|Model-based simulation to explore the cost-effectiveness of following practice guidelines for triglyceride and low-density lipoprotein cholesterol control among patients with diabetes mellitus and mixed dyslipidemia
|Sorensen SV, Frick KD, Wade A, Simko R, Burge R
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.
This study assessed the cost-effectiveness of guidelines for treatment to reduce cholesterol and triglycerides, in the prevention of coronary heart disease, in patients with diabetes mellitus and mixed dyslipidaemia. The authors concluded that the guidelines were cost-effective compared with usual care from the perspective of the health care payer. The methods appear to have been valid and the results were clearly presented. The authors’ conclusions appear to be robust.
Type of economic evaluation
Cost-effectiveness analysis, cost-utility analysis
This study assessed the cost-effectiveness of guidelines for the prevention of coronary heart disease (CHD) by reducing cholesterol and triglycerides in a typical 60-year-old patient with diabetes mellitus and mixed dyslipidaemia.
The National Cholesterol Education Program Adult Treatment Panel III (NCEP ATP III) guidelines were compared with the usual care. These guidelines had the following target lipid levels: low-density lipoprotein cholesterol (LDL-C) 100mg per dL, triglycerides 150mg per dL, high-density lipoprotein cholesterol (HDL-C) 40mg per dL for men or 50mg per dL for women, and total cholesterol 200mg per dL.
The treatment depended on the patient’s uncontrolled lipid levels. Patients with uncontrolled LDL-C, who were not on drug therapy, began treatment with statin monotherapy. Those with uncontrolled triglycerides, who were not on drug therapy, began treatment with fibrate monotherapy. All other patients began on a combination of a statin and a fibrate.
The analysis was based on a simulation with two time horizons of one year and 20 years. The authors stated that the perspective of the health care payer was adopted.
The clinical data were from a selection of relevant studies. The baseline patient data (including lipid levels) were from a sample of 20,938 patients in an electronic database of records of over 30 million patients across 49 US states. The efficacy of treatment was the key input to the model and was from a published clinical trial. The long-term data on the incidence of CHD and stroke were from the Framingham Heart Study. Assumptions were made, based on published data, for the impact of body mass index (BMI), smoking, menopause, and the use of antihypertensive medication. Assumptions were also made for the efficacy of various drug doses.
Monetary benefit and utility valuations:
The utility values were based on European Quality of life (EQ-5D) scores derived from a sample of 38,678 adults. Multivariate analysis considered the impact of the baseline characteristics of patients.
Measure of benefit:
Quality-adjusted life-years (QALYs) and CHD event rates were the summary benefit measures and they were discounted at an annual rate of 3%.
The economic analysis included the costs of drugs, treatment of medical events (stroke, myocardial infarction, other CHD events, and death), and management of diabetes. The distribution of drug therapies among patients was based on data from the electronic database, while the drug costs were based on average wholesale prices. The costs and quantities associated with medical events were from the hospital databases of six US states and published literature. The long-term costs were based on the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes and the price year was 2007. All costs were in US dollars ($) and those incurred after the first year were discounted at an annual rate of 3%.
Analysis of uncertainty:
A probabilistic analysis was undertaken to assess uncertainty, using probability distributions that best fitted the original data. One-way sensitivity analyses were also conducted on specific inputs, such as age, sex, adherence, drug efficacy and costs, and the time horizon. The ranges of values were from the literature or authors’ opinions.
In the short-term (one-year), the guidelines were more effective than usual care, with 72% of patients achieving two lipid targets and 44% achieving three lipid targets, and they led to an incremental cost per patient of $1,020.
Over 20 years, the guidelines led to an incremental gain of 0.18 QALYs, with 176 fewer CHD events (1,000 patients), and an additional cost of $8,993 per patient, resulting in an incremental cost per QALY gained of $50,315 and an incremental cost per CHD event avoided of $52,107. The cost per QALY ranged from $12,690, in patients with uncontrolled LDL-C and triglycerides, who were on a fibrate and moved to combination therapy, to $376,772, in patients with uncontrolled triglycerides, who were not on therapy and began on a fibrate.
The probabilistic analysis showed that the probability of the guidelines being cost-effective was 45% at a threshold of $50,000 per QALY and 90% at a threshold of $100,000 per QALY.
The univariate analysis showed that the results were sensitive to variations in the baseline characteristics of the patients. The guidelines were generally more cost-effective for men than for women, in older cohorts, and with longer horizons. In most scenarios, the incremental cost per QALY gained remained below the threshold of $100,000, but assumptions on drug adherence strongly influenced the results. For example, 53% adherence (100% in the base case) increased the cost-effectiveness ratio to over $115,000 per QALY.
The authors concluded that treatment according to the guidelines was cost-effective compared with usual care for patients with diabetes and mixed dyslipidaemia, from the perspective of the health care payer.
The rationale for the selection of the comparators was clear as the proposed intervention was compared with the usual care. The usual care was not clearly described, which reduces the ability to apply the results to other settings.
The sources of clinical evidence were selected by the authors, who did not report a literature review. The use of a large database to derive the baseline demographic and clinical characteristics of the patients was appropriate as it was representative of the actual impact of disease. The use of efficacy data from clinical trials was also appropriate as their methods are considered to be reliable. The authors extensively described the source for the clinical data. Both benefit measures were appropriate, with disease-related events being relevant for clinicians, while QALYs are more generalisable and capture the impact of the disease on both survival and quality of life. The utility values were appropriately derived from a large database of patients and conventional discounting was applied.
The analysis of costs was consistent with the perspective and only the direct medical costs were included. Their sources reflected the US reimbursement system and might not be applicable to other health care systems. Limited resource use data were provided, as often occurs when US tariffs are used. Other characteristics of the economic analysis, such as the discount rate and the price year, were reported.
Analysis and results:
The analytic approach used to synthesise the costs and benefits was valid and the results were clearly presented. The issue of uncertainty was extensively investigated, using appropriate approaches, and the methods and findings were clearly presented and discussed. The authors conducted an extensive comparison of their study with other published economic evaluations and highlighted the large variation in cost-effectiveness results for statins. They acknowledged some limitations of their analysis, such as the exclusion of adverse events, the assumptions of equal efficacy for each drug dose, and the use of short-term clinical trials for some efficacy estimates.
The methods appear to have been valid and the results were clearly presented. The authors’ conclusions appear to be robust.
Funded by Abbott Laboratories.
Sorensen SV, Frick KD, Wade A, Simko R, Burge R. Model-based simulation to explore the cost-effectiveness of following practice guidelines for triglyceride and low-density lipoprotein cholesterol control among patients with diabetes mellitus and mixed dyslipidemia. Clinical Therapeutics 2009; 31(4)
Subject indexing assigned by NLM
Adult; Aged; Aged, 80 and over; Cholesterol, LDL /blood /drug effects; Computer Simulation; Coronary Disease /etiology /prevention & Cost-Benefit Analysis; Databases, Factual; Diabetes Mellitus /blood; Drug Therapy, Combination; Dyslipidemias /complications /drug therapy /economics; Female; Humans; Hypolipidemic Agents /economics /therapeutic use; Male; Middle Aged; Models, Statistical; Practice Guidelines as Topic /standards; Quality-Adjusted Life Years; Risk Factors; Time Factors; Triglycerides /blood; United States /epidemiology; control
Date bibliographic record published
Date abstract record published