, or other causes. Those who survived re-entered the model after one year, until all patients died.

The health outcomes assessed from the published studies, which were used as model inputs, were:

the distribution of simvastatin dosage;

the reductions in total cholesterol and low-density lipoprotein (LDL) cholesterol obtained with simvastatin; and

the increase in high-density lipoprotein (HDL) cholesterol obtained with simvastatin.

Not reported.

Not stated.

Not stated.

Not stated.

The effectiveness evidence was derived from three primary studies.

A narrative method was used to combine the primary studies.

Not reported.

The simvastatin dosage was distributed as follows:

61.6% of the patients were given 20 mg/day, 31.6% were given 40 mg/day, 0.1% were given 10 mg/day, and 6.7% discontinued the therapy.

There was a 25% reduction in total cholesterol with simvastatin, a 35% reduction in LDL cholesterol, and an 8% increase in HDL cholesterol.

The authors made some assumptions to create two homogeneous patient populations.

The authors assumed that all of the patients had specific lipid profiles. These included total cholesterol levels of 261 mg/dL, a LDL cholesterol of 188 mg/dL, a HDL cholesterol of 46 mg/dL, and triglyceride levels of 132 mg/dL.

The blood pressure was assumed to be 138.5 mmHg over 83.2 mmHg.

All patients were assumed to be non-smokers.

All risk factor levels and the impact of treatment on lipid levels were assumed to be uniform across the age groups.

The benefit measure used in both study groups was the number of years of life saved (YOLS) with simvastatin over placebo. The YOLS were derived from the decision model and discounted at 3%.

A 3% discount rate was used for the costs incurred in future years. The quantities of resources used were not reported. Only the average hospitalisation costs (including inpatient physician fees) were reported. The cost/resource boundary adopted was that of the third-party payer. The cost items included in the analysis were hospitalisation, physician fees, outpatient care, emergency services, and drug prescriptions. It was estimated that a diabetic patient with no prior CVD would have two physician visits, two routine biochemical panels, and two haemoglobin (Hb) A1c tests per year. All patients also had a lipid profile, a urinalysis, and an annual consultation with an ophthalmologist. For those patients who developed CVD, the management of diabetes only included the additional cost of the ophthalmologist visit, the HbA1c test, and the urinalysis.

US costs were derived from published reports such as the national Medicare Provider Analysis and Review data, and the national sample of the Healthcare Cost and Utilization Project. The costs for the other countries were mainly derived from independent local costing studies or published data. The total costs were derived from the decision model. The price year was 1998.

Statistical analyses of the costs were not carried out.

The Indirect costs were not included.

US dollars ($). Local currencies were used for Canada (Canadian dollars, Can$), France (French francs, Ffr), Germany (German marks, DM), Italy (Italian lira, L), Spain (Spanish pesetas, Pta), and the UK (UK pounds sterling, ).

Sensitivity analyses were carried out to test for the robustness of the estimated cost-effectiveness ratios. The model inputs varied were the age and gender of the patients, and the health care costs across the seven countries.

For men, the undiscounted YOLS after simvastatin therapy in patients with diabetes but no CVD were 5.36 at age 40 years, 4.13 at age 50 years, 2.56 at age 60 years, and 0.81 at age 70 years. The corresponding values for patients with CVD but not diabetes were 3.90 at age 40 years, 3.16 at age 50 years, 2.12 at age 60 years, and 0.78 at age 70 years.

For women, the undiscounted YOLS after simvastatin therapy in patients with diabetes but no CVD were 2.83 at age 40 years, 2.36 at age 50 years, 1.59 at age 60 years, and 0.54 at age 70 years. The corresponding values for patients with CVD but not diabetes were 2.72 at age 40 years, 2.29 at age 50 years, 1.62 at age 60 years, and 0.63 at age 70 years.

The total costs in the two study groups were not reported.

The costs and the benefits were combined using an incremental cost-effectiveness analysis. The (incremental) cost per YOLS with simvastatin (over no treatment) ranged from $5,063 to $14,156 in diabetic men without CVD, from $8,799 to $14,996 in non-diabetic men with CVD, from $13,121 to $23,792 in diabetic women without CVD, and from $14,164 to $21,628 in non-diabetic women with CVD.

The results for both study populations were consistent in all countries examined. The results were not sensitive to variations of the key model inputs varied in the sensitivity analyses.

In all seven countries examined in the study, the lipid treatment with simvastatin was as cost-effective in diabetic patients without cerebrovascular disease (CVD) as in non-diabetics with CVD.

The rationale for the choice of the comparator was not explicitly stated. The simvastatin-based therapy could have been compared with no treatment in order to assess the active value of the intervention. You should assess whether no intervention represents a feasible comparator in your own setting.

The authors provided a narrative overview of the effectiveness data derived from the literature and their assumptions, which were used in the decision model. Details on the design of the primary studies were not given. In addition, differences across the studies were not taken into account.

The number of life-years saved represented the benefit measure used in the economic analysis. It was derived from the decision model and appears to have been appropriate for the study question. However, valuations in terms of quality of life would have been useful.

The unit costs and the quantities of resources were not reported separately. The details of the cost items included in the study were inadequate. Only average hospitalisation costs were reported. Statistical analyses of the costs were not carried out and, although sensitivity analyses were performed on the total health care costs, the results of the sensitivity analyses were not reported clearly.

The authors made some comparisons of their findings with those from other studies. The issue of the external validity of the study was addressed by considering cost data from seven countries. However, few sensitivity analyses were carried out. The study population comprised both male and female patients with CVD and without diabetes, or with diabetes and without CVD, and this was reflected in the authors' conclusions. The authors pointed out that their analysis was based on a simulation model and some assumptions were made to select two homogeneous study populations, while in the real world the relative risk of myocardial infarction may be different across patient sub-groups.

The authors state that the results of the study strongly support the cost-effectiveness of simvastatin-based lipid therapy for the prevention of myocardial infarction in patients with diabetes or with CVD. They recommend that further research should use data from real populations.

Supported by Merck & Co., Inc.