|
The cost effectiveness of ACE inibitors as first-line antihypertensive therapy |
Nordmann A J, Krahn M, Logan A G, Naglie G, Detsky A S |
|
|
Record Status 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. Health technology Four strategies of angiotensin-converting enzyme (ACE) inhibitor therapy for the treatment of uncomplicated hypertension were examined.
Strategy 1 was to prescribe ACE inhibitor therapy to all patients.
Strategy 2 was to prescribe conventional therapy (beta-adrenoceptor antagonists or diuretics) to all patients.
Strategy 3 was individualised antihypertensive therapy based on the presence or absence of left ventricular hypertrophy (LVH) on electrocardiography (ECG).
Strategy 4 was individualised antihypertensive therapy based on the presence or absence of LVH on echocardiography (echo).
Economic study type Cost-effectiveness analysis and cost-utility analysis.
Study population The study population comprised a hypothetical cohort of men aged 40 years without cardiovascular co-morbidity requiring antihypertensive drug therapy. No other details were reported.
Setting The setting was primary care. The economic study was carried out in Canada.
Dates to which data relate The data for probabilities and costs were derived from studies published from 1966 to September 2000. The price year was 1999.
Source of effectiveness data The effectiveness evidence was derived from a review of published studies, while some utility values were derived from patients on antihypertensive monotherapy. Several estimates were made to augment the data from these sources.
Modelling A Markov model (Data 3.5, Treeage) was constructed to evaluate the costs and consequences of the four clinical strategies in the treatment of uncomplicated hypertensive men. The time horizon was lifetime.
Outcomes assessed in the review The outcomes estimated from the literature as input parameters for the model were the following probability values:
the prevalence of LVH on echo;
the sensitivity of ECG to detect LVH;
the specificity of ECG to detect LVH;
intolerable adverse effects or lack of efficacy with ACE inhibitors;
intolerable adverse effects or lack of efficacy with conventional therapy;
the incidence of coronary artery disease (CAD), cerebrovascular disease (CVD and congestive heart failure (CHF) per year;
the relative risk of developing CAD given CVD and/or CHF;
the relative risk of developing CVD given CAD;
the relative risk of developing CVD given CHF;
the relative risk of developing CHF given CAD;
the relative risk of developing CHF given CVD;
the relative risk of developing CAD in LVH (echo);
the relative risk of developing CVD in LVH (echo);
the relative risk of developing CHF in LVH (echo);
the effectiveness of ACE inhibitor therapy in LVH;
the relative risk of ECG evidence of LVH;
the age-specific mortality per year;
the case fatality rate from acute CAD per episode;
the case fatality rate from acute CVD;
the disease-specific mortality rate in CHF;
the disease-specific mortality rate in chronic CAD;
the disease-specific mortality rate in chronic CVD;
the relative risk of disease-specific mortality in CAD with pre-existing CHF;
the relative risk of disease-specific mortality in CAD with pre-existing CVD;
the relative risk of disease-specific mortality in CVD with pre-existing CAD and/or CHF;
the utility for CAD;
the utility for CVD; and
the utility for CHF.
Study designs and other criteria for inclusion in the review Sources searched to identify primary studies MEDLINE was searched for 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 Twenty-four primary studies were included in the review. The results of a recent meta-analysis (which included 50 double-blind randomised controlled trials) were also used to estimate the effectiveness of ACE inhibitor therapy.
Methods of combining primary studies Many studies were the only source of evidence, so combination was not required in these instances. In other instances, the method used was not reported.
Investigation of differences between primary studies Results of the review The probability values were too numerous to report in this abstract. The important drivers were as follows:
prevalence of LVH on echo, 0.25 (range: 0.23 - 0.48);
sensitivity of ECG to detect LVH, 0.26 (range: 0.19 - 0.33);
specificity of ECG to detect LVH, 0.96 (range: 0.93 - 0.98);
relative risk of developing CAD given CVD and/or CHF, 1 (no range given);
relative risk of developing CVD given CAD, 2.5 (range: 1.4 - 4.4);
relative risk of developing CVD given CHF, 4.1 (range: 1 - 6);
relative risk of developing CHF given CAD, 4.0 (range: 2.9 - 5.7);
relative risk of developing CHF given CVD, 1.56 (range: 1.07 - 2.27);
relative risk of disease-specific mortality in CAD with pre-existing CHF, 2.7 (range: 1.31 - 3.59);
relative risk of disease-specific mortality in CAD with pre-existing CVD, 1.46 (range: 0.94 - 2.26); and
relative risk of disease-specific mortality in CVD with pre-existing CAD and/or CHF, 1.3 (range: 1 - 3).
Methods used to derive estimates of effectiveness The authors made assumptions where data for the model were not available from the literature.
Estimates of effectiveness and key assumptions The authors assumed the following:
the effectiveness of ACE inhibitor therapy in LVH was 0.05 (range: 0 - 30);
the utility for CAD was 0.90 (range: 0.88 - 0.92);
the utility for CVD was 0.65 (range: 0.4 - 0.9);
the utility for CHF was 0.88 (range: 0.5 - 0.99);
there was no difference in adherence to therapy between the two treatment options;
echo was 100% specific for the diagnosis of LVH; and
the relative risk for the development of cardiovascular complications remained constant in every year.
Measure of benefits used in the economic analysis The two summary benefit measures used were the quality-adjusted life-years (QALYs) and the life-years saved (LYS). The utility values for the patients' current health state were obtained from patients on antihypertensive monotherapy in two clinics in Toronto, Canada. Seventy-five of the 110 patients identified were interviewed. Of these 75 patients, 25 were taking ACE inhibitors, 27 were taking a diuretic and 23 were taking a beta-adrenoceptor antagonist. The utilities were generated using the standard gamble and time trade-off techniques. Mean utility values derived by the standard gamble technique were used for the baseline analysis. The utility for being on conventional antihypertensive therapy was 0.9998 (range: 0.972 - 1), while that for being on ACE inhibitor therapy was 0.9988 (range: 0.972 - 1). The health benefits were discounted at a rate of 5%.
Direct costs The analysis was performed over a time horizon of a patient(s life time, and all future costs were discounted at a rate of 5% per year. The unit costs were reported, but details on resource use were not. As such, the costs and the quantities were not reported separately. The direct costs in the economic analysis were office visits for hypertension, antihypertensive drug treatment, hospitalisation, and ambulatory outpatient treatment for CAD, CVD and CHF.
The costs for office visits to a medical practitioner were calculated from the Fee Schedule of the Ministry of Health in Ontario, Canada. The drug costs were based on information in the Drug Benefit Formulary of the Ontario Ministry of Health. The costs of acute CAD were derived from two recently published studies, one of which analysed cumulative resource use data while the other reported direct medical costs of CAD in the USA. The annual follow-up costs of CAD after the first year were estimated using resource use estimates from another study (1996) and costs from the 1999 Ontario Ministry of Health Drug Benefit Formulary. The costs of CVD were derived from a prevalence-based study of stroke-related costs in 1994/95 and from a model estimating the lifetime costs of stroke in the USA. The total direct costs of CHF were derived from a study estimating the annual costs for inpatient and outpatient care of CHF within the US Health Care and Cost Utilisation Project. The cost for the technical and professional component of ECG and echo were derived from the Schedule of Benefits of the Ontario Ministry of Health.
The cost estimates from older studies were inflated to 1999 Canadian dollars according to the Consumer Price Index for health and personal care.
Statistical analysis of costs The costs were treated deterministically.
Indirect Costs The indirect costs were not considered in the economic evaluation.
Currency All of the costs were calculated in 1999 Canadian dollars (Can$) but were reported in US dollars (US$) according to the 1999 purchasing power parity rate for medical and health care.
Sensitivity analysis Sensitivity analyses were conducted to address the issue of uncertainty in the estimates used in the economic model. A one-way sensitivity analysis was performed on all variables using the ranges reported for the probability and cost variables. The method used to derive these ranges from the literature was not reported.
Estimated benefits used in the economic analysis Health outcomes, as measured by QALYs, were 15.052 for strategy 1, 15.049 for strategy 2, 15.050 for strategy 3 and 15.052 for strategy 4. The incremental benefit (i.e. QALYs gained) when compared with strategy 2 (conventional therapy) was 0.003 for strategy 1 (ACE inhibitors), 0.003 for strategy 3 (ECG) and 0.001 for strategy 4 (echo).
Health outcomes, as measured by LYS, were 15.394 for strategy 1, 15.390 for strategy 2, 15.390 for strategy 3 and 15.393 for strategy 4. The incremental benefit (LYS) when compared with strategy 2 was 0.004 for strategy 1, 0.003 for strategy 3 and 0.000 for strategy 4.
Cost results The total cost was $22,600 for strategy 1, $20,500 for strategy 2, $20,500 for strategy 3 and $21,100 for strategy 4. The differential cost when compared with strategy 2 (conventional therapy) was $2,100 for strategy 1 (ACE inhibitors), nothing for strategy 3 (ECG) and $600 for strategy 4 (echo).
Synthesis of costs and benefits An incremental cost-effectiveness ratio (ICER) was calculated to combine the costs and benefits of the three strategies relative to strategy 2 (conventional therapy). These results were reported in the paper (Table 3). In the decision model, strategy 1 (ACE inhibitors) was more effective, but more costly, with an ICER of $700,000 per QALY gained ($525,000 per LYS). Strategy 4 (echo) was also more effective, but was more costly with an ICER of $200,000 per QALY gained ($200,000 per LYS). Strategy 3 (ECG) was more effective, but of equal cost, with an ICER of $0 per QALY gained (note: with LYS there was no difference in cost or effectiveness so the ICER is not relevant).
However, in the text, the discussion related to the incremental cost-effectiveness of echo and ACE inhibitors where the comparator was the ECG strategy (strategy 3). This is incorrect. Using the results as produced in the table, the ICER for echo (strategy 4) compared with ECG (strategy 3) is $300,000 (not $200,000 as stated), and the ICER for ACE inhibitors (strategy 2) compared with ECG (strategy 3) is $1,050,000 (not $700,000 as stated).
The sensitivity analysis found that the results were sensitive to the difference in annual cost between conventional and ACE inhibitor therapy, and the effectiveness of ACE inhibitor versus conventional therapy in reducing cardiovascular complications in patients with LVH.
When the annual cost of ACE inhibitor therapy exceeded the annual cost of conventional therapy by less than $20, the ICERs of strategy 3 compared with strategies 1 and 4 fell below $100,000/QALY. When conventional antihypertensive therapy was assumed to be as effective as ACE inhibitor therapy, strategy 1 dominated, being marginally more effective and less costly.
Authors' conclusions Prescribing angiotensin-converting enzyme (ACE) inhibitors as first-line antihypertensive therapy in patients without cardiovascular morbidity cannot be recommended at this time, unless the acquisition costs of ACE inhibitors become substantially more attractive.
CRD COMMENTARY - Selection of comparators The authors did not provide any specific justification for the choice of the comparators. However, the range of comparators appears to have been derived from current guidelines that differ in their recommendations for first-line antihypertensive therapy. You should decide if they are widely used interventions in your own setting.
Validity of estimate of measure of effectiveness The analysis of effectiveness was largely based on data derived from the literature. The authors stated that they conducted a systematic MEDLINE search of the literature, but it was unclear what methodology was employed or how the review was conducted. The authors provided a brief commentary on each of the primary studies that supplied evidence. Many of the primary studies represented the only available evidence, thus limiting the need to combine the data. The results of a meta-analysis were also included. Where information was unavailable, the authors made a number of reasonable assumptions. These assumptions, and other limitations concerning the reliability of the data, were addressed in the sensitivity analyses.
Validity of estimate of measure of benefit Both benefit measures used in the economic analysis (QALYs and LYS) are generic, meaning that comparisons across different interventions and diseases can be made. Some utility values were derived from the literature, while others were generated using the standard gamble and time trade-off techniques. These techniques are appropriate, but the authors did not provide any detail of what was involved in eliciting these values. The QALYs and LYS were obtained from the Markov model, which appears to be an appropriate representation of the problem.
Validity of estimate of costs The authors explicitly stated the perspective adopted. It appears that all the relevant cost categories have been included in the analysis. A table of the costs employed in the Markov model was provided. Several of the costs came from fee schedules and a drug formulary, while others were derived from published papers (references given). The costs from early years were inflated using an appropriate index, and converted from Canadian to US dollars using an appropriate conversion factor.
Other issues The model, probabilities and costs were described well, such that it should be possible to replicate the study in other settings. Both the costs and outcomes were discounted at a rate of 5% per year. The sensitivity of this was tested by repeating the analysis without discounting both costs and outcomes. However, it might have been more appropriate, given the controversy surrounding discounting outcomes, to have repeated the analysis with discounted costs and undiscounted outcomes.
The authors noted some limitations of the analysis. First, the need to use circumstantial evidence to calculate the relative risk reductions associated with the use of ACE inhibitors, although they tested this in the sensitivity analysis. Second, there was some confusion about the comparator used in the analysis. From the initial introduction in the paper and the results table it would appear to have been conventional therapy, but the textual comments suggested that the comparator was ECG (although the quoted numbers were those from the table that represented the ICER with conventional therapy as the comparator). However, this makes little difference to the overall conclusion. ACE inhibitors are not a cost-effective alternative to either conventional therapy or risk stratification by ECG and individualised therapy.
Implications of the study The study results suggested that ACE inhibitors as first-line antihypertensive therapy in patients without cardiovascular morbidity do not represent a cost-effective alternative to management strategies for hypertension. ACE inhibitors would only be considered a viable option if they became substantially cheaper.
Bibliographic details Nordmann A J, Krahn M, Logan A G, Naglie G, Detsky A S. The cost effectiveness of ACE inibitors as first-line antihypertensive therapy. PharmacoEconomics 2003; 21(8): 573-585 Indexing Status Subject indexing assigned by NLM MeSH Adrenergic beta-Antagonists /economics /therapeutic use; Angiotensin-Converting Enzyme Inhibitors /economics /therapeutic use; Canada /epidemiology; Cohort Studies; Cost-Benefit Analysis; Diuretics /economics /therapeutic use; Drug Utilization /economics /statistics & Humans; Hypertension /complications /drug therapy; Hypertrophy, Left Ventricular /complications /drug therapy /economics; Markov Chains; Practice Guidelines as Topic; Quality-Adjusted Life Years; Randomized Controlled Trials as Topic; numerical data AccessionNumber 22003008148 Date bibliographic record published 30/11/2005 Date abstract record published 30/11/2005 |
|
|
|