Disclaimer: The Medical Advisory Secretariat uses a standardized costing method for its economic analyses of interventions. The main cost categories and the associated methods from the province’s perspective are as follows:
Hospital: Ontario Case Costing Initiative cost data are used for in-hospital stay, emergency visit and day procedure costs for the designated International Classification of Diseases (ICD) diagnosis codes and Canadian Classification of Health Interventions procedure codes. Adjustments may be required to reflect accuracy in estimated costs of the diagnoses and procedures under consideration. Due to the difficulties of estimating indirect costs in hospitals associated with a particular diagnosis or procedure, the secretariat normally defaults to considering direct treatment costs only.
Nonhospital: These include physician services costs obtained from the Ontario Schedule of Benefits, laboratory fees from the Ontario Schedule of Laboratory Fees, drug costs from the Ontario Drug Benefit Formulary, and device costs from the perspective of local health care institutions whenever possible or its manufacturer.
Discounting: For cost-effectiveness analyses, a discount rate of 5% is applied as recommended by economic guidelines.
Downstream costs: All numbers reported are based on assumptions on population trends (i.e. incidence, prevalence and mortality rates), time horizon, resource utilization, patient compliance, healthcare patterns, market trends (i.e. rates of intervention uptake or trends in current programs in place in the Province), and estimates on funding and prices. These may or may not be realized by the system or individual institutions and are often based on evidence from the medical literature, standard listing references and educated hypotheses from expert panels. In cases where a deviation from this standard is used, an explanation is offered as to the reasons, the assumptions, and the revised approach. The economic analysis represents an estimate only, based on the assumptions and costing methods that have been explicitly stated above. These estimates will change if different assumptions and costing methods are applied to the analysis.
The objective of this economic analysis is to determine the cost effectiveness of cardiac magnetic resonance imaging (cardiac MRI) for the diagnosis of patients with suspected CAD as compared to: stress ECHO, stress contrast ECHO, SPECT, and CT angiography. The relative cost-effectiveness of these five non-invasive cardiac imaging technologies was assessed in two patient populations: 1) out-patients presenting with stable chest pain; and 2) in-patients presenting with acute, unstable chest pain. Note that the term “contrast ECHO” used in the following sections refers to stress ECHO performed with a contrast medium.
Economic Analysis Overview
A decision-analytic cost-effectiveness analysis was conducted to evaluate the relative cost-effectiveness of five non-invasive cardiac imaging technologies for diagnosing CAD in two patient populations: 1) out-patients presenting with stable chest pain; and 2) in-patients presenting with acute, unstable chest pain. Two decision analytic models were developed for these patient populations with two reported outcomes: the cost per accurate diagnosis of CAD and the cost per true positive diagnosis of CAD.
The physician and hospital costs for the non-invasive imaging tests were taken from 2009 Ontario Health Insurance Plan (OHIP) and the Ontario Case Costing Initiative (OCCI) administrative databases. (14
) A budget impact analysis (BIA) was performed assessing the effect of replacing a certain proportion of stress ECHO tests with other cost-effective, non-invasive modalities. The costs presented in this BIA were estimated from Ontario data sources from 2009; the volumes of tests performed were estimated from data from fiscal years 2002 to 2008.
Economic Literature Review
The purpose of the systematic review of economic literature was to identify, retrieve, and summarize studies evaluating the cost-effectiveness of selected cardiac imaging tests for the diagnosis of CAD. Medline and the National Health Service Economic Evaluation Database (NHSEED) were searched from their inception up to October 2009. Included studies were those full economic evaluations describing both costs and consequences of a) CT angiography, b) Cardiac MRI, c) SPECT, d) stress ECHO, and e) stress contrast ECHO in the diagnosis of CAD. Article selection was performed by independent pairs of researchers. Target data for extraction included: study first author and year of publication, imaging tests compared, type of economic analysis, reported costs and outcomes, incremental cost-effectiveness ratio (ICER), currency, and patient characteristics (i.e., known or suspected CAD and risk of CAD). The primary outcome of interest for the present systematic review was the ICER of each imaging test in relation to another test of interest.
Literature Search Results
A total of 883 non-duplicate citations were found from the two electronic databases after applying the literature search strategy. Of these, 147 full-text articles were retrieved for further assessment of their inclusion/exclusion, following which, 122 were rejected leaving 25 articles for inclusion in the systematic review. After the data extraction process, 13 studies were excluded (16
), with 12 studies being ultimately selected for analysis.(28
) Reasons for the exclusion of articles are described in The Relative Cost-effectiveness of Five Non-invasive Cardiac Imaging Technologies for Diagnosing Coronary Artery Disease in Ontario
Characteristics of Included Studies
From the 12 studies included in the present systematic review, eight assessed the cost-effectiveness of two of the selected imaging tests (31
), three studies evaluated three concomitant technologies (28
), and one study evaluated five technologies.(29
Five studies were cost-effectiveness analyses, in which the most common outcome was cost per correct/successful CAD diagnosis.(28
) The remaining seven studies were cost-utility analyses that used cost per quality adjusted life years (QALYs) as their primary outcome.(30
) The time-horizon used across the included studies ranged from 30 days to lifetime and five had 25 years or more of follow-up.(30
) The remaining studies used 18 months (37
), 3 months (39
), and 30 days of analytical time horizon.(33
) Four studies did not report the time-horizon used in their analysis. (28
All included studies evaluated at least one form of ECHO against one of the other remaining selected imaging tests.(28
) The cost-effectiveness of SPECT was studied in nine studies.(28
) Three studies assessed CT angiography in comparison to stress ECHO or MRI.(29
) Cardiac MRI was compared to each of the three other selected imaging tests in two studies.(29
) No full economic analysis between CT angiography and SPECT was found in the published literature.
Literature results for cardiac MRI
The cost-effectiveness of cardiac MRI was assessed against three selected cardiac imaging tests: stress ECHO, SPECT and CT angiography (see ). Two studies evaluated the cost-effectiveness of MRI versus CT angiography, SPECT, or stress ECHO.(29
) In one analysis, cardiac MRI was the alternative with lower costs and worst outcome – and thus not cost-effective – with an ICER per QALY of GBP £13,200 against stress ECHO.(37
Summary incremental cost-effectiveness ratios across selected studies evaluating cardiac MRI
Conclusion of systematic review
Overall, CT angiography was found to be cost-effective or cost-saving in all four comparisons of that technology. Stress ECHO was found cost-effective in eight of the 13 comparisons in which it was evaluated, while SPECT was found cost-effective in three of the 9 comparisons. Cardiac MRI was not found to be cost-effective or cost-saving in any of the four comparisons found.
According to the published economic data from the literature, CT angiography is often found to be cost-effective when compared to other technologies. SPECT and stress ECHO were also found to be cost-effective in several of the comparative studies examined, while cardiac MRI was not cost-effective in any study. Limitations to these conclusions apply, such as the analyses found in the literature evaluated other forms of the selected cardiac imaging tests, which may change the proposed relative cost-effectiveness.
Decision analytic Cost Effectiveness Analysis
This study was designed as a cost effectiveness analysis, with primary results reported as incremental cost per true positive diagnosis or incremental cost per accurate diagnosis.
Two populations were defined for evaluating the cost-effectiveness of an accurate diagnosis (i.e., true positive and true negative diagnoses) of CAD:
- out-patients presenting with stable chest pain; and
- in-patients presenting with acute, unstable chest pain.
The first population was defined as persons presenting with stable chest pain, with an intermediate risk of CAD following physical examination and a graded exercise test, as defined by the American College of Cardiology / American Heart Association 2002 Guideline Update for the Management of Patients with Chronic Stable Angina.(41
) The second population was defined as persons presenting to emergency for acute, unstable chest pain, and who are admitted to hospital, as defined by the American College of Cardiology / American Heart Association 2007 Guidelines for the Management of Patients with Unstable Angina/Non-ST-Elevation Myocardial Infarction.(42
The analytic perspective was that of the Ontario Ministry of Health and Long-Term Care (MOHLTC).
Comparators & Parameter Estimates
The imaging technologies that were compared in the current cost-effectiveness analysis included: 1) CT angiography, 2) stress ECHO, 3) stress ECHO with the availability of contrast medium if needed, 4) cardiac perfusion stress MRI, and 5) attenuation-corrected SPECT. Test characteristic estimates (i.e. specificity, sensitivity, accuracy) for each cardiac imaging technology were obtained from the systematic review and meta-analysis conducted by MAS and the MOHLTC. shows a list of the parameters with corresponding 95% confidence intervals used for both the outpatient and inpatient decision-analytic cost-effectiveness models.
Summary parameter estimates for cardiac MRI tests: sensitivity, specificity; additional days needed to wait for specific cardiac tests; proportion of non-invasive tests considered uninterpretable
Two additional sets of parameters are shown in and described fully in The Relative Cost-effectiveness of Five Non-invasive Cardiac Imaging Technologies for Diagnosing Coronary Artery Disease in Ontario
). The average wait-time for each cardiac imaging test is measured as the additional days needed to wait for a non-invasive test compared to the average wait time for a typical graded exercise stress test (GXT). The proportion of tests deemed uninterpretable by expert opinion is also shown in , with a corresponding range of high and low values. The probability of receiving pharmacological stress versus exercise stress is not shown in the table, but reported here for completeness: approximate values of 30% for the stable, outpatient population and 80% for the unstable, inpatient population.
Time Horizon & Discounting
The time horizon for both decision-analytic models (i.e. for outpatient and inpatient populations) was the time required to determine an accurate, or true positive diagnosis of CAD. As a result, the actual time taken to determine the CAD status of patients may differ across non-invasive test strategies.
provides a simplified illustration of the decision-analytic model structure used for the outpatient and inpatient populations. The following two simplifying assumptions were made for the models:
Decision analytic model used to evaluate the cost-effectiveness of cardiac imaging technologies for the diagnosis of CAD
- When results of the first cardiac imaging test are un-interpretable, a patient will undergo a second cardiac test; the second test will be one of the four remaining tests that were not used as the first.
- Should a second test be required, the type of stress (pharmacological or exercise) that a patient receives for the second test will be the same type of stress as the first.
Various sensitivity analyses were conducted for the outpatient and inpatient populations. First, the prevalence of CAD was varied from 5% to 95% in 5% increments, while all other model estimates were held constant; willingness-to-pay (WTP) was also varied and a range of results were presented. Second, one-way sensitivity analyses were conducted in which selected estimates were varied over plausible ranges, such as sensitivity and specificity estimates, wait times for imaging tests performed in hospital, and costs of CT angiography, ECHO with contrast available and cardiac MRI. A third series of sensitivity analyses was conducted that specifically addressed the issue of possibly unavailable imaging technologies.
Additional details of the sensitivity analyses performed can be found in The Relative Cost-effectiveness of Five Non-invasive Cardiac Imaging Technologies for Diagnosing Coronary Artery Disease in Ontario
) The results of the sensitivity analyses are summarized in the Results and Discussion section below.
Resource Use and Costs
Resource use and costs were derived from Ontario data sources: the OHIP and OCCI administrative databases.(14
) The cost of conducting each cardiac test was calculated as the sum of the test’s respective professional fees and technical fees, as described in the Ontario Schedule of Benefits, are listed in . Note that for ECHO tests with available contrast agent, the cost for the contrast medium was added whenever the contrast was used in the event of uninterpretable ECHO test result. The cost of the contrast medium was estimated as $170 per vial (single use) through consultation with industry experts; only this cost was added to the base test cost of contrast ECHO. In general, where an imaging test result was uninterpretable, an additional cost of follow-up with the patient (physician fee) was incurred, as well as the cost for conducting another cardiac imaging test. For out-patients presenting with stable chest pain, a consultation professional fee of $30.60 (OHIP code A608 for “partial assessment”) was used after an uninterpretable test result (one time cost).
List of cardiac imaging tests and associated OHIP 2009 costs
In the case of patients presenting with acute, unstable chest pain, costs for inpatient hospitalization were also included in the model. The total cost of hospitalization was calculated based on the average wait time for each cardiac imaging test and a cost per diem for each day spent in hospital (for the cardiac MRI wait time, see ). An additional consultation fee was also used only for the inpatient population: $29.20 (OHIP code C602 for “subsequent visit- first five weeks”) was used for each inpatient day (per diem) spent in hospital.
The WTP must be determined by the MOHLTC. For the sensitivity analyses, all reasonable WTP values are presented (see Results and Discussion below) and interpreted at two WTP “anchors”. The two anchors represent the estimated cost of the most expensive non-invasive test considered in our model (cardiac MRI perfusion, $804) and the estimated cost of a coronary angiography ($1,433). These anchors are intended to guide discussion only.
Note that the following points might be useful in determining the WTP:
- An “accurate diagnosis” of CAD can be obtained through a coronary angiography for $1,433, thus one might expect the WTP for an accurate diagnosis through a non-invasive test to resemble this amount. It should be remembered, however, that an “accurate diagnosis” does not include the value or benefit of providing additional diagnostic or prognostic information from either non-invasive imaging and coronary angiography.
- The MOHLTC is currently willing to pay up to $804 for a non-invasive test with less-than-perfect diagnostic accuracy – its willingness to pay for an “accurate diagnosis” from such a test may, therefore, be greater.
- These tests are non-invasive, whereas coronary angiography is invasive. This would presumably be “worth” more (i.e., paying a higher premium); Conversely, these tests carry risks not applicable to coronary angiography, such as increased radiation exposure or adverse reaction to contrast agents.
- These tests are not perfectly accurate – an accurate diagnosis from such a test may be valued less than one from a coronary angiography.
Results and Discussion
The base case results are summarized in and . The analysis revealed that, for both populations (stable outpatients and acute inpatients), cardiac MRI was dominated by CT angiography. That is, it had higher costs and was less effective.
Cost-effectiveness analysis base case results for stable outpatients
Cost-effectiveness analysis base case results for acute inpatients
In sensitivity analyses, MRI was not found to be cost-effective at any reasonable willingness-to-pay for an incremental accurate diagnosis, even after removing CT angiography from the analysis. The present analysis suggests that MRI is not a cost-effective technology for the diagnosis of CAD.
Budget Impact Analysis
The budget impact analysis (BIA) was performed taking the perspective of the MOHLTC and includes both physician and hospital (clinic) costs of non-invasive cardiac imaging tests. Volumes of cardiac tests in Ontario were taken from administrative databases (OHIP, DAD, NACRS) for fiscal years 2004 to 2008 using methodology summarized in The Relative Cost-effectiveness of Five Non-invasive Cardiac Imaging Technologies for Diagnosing Coronary Artery Disease in Ontario
) The following technologies were considered in the current BIA for the diagnosis of CAD: ECHO (including both stress and stress with contrast agent available), nuclear cardiac imaging (including MPI and SPECT tests), cardiac MRI, and CT angiography.
In the current BIA, the effect of moving a certain proportion of the volume of specific tests to another, substitute technology was assessed for various scenarios. These scenarios are presented irrespective of whether a technology was found to be cost-effective and reported as general reference tables.
In summary, cardiac MRI tests were found to be the most expensive of the compared cardiac imaging modalities. When the volume of cardiac MRI tests is shifted to other technologies, all scenarios result in lower projected costs, however, the actual number of tests moved is relatively small. If 25% of cardiac MRI tests is moved to other imaging technologies, ensuing projected costs would be lower: from the largest cost avoidance of about $62.1K per year for stress ECHO testing to the smallest cost avoidance of $28.3K for nuclear cardiac imaging. The largest possible cost avoidance corresponds to replacing 50% of cardiac MRI tests with stress ECHO imaging ($124.2M per year); the smallest cost avoidance occurs by replacing 5% of cardiac MRI tests with nuclear cardiac imaging ($5.7K per year).