|Cost-effectiveness of screening for coronary artery disease in asymptomatic patients with Type 2 diabetes and additional atherogenic risk factors
|Hayashino Y, Nagata-Kobayashi S, Morimoto T, Maeda K, Shimbo T, Fukui T
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.
Screening strategies for coronary artery disease (CAD) in asymptomatic diabetic patients with two additional atherogenic risk factors were examined. The strategies were:
exercise electrocardiography followed by coronary angiography (CAG) if positive;
exercise echocardiography followed by CAG if positive; and
exercise single-photon emission computed tomography (SPECT) followed by CAG if positive.
Economic study type
The study population comprised cohorts of men aged between 50 and 70 with Type 2 diabetes mellitus and 10 possible pairs of atherogenic risk factors. Such risk factors included hypertension, smoking, low-density lipoprotein level above 160 mg/dL, high-density lipoprotein level below 35 mg/dL, and proteinuria.
The setting was secondary care. The economic study was conducted in the USA.
Dates to which data relate
The effectiveness evidence was derived from literature published between 1983 and 2003. The resource use data were derived from literature published between 1998 and 2003. The price year was 2003.
Source of effectiveness data
The effectiveness data were derived from a review of published literature and authors' assumptions.
A Markov model was conducted to examine the benefits and costs of the screening strategies. The seven CAD states in the model were normal, silent ischaemia, symptomatic ischaemia, history of myocardial infarction (MI), post-percutaneous transluminal coronary angiography (post-PTCA), post-coronary artery bypass grafting (post-CABG) and death. The cycle length was not reported. The time horizon was 30 years. The model assumed the following:
a diagnostic test was performed once only, at the first stage;
patients with negative screening test results received no specific therapy;
treatment after angiography was based on the anatomic pattern of vessel obstruction, and patients without vessel obstruction on angiography received no specific treatment;
patients with 1- or 2-vessel CAD on angiography underwent PTCA, and PTCA reduced late MI in those with 1- or 2-vessel disease;
patients with 3-vessel or left-main trunk disease on angiography underwent CABG;
patients who developed restenosis after PTCA underwent PTCA once more only, so that PTCA was performed not more than twice;
patients with MI developed relevant symptoms and received the appropriate treatment; and
patients with silent ischaemia did not receive specific anti-anginal therapy.
Outcomes assessed in the review
The outcomes derived from the review were:
the prevalence of asymptomatic ischaemia in the base-case (estimated);
the annual incidence of CAD;
the proportion of silent ischaemia in patients with MI;
the sensitivity and specificity of the tests;
the mortality risk reduction by CABG;
the risk reduction in late MI with revascularisation;
the risk reduction in the incidence of CAD by aspirin (primary prevention);
the risk reduction in late CAD events by aspirin (secondary prevention);
the risk reduction in all-cause mortality by aspirin (secondary prevention);
the risk reduction in the incidence of CAD by simvastatin (primary prevention); and
the annual risk for revascularisation by initial treatment and extent of disease.
Study designs and other criteria for inclusion in the review
The sensitivity and specificity of exercise electrocardiography were obtained from a meta-analysis of diagnostic tests from 150 studies. The sensitivity and specificity of exercise echocardiography and exercise SPECT came from another meta-analysis, based on a summary of receiver operating characteristics curve analysis. The other study designs were not reported.
Sources searched to identify primary studies
Criteria used to ensure the validity of primary studies
Methods used to judge relevance and validity, and for extracting data
Number of primary studies included
Thirteen studies were reviewed in order to extract effectiveness evidence.
Methods of combining primary studies
Transition probabilities were calculated using Framingham data, estimated from equations developed by Anderson et al. (see 'Other Publications of Related Interest' below for bibliographic details). The prevalence of CAD in each risk state was calculated from the prevalence of CAD in the general population and the odds ratio for associated atherogenic risk factors for CAD, stratified by age and gender (see Appendix II, which is available online at www.jgim.org). The prevalence of CAD in different age groups of diabetics was derived from the Third National Health and Nutrition Examination Survey (NHANES III), and the prevalence data then calibrated to match the estimated number of patients with stable angina according to the American College of Cardiology/American Heart Association, as NHANES data were likely to be influenced by self-reporting bias. The relative risks were estimated by multivariate logistic regression using NHANES III data. The cycle-specific mortality in each health state was estimated on the basis of a declining exponential approximation to life expectancy.
Investigation of differences between primary studies
The authors do not appear to have investigated differences between the primary studies.
Results of the review
The prevalence of asymptomatic ischaemia was 0.39.
The annual incidence of CAD per 1,000 was 23.4.
The proportion of silent ischaemia in patients with MI was 0.4.
The sensitivities and specificities were, respectively, 0.68 and 0.77 for exercise electrocardiography, 0.85 and 0.77 for exercise echocardiography, and 0.87 and 0.64 for exercise SPECT.
The risk reduction in mortality due to CABG was 15% for 1- or 2-vessel disease, 48% for 3-vessel disease, and 67% for left-main trunk disease.
The risk reduction in late MI with revascularisation was 17% for PTCA and 42% for CABG.
The risk reduction was 15% in the incidence of CAD by aspirin (primary prevention), 31% in late CAD events by aspirin (secondary prevention), 28% in all-cause mortality by aspirin (secondary prevention), and 25% in the incidence of CAD by simvastatin (primary prevention).
The annual risk for revascularisation by initial treatment and extent of disease was 0.01 for non-specific therapy (1-vessel disease), 0.042 for non-specific therapy (2-vessel disease), 0.075 for non-specific therapy (3-vessel disease or left-main trunk disease), 0.036 for PTCA and 0.018 for CABG.
Methods used to derive estimates of effectiveness
The authors made several assumptions concerning short-term sequels to intervention.
Estimates of effectiveness and key assumptions
The probabilities of death and nonfatal MI associated with CAG were 0.1% and 0.06%, respectively.
The mortality rates associated with PTCA were 0.2% in 1-vessel disease and 0.9% in 2-vessel disease, while the mortality rates of nonfatal MI were 3.5% (1-vessel) and 5.2% (2-vessel), respectively.
The mortality rate associated with CABG was 3.2% and the probability of nonfatal MI was 7.0%.
The efficacy of PTCA in patients with 1- or 2-vessel disease was assumed to be the same as that of CABG.
The risks of subsequent nonfatal MI, PTCA, or CABG were assumed to depend on the extent of the CAD and the type of initial treatment.
Measure of benefits used in the economic analysis
The summary benefit measures were the quality-adjusted life-years (QALYs). An annual discount rate of 3% was applied. The QALYs were obtained from the Markov model. The health utility was obtained from the review of literature.
The cost boundary adopted in the economic analysis was that of society. The cost categories included were costs of diagnostic tests, drug costs and costs of outpatient visits, self-testing and care management (including hospitalisation, professional costs and outpatient treatment). Non medical costs, such as patient travel, waiting and treatment time associated with office visits, were also included. The resource use data were obtained from published literature. The unit costs were not reported separately from the quantities of resources used for all cost items. The exercise test costs were obtained from Medicare-allowed charges, hospital costs were taken from Medicare administrative data, and professional costs came from the Medicare schedule. The cost of simvastatin was calculated from average wholesale prices. Other unit costs were derived from literature. A discount rate of 3% was used. The price year was 2003.
Statistical analysis of costs
The costs were treated deterministically in the base-case.
The indirect costs were not considered.
One- and two-way sensitivity analyses were performed to test the robustness of the estimated cost-effectiveness and incremental cost-effectiveness ratios to variations in the baseline model inputs. Ninety-five per cent confidence intervals (CIs) were adopted as the ranges of values, where applicable, otherwise a range of +/- 30% was used.
Estimated benefits used in the economic analysis
For patients aged 55 years, the expected QALYs were 11.754 with no screening, 11.826 with exercise electrocardiography, 11.849 with exercise echocardiography, and 11.841 with exercise SPECT. Compared with no screening, the incremental QALYs gained were 0.072 with exercise electrocardiography and 0.095 with exercise echocardiography.
For patients aged 60 years, the expected QALYs were 9.911 with no screening, 10.063 with exercise electrocardiography, 10.104 with exercise echocardiography, and 10.100 with exercise SPECT. Compared with no screening, the incremental QALYs gained were 0.152 with exercise electrocardiography and 0.193 with exercise echocardiography.
For patients aged 55 years, the expected costs per patient were $135,447 with no screening, $142,179 with exercise electrocardiography, $143,847 with exercise echocardiography, and $144,806 with exercise SPECT. Compared with no screening, the incremental costs were $6,732 with exercise electrocardiography and $8,400 with exercise echocardiography.
For patients aged 60 years, the costs per patient were $123,301 with no screening, $129,617 with exercise electrocardiography, $131,180 with exercise echocardiography, and $132,129 with exercise SPECT. Compared with no screening, the incremental costs were $6,316 with exercise electrocardiography and $7,879 with exercise echocardiography.
Synthesis of costs and benefits
Compared with no screening, the incremental cost-effectiveness ratio of exercise electrocardiography was $93,500/QALY gained and that of exercise echocardiography was $88,400/QALY gained in 55-year-old patients. Therefore, exercise electrocardiography was weakly dominated by exercise echocardiography. In 60-year-old patients, the incremental cost-effectiveness ratio of exercise electrocardiography was $41,600/QALY gained and that of exercise echocardiography was $40,800/QALY gained. Therefore, exercise electrocardiography was weakly dominated by exercise echocardiography. The strategy exercise SPECT was strictly dominated in both age groups.
The results were sensitive to age and combination of additional atherogenic risk factors. The incremental cost-effectiveness ratio of exercise echocardiography decreased from $327,400 to $25,600/QALY gained as the age rose from 50 to 70 years. The incremental cost-effectiveness ratio of exercise echocardiography ranged from $83,700 to $126,300/QALY gained in 55-year-old men, depending on the additional atherogenic risk factors. In 60-year-old men, the model was fairly insensitive to the risk factors.
The sensitivity analyses also revealed that the results were sensitive to diagnostic test performance. For both exercise echocardiography and electrocardiography, the incremental cost-effectiveness ratio exceeded $50,000/QALY gained if the sensitivity was very high, so that the specificity was very low.
The screening of asymptomatic patients with diabetes and two or more additional atherogenic risk factors was cost-effective from a societal perspective. Exercise echocardiography dominated over electrocardiography and exercise single-photon emission computed tomography (SPECT).
CRD COMMENTARY - Selection of comparators
The rationale for the choice of the comparator appears to have been appropriate, as it covered all possible screening strategies that met the requirement for the research aim. The strategy of no screening was also included for comparative purposes. You should decide if these are widely used health technologies in your own setting.
Validity of estimate of measure of effectiveness
The effectiveness evidence used as model inputs was mainly derived from a review of published literature. However, it was unclear whether a systematic review had been conducted. The methods used to extract the primary estimates were not reported, nor were the criteria used to ensure the validity of the primary studies. However, the methods used to combine the studies were described in detail. The study design for some of the primary studies was a meta-analysis, which may raise the validity of the sources used. Some values of the clinical parameters were based on authors' assumptions, which were supported by published studies.
Validity of estimate of measure of benefit
The choice of QALYs as a summary measure of benefits was appropriate as it reflected the impact of the screening strategies on the patients' health status. However, the authors provided no details on the method used to measure the utilities. Therefore, it is not possible to be certain of the validity of the utility weights. Discounting was appropriately conducted.
Validity of estimate of costs
The perspective adopted in the study was explicitly stated. Although a societal perspective was adopted, the indirect costs (i.e. lost productivity) were not included and the authors did not provide any justification for their exclusion. It appears that all the relevant categories of costs have been included in the economic analysis. However, the authors did not report how they estimated the costs associated with false-positive and false-negative results with diagnostic strategies. Therefore, it was unclear whether or not such costs were accounted for in the cost analysis. Information on the unit costs and quantity of resources used was not provided, which may limit the possibility of replicating the study. The costs were treated deterministically in the base-case, but sensitivity analyses were conducted. The price year was reported. Discounting was conducted, which was appropriate since the follow-up period was more than 2 years. This aids reflation exercises in other settings.
The authors compared their findings with those from another published study and stated that its conclusion supported their results. The issue of the generalisability to other settings was addressed by performing sensitivity analyses that enhanced the external validity of the study. The authors noticed some limitations resulting from the assumptions made in the Markov model. For instance, they stated that some of the assumptions might have resulted in an underestimation of the cost-effectiveness of CAD screening.
Implications of the study
This study suggested that exercise echocardiography offered the most rational use of health care resources for CAD screening in asymptomatic patients with diabetes and additional atherogenic risk factors, followed by exercise electrocardiography.
Hayashino Y, Nagata-Kobayashi S, Morimoto T, Maeda K, Shimbo T, Fukui T. Cost-effectiveness of screening for coronary artery disease in asymptomatic patients with Type 2 diabetes and additional atherogenic risk factors. Journal of General Internal Medicine 2004; 19(12): 1181-1191
Other publications of related interest
Fleischmann KE, Hunink MG, Kuntz KM, Douglas PS. Exercise echocardiography or exercise SPECT imaging? A meta-analysis of diagnostic test performance. JAMA 1998;280:913-20.
Littenberg B, Garber AM, Sox HC Jr. Screening for hypertension. Annals of Internal Medicine 1990;112:192-202.
Sox HC Jr, Littenberg B, Garber AM. The role of exercise testing in screening for coronary artery disease. Annals of Internal Medicine 1989;110:456-69.
Turner RC, Millns H, Neil HA, et al. Risk factors for coronary artery disease in non-insulin dependent diabetes mellitus: United Kingdom Prospective Diabetes Study (UKPDS: 23). BMJ 1998;316:823-8.
Weinstein MC, Coxson PG, Williams LW, et al. Forecasting coronary heart disease incidence, mortality, and cost: the Coronary Heart Disease Policy Model. American Journal of Public Health 1987;77:1417-26.
Subject indexing assigned by NLM
Aged; Coronary Disease /diagnosis; Cost-Benefit Analysis; Decision Support Techniques; Diabetes Mellitus, Type 2 /complications; Echocardiography /economics; Electrocardiography /economics; Exercise Test /economics; Humans; Hypertension /complications; Male; Markov Chains; Middle Aged; Prognosis; Sensitivity and Specificity; Smoking /adverse effects; Tomography, Emission-Computed, Single-Photon /economics
Date bibliographic record published
Date abstract record published