Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Statistical primer: a cost–effectiveness analysis

Statistical primer: a cost–effectiveness analysis Abstract Cost–effectiveness analyses (CEAs) of new treatment strategies are increasingly reported. This can be a part of a clinical trial or as a separate study. Governments and healthcare payers frequently require a CEA to decide whether a new treatment strategy will be reimbursed. CEA is a framework to assess the effectiveness and costs of a new treatment strategy (e.g. a drug or intervention) when compared with a reference strategy. Effectiveness is often measured in life-years or quality-adjusted life-years, whereas costs consist of direct costs (the costs of the treatment), induced costs (downstream costs and cost offsets) and indirect costs. In this statistical primer, the rationale for assessing the economic consequences of new therapies is explained, followed by the fundamental concepts of CEAs, the different types of CEAs and an introduction to interpretation of CEAs. Finally, a real-world example of a CEA is discussed, comparing cost–effectiveness of transcatheter versus surgical aortic valve replacement in patients with severe aortic stenosis at intermediate surgical risk. Cost–effectiveness analysis, Incremental cost–effectiveness ratio, Cost–effectiveness acceptability curve, Willingness to pay threshold, Health economics Healthcare expenditures have been increasing for some years now (Fig. 1). On an average, healthcare expenditures have increased from approximately 9% of the gross domestic product in 2000 to more than 12% in 2016 [1]. Figure 1: View largeDownload slide Healthcare expenditure was reflected as a percentage of the GDP in several Western countries from 2000 to 2016 [1]. GDP: gross domestic product. Figure 1: View largeDownload slide Healthcare expenditure was reflected as a percentage of the GDP in several Western countries from 2000 to 2016 [1]. GDP: gross domestic product. Reasons for this growth include the aging Western populations, the absence of a (truly) free market and increasing administrative costs [2]. In addition, technological innovation is a major driver of increasing healthcare expenditures [3]. These innovations require capital (new devices and catheterization laboratories), labour and education costs. Although in some circumstances, technological innovations may eventually lead to lower costs per patient treated, there may also be an increase in the total number of patients treated, resulting in a net increase in total healthcare costs. For instance, patients with severe aortic stenosis who were previously considered inoperable due to high surgical risk can now undergo transcatheter aortic valve implantation (TAVI). This results in a net increase in the total number of patients who are treated for aortic stenosis (either surgically or using TAVI) [4]. Furthermore, a new therapeutic option for previously inoperable patients might also result in an increased number of diagnoses, ultimately leading to more procedures being performed. In light of these developments and in order to sustain affordability of healthcare systems, it has become vital to demonstrate not only the clinical effectiveness of new treatment strategies but also their economic consequences. In this statistical primer, the fundamental concepts of cost–effectiveness studies will be highlighted and a real-world example will be provided. FUNDAMENTAL CONCEPTS OF COST–EFFECTIVENESS ANALYSIS Costs and effectiveness measures Cost–effectiveness analyses (CEAs) strive to answer 2 questions. First, is a new intervention more effective than the current treatment? Secondly, what are the costs of the new intervention when compared with the current one? Subsequently, CEAs use data concerning effectiveness and costs for a comprehensive answer to the question how economically attractive the new treatment strategy when compared with the reference treatment strategy. The costs to be taken into consideration in CEAs are direct medical costs caused by the new intervention (e.g. a new medical device, drug therapy or magnetic resonance imaging scan), induced or avoided downstream costs (e.g. stroke and its treatment) and indirect costs (e.g. costs of travelling to and from the hospital, costs of informal care and unemployment benefits) [5]. For effectiveness, the most frequently used measure is quality-adjusted life-years (QALYs)—a metric which aims to incorporate both changes in life expectancy and improvements in quality of life (QoL). The most frequently used instrument for QoL is the EuroQol 5 dimensions, which consists of a descriptive system for 5 dimensions of health status and the EuroQol visual analogue scale for the patient’s health perception [6]. QALYs are calculated by multiplying utility weights—a measure of a patient’s preference for his or her health state on a scale from 0 (death) to 1 (excellent health) times the duration lived in that health state (Fig. 2). In this example, Patient 2 lives longer than Patient 1, but QALYs are higher for Patient 1 owing to the higher average utility of this patient. Although QALYs are currently the most accepted way of measuring effectiveness for the purposes of cost–effectiveness studies, their use has also been criticized due to limitations of the various and generic health state questionnaires that are used to estimate utility weights [7]. Figure 2: View largeDownload slide Quality-adjusted life expectancy of 2 fictional patients. Utility weights are measured from 0 (death) to 1 (perfect health). Area under the curve represents the number of QALYs: multiplication of years and utility. QALY: quality-adjusted life-years. Figure 2: View largeDownload slide Quality-adjusted life expectancy of 2 fictional patients. Utility weights are measured from 0 (death) to 1 (perfect health). Area under the curve represents the number of QALYs: multiplication of years and utility. QALY: quality-adjusted life-years. Together, the effectiveness and costs are combined in the incremental cost–effectiveness ratio (ICER) by applying the following formula: ICER=Costnew-CostreferenceQALYnew- QALYreference In this formula, a new treatment strategy is compared with the reference strategy based on costs and effectiveness. The ICER indicates the additional cost required to gain 1 QALY using the new treatment strategy when compared with the reference strategy. Different types of cost–effectiveness analyses Roughly 3 distinct approaches exist for CEAs: model-based, trial-based and hybrid CEAs [8]. Model-based analyses are the most commonly employed. In this approach, mathematical models are used to incorporate fundamental characteristics of the treatment strategies. Markov simulations, often used in model-based CEAs, track cohorts of patients through the different states in which a patient can reside after either treatment strategy (e.g. healthy state, adverse event state or death). Each health state is associated with specific utility, and costs and patients can transition between states based on distinctive likelihoods. A major advantage of the model-based approach is that a wide variety of sources can be used to inform the model, e.g. data from trials and observational studies, QoL registries [for instance, the Center for the Evaluation of Value and Risk in Health (CEVR) [9]] and financial data. After a model has been constructed, sensitivity analyses are generally performed to understand the impact parameter uncertainty (e.g. costs, length of stay and adverse event probabilities) and modelling or structural uncertainty on the results of the analysis [10, 11]. In addition, Monte Carlo simulations are often performed numerous times to calculate the ICER along with its uncertainty. Uncertainty is depicted by the cloud of bootstrap replicates in Fig. 3 [12, 13]. Figure 3: View largeDownload slide A distribution of incremental costs and QALYs gained for the SYNTAX trial on CABG versus PCI as a lifetime projection based on bootstrap replication plotted in a cost–effectiveness plane. The difference in costs is presented on the y-axis, whereas the difference in QALYs is shown on the x-axis. The black dot represents the mean value of costs per 0.307 QALY gained for CABG when compared with PCI. Reprinted with permission from Cohen et al. [13]. CABG: coronary artery bypass grafting; PCI: percutaneous coronary intervention; QALY: quality-adjusted life-years. Figure 3: View largeDownload slide A distribution of incremental costs and QALYs gained for the SYNTAX trial on CABG versus PCI as a lifetime projection based on bootstrap replication plotted in a cost–effectiveness plane. The difference in costs is presented on the y-axis, whereas the difference in QALYs is shown on the x-axis. The black dot represents the mean value of costs per 0.307 QALY gained for CABG when compared with PCI. Reprinted with permission from Cohen et al. [13]. CABG: coronary artery bypass grafting; PCI: percutaneous coronary intervention; QALY: quality-adjusted life-years. Dedicated software exists for Markov and/or Monte Carlo simulations including TreeAge (TreeAge Software, Williamstown, MA, USA) and R (R Development Core Team), with specific CRAN packages, for example, Bayesian cost–effectiveness analysis or following a manual [14]. Modelling can also be performed using specific Macros Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) [15]. In the trial-based approach, an economic evaluation of a new treatment strategy is performed alongside a clinical trial. The high-quality data acquirement and randomization are major benefits of this type of analysis. On the other hand, patients included in trials might not always reflect the real world, leading to limited generalizability. A further limitation of this approach is the strict (and often relatively short) time horizon, which may be insufficient to fully capture the costs or benefits of a given treatment strategy. Finally, in the hybrid approach, the concepts of trial-based and model-based CEAs are blended together such that the high-quality data of a trial are combined with external data to analyse long-term costs and effects. The ability to extend the results of a clinical trial to reflect a lifetime horizon is an important benefit of this approach as cost–effectiveness analyses should generally consider a lifetime horizon. An interesting example of the importance of an adequate time horizon on the results of a cost–effectiveness analysis can be found in the Synergy between percutaneous coronary intervention (PCI) with TAXUS and Cardiac Surgery (SYNTAX) trial, a randomized trial comparing coronary artery bypass grafting (CABG) versus PCI using drug-eluting stents (DESs) for revascularization of patients with 3-vessel and left main coronary artery disease [13, 16]. The initial 1-year cost–effectiveness analysis, which was a trial-based CEA, showed that PCI was economically more attractive than CABG, as total costs of CABG were higher ($5693) than DES-PCI, while QALYs gained were higher for DES-PCI when compared with CABG (0.82 QALYs and 0.80 QALYs, respectively), making DES-PCI economically dominant. However, a hybrid-based CEA based on 5-year trial data with lifetime extrapolations showed spectacularly different results; although CABG was more expensive than DES-PCI, it was also more effective with a favourable ICER. INTERPRETATION OF COST–EFFECTIVENESS ANALYSIS An important topic to consider in CEAs is the analytic perspective. In 1996, the US Panel on Cost–Effectiveness in Health and Medicine recommended the use of a ‘societal perspective’ [5]. CEAs with a societal perspective strive to answer if a new treatment strategy is economically attractive for the society and not only for individual parties. This perspective is the most comprehensive and aims to incorporate all costs (direct, indirect and avoided or induced) associated with the use of a new treatment strategy. The healthcare perspective is another frequently used perspective. In this perspective, the current and future healthcare costs (paid by any party including health insurances, patients, etc.) are considered, but other costs (e.g. sick leave) are not taken into account. Although papers often report to use the societal perspective in CEAs, the actual perspective being used is the healthcare perspective. In total, 40.6% of the CEAs stated that they used the societal perspective, while the reviewers found this to be true for 29.3%, whereas the use of the healthcare perspective was reported to be 32.8%, while this was actually 68.6% [17]. Cost–effectiveness plane Once a CEA has been performed, the results can be visualized in the cost–effectiveness plane (Fig. 4). In this graph, the lower right quadrant reflects therapies that are ‘economically dominant’, meaning that they are both more effective and are associated with lower costs when compared with the reference intervention (Fig. 4, letter B). In contrast, the upper left quadrant indicates those therapies that are ‘economically dominated’ (Fig. 4, letter D)—i.e. treatments that are both more costly and less effective than the reference strategy. When a therapy falls into 1 of these 2 quadrants, policy decisions are straightforward; economically dominant strategies are always preferred, and economically dominant strategies are never preferred. Figure 4: View largeDownload slide Cost–effectiveness plane. The upper left quadrant reflects interventions that are more expensive and less effective, whereas the lower right quadrant reflects studies with higher effectiveness combined at lower costs. Most new treatment strategies will fit in the upper right quadrant (more effective and more expensive), where ICERs are considered with relation to cost–effectiveness thresholds: new treatment strategies that fall into the red-shaded part of the quadrant are not considered economically attractive, whereas new treatment strategies that fall in the green-shaded part are considered economically attractive. Treatment strategies are indicated as A–D. A, increased effectiveness, with increased cost, below cost–effectiveness threshold and, therefore, considered economically attractive. B, cost saving and increased effectiveness: dominant treatment strategy. C, cost-saving strategy: marginally less effective but also significantly less expensive. D, less effective while more expensive: dominated. ICER: incremental cost–effectiveness ratio. Figure 4: View largeDownload slide Cost–effectiveness plane. The upper left quadrant reflects interventions that are more expensive and less effective, whereas the lower right quadrant reflects studies with higher effectiveness combined at lower costs. Most new treatment strategies will fit in the upper right quadrant (more effective and more expensive), where ICERs are considered with relation to cost–effectiveness thresholds: new treatment strategies that fall into the red-shaded part of the quadrant are not considered economically attractive, whereas new treatment strategies that fall in the green-shaded part are considered economically attractive. Treatment strategies are indicated as A–D. A, increased effectiveness, with increased cost, below cost–effectiveness threshold and, therefore, considered economically attractive. B, cost saving and increased effectiveness: dominant treatment strategy. C, cost-saving strategy: marginally less effective but also significantly less expensive. D, less effective while more expensive: dominated. ICER: incremental cost–effectiveness ratio. However, new interventions are often in neither of these 2 categories and will often require further analysis. If a new treatment strategy occupies the lower left quadrant, its effectiveness is less than the reference strategy. From an economic perspective, such a strategy might be considered if it is marginally less clinically effective than the reference, while the economic savings are substantial (Fig. 4, letter C); however, this situation is exceptional in many Western countries. Most new technologies, however, will fit in the upper right quadrant. These technologies, which are more effective, but also more expensive, must then be considered with relation to a cost–effectiveness threshold. A CE threshold is defined as the amount of money that is considered acceptable to gain 1 QALY. Thresholds are not uniformly defined but are often related to the economic wealth of a country. For instance, the acceptable thresholds for the ICER in the USA are substantially higher than those in Brazil. The threshold is formed by a balance between the ability-to-pay and willingness-to-pay with societal and political factors playing a role [18]. Cost–effectiveness thresholds In the USA, the cost–effectiveness threshold is traditionally set at $50 000/QALY. This value is derived from the 1970s, where the cost per QALY for patients with end-stage renal failure who received dialysis was approximately $50 000 [19]. More recently, a statement by the American College of Cardiology and American Heart Association on cost–effectiveness considered an ICER of <$50 000 to represent a high value and an ICER between $50 000 and $150 000 per QALY to represent a intermediate value [20]. The National Institute for Health and Care Excellence in the UK considers an ICER of <£20 000 per QALY to be economically attractive [21]. If the ICER is higher, an advisory board will carefully consider whether a new treatment strategy is supported or not. The Dutch Council for Public Health and Care proposes a CEA threshold that depends on the disease burden of a patient and ranges between €10 000 and €80 000 per QALY [22]. Sensitivity analyses Both trial-based and model-based CEAs harbour a number of uncertain parameters. To assess this uncertainty, sensitivity analyses are performed, which can be broken down in univariable and probabilistic sensitivity analyses. A univariable sensitivity analysis involves varying 1 parameter that carries uncertainty across a range of plausible values to assess the impact of that particular parameter on the results. In a probabilistic sensitivity analysis, all variables in the model are varied, allowing the assessment of uncertainty in all parameters simultaneously. For trial-based CEAs, a bootstrap method is generally used to assess the uncertainty in all parameters. Patients are randomly selected from a random dataset (with replacement), and the analysis is replicated on this new dataset. This procedure is then repeated a large number of times (e.g. 1000), leading to 1000 estimates of the ICER (e.g. Fig. 3) [23]. Both the probabilistic sensitivity analysis and bootstrap method procedures provide insight in the uncertainty of costs and effectiveness outcomes. The fraction of replicates of the resulting ICERs fitting into the different quadrants of the CE plane and the proportion of ICERs below or above a certain willingness-to-pay threshold can be calculated. Finally, as willingness-to-pay thresholds are not set in stone, the data from bootstrap replication can be used to produce a cost–effectiveness acceptability curve, which demonstrates the probability that a given treatment is economically attractive as a function of the cost–effectiveness threshold (Fig. 5) [24]. Figure 5: View largeDownload slide Cost–effectiveness acceptability curve. The y-axis represents the percentage of bootstraps that had an acceptable incremental cost–effectiveness ratio, and the x-axis represents the costs in Euros (€) per QALY gained. The 3 curves have different cost–effectiveness probabilities at the same costs in euros per QALY. QALY: quality-adjusted life-years. Figure 5: View largeDownload slide Cost–effectiveness acceptability curve. The y-axis represents the percentage of bootstraps that had an acceptable incremental cost–effectiveness ratio, and the x-axis represents the costs in Euros (€) per QALY gained. The 3 curves have different cost–effectiveness probabilities at the same costs in euros per QALY. QALY: quality-adjusted life-years. REAL-WORLD EXAMPLE: TRANSCATHETER VERSUS SURGICAL AORTIC VALVE REPLACEMENT One area of considerable interest for economic analyses has been the use of TAVI as an alternative to surgical aortic valve replacement (SAVR) to treat patients with severe aortic stenosis. Reasons for this interest include the large population of patients who may be candidates for either procedure and the substantial cost of the transcatheter valve prosthesis. The CoreValve US High-Risk Pivotal trial enrolled 795 patients who were randomized to TAVI or SAVR. For the CEA, detailed resource utilization (e.g. indirect costs) and hospital fees were available from randomization until 12 months later or death. QoL data were available for 2 years (1, 6, 12 and 24 months) by means of the EuroQol 5 dimensions questionnaire. The data collected during the trial were then used to predict survival, QALYs and costs (from a healthcare system perspective) over a lifetime horizon. The study showed that TAVI, when compared with SAVR, resulted in a 0.32-year increase in quality-adjusted life expectancy, with an increase in lifetime costs of approximately $18 000, resulting in an ICER of just above $55 000 [25]. Interestingly, a more recent CEA based on an intermediate-risk population showed TAVI to be cost saving when compared with SAVR, with a decrease in costs between $7900 and $9700 per QALY depending on the TAVI valve chosen [26]. These results indicate that TAVI could be considered an economically dominant treatment strategy when compared with SAVR—at least for the intermediate-risk population. A few factors could have contributed to the fact that TAVI was economically dominant in the intermediate-risk population: a 6.5-day reduction in length of stay for TAVI patients, compared to SAVR patients. This reduction in length of stay is more pronounced than in the High-Risk Pivotal trial where a difference of 4.4 days was observed when compared with SAVR [25]. Furthermore, a significant drop in rehospitalizations for TAVI and >50% reduction in rehabilitation days were observed. All 3 factors contribute to lower costs after TAVI, thereby closing the cost difference gap found in the High-Risk Pivotal trial. LIMITATIONS AND FUTURE PERSPECTIVES CEAs are an increasingly important form of research to aid and facilitate the sustainability of healthcare systems. Although CEAs are widely accepted, some have questioned the methodology. For instance, the willingness to pay threshold (often set at $50 000/QALY) is frequently questioned. However, we argue that the willingness-to-pay threshold should be seen as a guidance and not as a strict cut-point above which a treatment strategy is deemed to be too expensive. Moreover, the applicability of CEAs is questioned, as it would ignore that what is considered good on average does not necessarily have to be appropriate for individuals [27]. Others accept the inherent limitations of CEAs and see its value if used appropriately [28]. Finally, the quality of CEAs greatly depends on the quality and generalizability of the data used as input. Despite its limitations, CEAs offer us the best opportunity to evaluate not only the possible greater effectiveness of new treatment strategies but also its value for money. The development of new costly treatment strategies will continue, whereas healthcare budgets increasingly threaten other parts of the economy. This stresses the important role CEAs not only have now but also in the future. Conflict of interest: Christiaan F.J. Antonides has no conflict of interest. David J. Cohen discloses a financial relationship with Edwards Lifesciences, Medtronic, and Boston Scientific. Ruben L.J. Osnabrugge has no conflict of interest. REFERENCES 1 Health Expenditure and Financing: Current Expenditure on Health (All Functions) Measured by the Share of Gross Domestic Product (2000-2016). Organisation for Economic Co-operation and Development (OECD ), 2017 . 2 Bodenheimer T. High and rising health care costs. Part 1: seeking an explanation . Ann Intern Med 2005 ; 142 : 847 – 54 . Google Scholar CrossRef Search ADS PubMed 3 Bodenheimer T. High and rising health care costs. Part 2: technologic innovation . Ann Intern Med 2005 ; 142 : 932 – 7 . Google Scholar CrossRef Search ADS PubMed 4 Osnabrugge RL , Mylotte D , Head SJ , Van Mieghem NM , Nkomo VT , LeReun CM et al. Aortic stenosis in the elderly: disease prevalence and number of candidates for transcatheter aortic valve replacement: a meta-analysis and modeling study . J Am Coll Cardiol 2013 ; 62 : 1002 – 12 . Google Scholar CrossRef Search ADS PubMed 5 Weinstein MC , Siegel JE , Gold MR , Kamlet MS , Russell LB. Recommendations of the panel on cost-effectiveness in health and medicine . JAMA 1996 ; 276 : 1253 – 8 . Google Scholar CrossRef Search ADS PubMed 6 Oppe M , Devlin NJ , Szende A. EQ-5D Value Sets: Inventory, Comparative Review and User Guide . Dordrecht, The Netherlands : Springer , 2007 . Google Scholar CrossRef Search ADS 7 Chapman RH , Berger M , Weinstein MC , Weeks JC , Goldie S , Neumann PJ. When does quality-adjusting life-years matter in cost-effectiveness analysis? Health Econ 2004 ; 13 : 429 – 36 . Google Scholar CrossRef Search ADS PubMed 8 Cohen DJ , Reynolds MR. Interpreting the results of cost-effectiveness studies . J Am Coll Cardiol 2008 ; 52 : 2119 – 26 . Google Scholar CrossRef Search ADS PubMed 9 Cost-Effectiveness Analysis Registry (CEA-Registry) . http://healtheconomics.tuftsmedicalcenter.org/cear4/SearchingtheCEARegistry/SearchtheCEARegistry.aspx (12 December 2017, date last accessed). 10 Jain R , Grabner M , Onukwugha E. Sensitivity analysis in cost-effectiveness studies: from guidelines to practice . Pharmacoeconomics 2011 ; 29 : 297 – 314 . Google Scholar CrossRef Search ADS PubMed 11 Strong M , Pilgrim H , Oakley J , Chilcott J. Structural uncertainty in health economic decision models. ScHARR Occasional Paper, ISBN: 1 900752 27 1, 2009 . 12 Sonnenberg FA , Beck JR. Markov models in medical decision making: a practical guide . Med Decis Making 1993 ; 13 : 322 – 38 . Google Scholar CrossRef Search ADS PubMed 13 Cohen DJ , Osnabrugge RL , Magnuson EA , Wang K , Li H , Chinnakondepalli K et al. Cost-effectiveness of percutaneous coronary intervention with drug-eluting stents versus bypass surgery for patients with 3-vessel or left main coronary artery disease: final results from the Synergy Between Percutaneous Coronary Intervention With TAXUS and Cardiac Surgery (SYNTAX) trial . Circulation 2014 ; 130 : 1146 – 57 . Google Scholar CrossRef Search ADS PubMed 14 Williams C , Lewsey JD , Briggs AH , Mackay DF. Cost-effectiveness analysis in R using a multi-state modeling survival analysis framework: a tutorial . Med Decis Making 2017 ; 37 : 340 – 52 . Google Scholar CrossRef Search ADS PubMed 15 Watt M , Mealing S , Eaton J , Piazza N , Moat N , Brasseur P et al. Cost-effectiveness of transcatheter aortic valve replacement in patients ineligible for conventional aortic valve replacement . Heart 2012 ; 98 : 370 – 6 . Google Scholar CrossRef Search ADS PubMed 16 Cohen DJ , Lavelle TA , Van Hout B , Li H , Lei Y , Robertus K et al. Economic outcomes of percutaneous coronary intervention with drug-eluting stents versus bypass surgery for patients with left main or three-vessel coronary artery disease: one-year results from the SYNTAX trial . Catheter Cardiovasc Interv 2012 ; 79 : 198 – 209 . Google Scholar CrossRef Search ADS PubMed 17 Neumann PJ. Costing and perspective in published cost-effectiveness analysis . Med Care 2009 ; 47 : S28 – 32 . Google Scholar CrossRef Search ADS PubMed 18 Schwarzer R , Rochau U , Saverno K , Jahn B , Bornschein B , Muehlberger N et al. Systematic overview of cost-effectiveness thresholds in ten countries across four continents . J Comp Eff Res 2015 ; 4 : 485 – 504 . Google Scholar CrossRef Search ADS PubMed 19 Neumann PJ , Cohen JT , Weinstein MC. Updating cost-effectiveness—the curious resilience of the $50,000-per-QALY threshold . N Engl J Med 2014 ; 371 : 796 – 7 . Google Scholar CrossRef Search ADS PubMed 20 Anderson JL , Heidenreich PA , Barnett PG , Creager MA , Fonarow GC , Gibbons RJ et al. ACC/AHA statement on cost/value methodology in clinical practice guidelines and performance measures: a report of the American College of Cardiology/American Heart Association Task Force on Performance Measures and Task Force on Practice Guidelines . J Am Coll Cardiol 2014 ; 63 : 2304 – 22 . Google Scholar CrossRef Search ADS PubMed 21 McCabe C , Claxton K , Culyer AJ. The NICE cost-effectiveness threshold: what it is and what that means . Pharmacoeconomics 2008 ; 26 : 733 – 44 . Google Scholar CrossRef Search ADS PubMed 22 Cost-effectiveness in practice. National Health Care Institute, The Netherlands, 26 June 2015. 23 Campbell MK , Torgerson DJ. Bootstrapping: estimating confidence intervals for cost-effectiveness ratios . QJM 1999 ; 92 : 177 – 82 . Google Scholar CrossRef Search ADS PubMed 24 Al MJ. Cost-effectiveness acceptability curves revisited . Pharmacoeconomics 2013 ; 31 : 93 – 100 . Google Scholar CrossRef Search ADS PubMed 25 Reynolds MR , Lei Y , Wang K , Chinnakondepalli K , Vilain KA , Magnuson EA et al. Cost-effectiveness of transcatheter aortic valve replacement with a self-expanding prosthesis versus surgical aortic valve replacement . J Am Coll Cardiol 2016 ; 67 : 29 – 38 . Google Scholar CrossRef Search ADS PubMed 26 Cohen DJ. Cost-effectiveness of transcatheter vs. surgical aortic valve replacement in intermediate risk patients. Results from the PARTNER 2A and SAPIEN 3 intermediate risk trials. Presented at TCT 2017, Denver, Colorado, USA, 2017 . 27 Diamond GA , Kaul S. Cost, effectiveness, and cost-effectiveness . Circ Cardiovasc Qual Outcomes 2009 ; 2 : 49 . Google Scholar CrossRef Search ADS PubMed 28 Weintraub WS , Cohen DJ. The limits of cost-effectiveness analysis . Circ Cardiovasc Qual Outcomes 2009 ; 2 : 55. Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Journal of Cardio-Thoracic Surgery Oxford University Press

Statistical primer: a cost–effectiveness analysis

Download PDF
5 pages

Loading next page...
1
 
/lp/ou_press/statistical-primer-a-cost-effectiveness-analysis-Pu1agXDIQ4

References (25)

Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ISSN
1010-7940
eISSN
1873-734X
DOI
10.1093/ejcts/ezy187
Publisher site
See Article on Publisher Site

Abstract

Abstract Cost–effectiveness analyses (CEAs) of new treatment strategies are increasingly reported. This can be a part of a clinical trial or as a separate study. Governments and healthcare payers frequently require a CEA to decide whether a new treatment strategy will be reimbursed. CEA is a framework to assess the effectiveness and costs of a new treatment strategy (e.g. a drug or intervention) when compared with a reference strategy. Effectiveness is often measured in life-years or quality-adjusted life-years, whereas costs consist of direct costs (the costs of the treatment), induced costs (downstream costs and cost offsets) and indirect costs. In this statistical primer, the rationale for assessing the economic consequences of new therapies is explained, followed by the fundamental concepts of CEAs, the different types of CEAs and an introduction to interpretation of CEAs. Finally, a real-world example of a CEA is discussed, comparing cost–effectiveness of transcatheter versus surgical aortic valve replacement in patients with severe aortic stenosis at intermediate surgical risk. Cost–effectiveness analysis, Incremental cost–effectiveness ratio, Cost–effectiveness acceptability curve, Willingness to pay threshold, Health economics Healthcare expenditures have been increasing for some years now (Fig. 1). On an average, healthcare expenditures have increased from approximately 9% of the gross domestic product in 2000 to more than 12% in 2016 [1]. Figure 1: View largeDownload slide Healthcare expenditure was reflected as a percentage of the GDP in several Western countries from 2000 to 2016 [1]. GDP: gross domestic product. Figure 1: View largeDownload slide Healthcare expenditure was reflected as a percentage of the GDP in several Western countries from 2000 to 2016 [1]. GDP: gross domestic product. Reasons for this growth include the aging Western populations, the absence of a (truly) free market and increasing administrative costs [2]. In addition, technological innovation is a major driver of increasing healthcare expenditures [3]. These innovations require capital (new devices and catheterization laboratories), labour and education costs. Although in some circumstances, technological innovations may eventually lead to lower costs per patient treated, there may also be an increase in the total number of patients treated, resulting in a net increase in total healthcare costs. For instance, patients with severe aortic stenosis who were previously considered inoperable due to high surgical risk can now undergo transcatheter aortic valve implantation (TAVI). This results in a net increase in the total number of patients who are treated for aortic stenosis (either surgically or using TAVI) [4]. Furthermore, a new therapeutic option for previously inoperable patients might also result in an increased number of diagnoses, ultimately leading to more procedures being performed. In light of these developments and in order to sustain affordability of healthcare systems, it has become vital to demonstrate not only the clinical effectiveness of new treatment strategies but also their economic consequences. In this statistical primer, the fundamental concepts of cost–effectiveness studies will be highlighted and a real-world example will be provided. FUNDAMENTAL CONCEPTS OF COST–EFFECTIVENESS ANALYSIS Costs and effectiveness measures Cost–effectiveness analyses (CEAs) strive to answer 2 questions. First, is a new intervention more effective than the current treatment? Secondly, what are the costs of the new intervention when compared with the current one? Subsequently, CEAs use data concerning effectiveness and costs for a comprehensive answer to the question how economically attractive the new treatment strategy when compared with the reference treatment strategy. The costs to be taken into consideration in CEAs are direct medical costs caused by the new intervention (e.g. a new medical device, drug therapy or magnetic resonance imaging scan), induced or avoided downstream costs (e.g. stroke and its treatment) and indirect costs (e.g. costs of travelling to and from the hospital, costs of informal care and unemployment benefits) [5]. For effectiveness, the most frequently used measure is quality-adjusted life-years (QALYs)—a metric which aims to incorporate both changes in life expectancy and improvements in quality of life (QoL). The most frequently used instrument for QoL is the EuroQol 5 dimensions, which consists of a descriptive system for 5 dimensions of health status and the EuroQol visual analogue scale for the patient’s health perception [6]. QALYs are calculated by multiplying utility weights—a measure of a patient’s preference for his or her health state on a scale from 0 (death) to 1 (excellent health) times the duration lived in that health state (Fig. 2). In this example, Patient 2 lives longer than Patient 1, but QALYs are higher for Patient 1 owing to the higher average utility of this patient. Although QALYs are currently the most accepted way of measuring effectiveness for the purposes of cost–effectiveness studies, their use has also been criticized due to limitations of the various and generic health state questionnaires that are used to estimate utility weights [7]. Figure 2: View largeDownload slide Quality-adjusted life expectancy of 2 fictional patients. Utility weights are measured from 0 (death) to 1 (perfect health). Area under the curve represents the number of QALYs: multiplication of years and utility. QALY: quality-adjusted life-years. Figure 2: View largeDownload slide Quality-adjusted life expectancy of 2 fictional patients. Utility weights are measured from 0 (death) to 1 (perfect health). Area under the curve represents the number of QALYs: multiplication of years and utility. QALY: quality-adjusted life-years. Together, the effectiveness and costs are combined in the incremental cost–effectiveness ratio (ICER) by applying the following formula: ICER=Costnew-CostreferenceQALYnew- QALYreference In this formula, a new treatment strategy is compared with the reference strategy based on costs and effectiveness. The ICER indicates the additional cost required to gain 1 QALY using the new treatment strategy when compared with the reference strategy. Different types of cost–effectiveness analyses Roughly 3 distinct approaches exist for CEAs: model-based, trial-based and hybrid CEAs [8]. Model-based analyses are the most commonly employed. In this approach, mathematical models are used to incorporate fundamental characteristics of the treatment strategies. Markov simulations, often used in model-based CEAs, track cohorts of patients through the different states in which a patient can reside after either treatment strategy (e.g. healthy state, adverse event state or death). Each health state is associated with specific utility, and costs and patients can transition between states based on distinctive likelihoods. A major advantage of the model-based approach is that a wide variety of sources can be used to inform the model, e.g. data from trials and observational studies, QoL registries [for instance, the Center for the Evaluation of Value and Risk in Health (CEVR) [9]] and financial data. After a model has been constructed, sensitivity analyses are generally performed to understand the impact parameter uncertainty (e.g. costs, length of stay and adverse event probabilities) and modelling or structural uncertainty on the results of the analysis [10, 11]. In addition, Monte Carlo simulations are often performed numerous times to calculate the ICER along with its uncertainty. Uncertainty is depicted by the cloud of bootstrap replicates in Fig. 3 [12, 13]. Figure 3: View largeDownload slide A distribution of incremental costs and QALYs gained for the SYNTAX trial on CABG versus PCI as a lifetime projection based on bootstrap replication plotted in a cost–effectiveness plane. The difference in costs is presented on the y-axis, whereas the difference in QALYs is shown on the x-axis. The black dot represents the mean value of costs per 0.307 QALY gained for CABG when compared with PCI. Reprinted with permission from Cohen et al. [13]. CABG: coronary artery bypass grafting; PCI: percutaneous coronary intervention; QALY: quality-adjusted life-years. Figure 3: View largeDownload slide A distribution of incremental costs and QALYs gained for the SYNTAX trial on CABG versus PCI as a lifetime projection based on bootstrap replication plotted in a cost–effectiveness plane. The difference in costs is presented on the y-axis, whereas the difference in QALYs is shown on the x-axis. The black dot represents the mean value of costs per 0.307 QALY gained for CABG when compared with PCI. Reprinted with permission from Cohen et al. [13]. CABG: coronary artery bypass grafting; PCI: percutaneous coronary intervention; QALY: quality-adjusted life-years. Dedicated software exists for Markov and/or Monte Carlo simulations including TreeAge (TreeAge Software, Williamstown, MA, USA) and R (R Development Core Team), with specific CRAN packages, for example, Bayesian cost–effectiveness analysis or following a manual [14]. Modelling can also be performed using specific Macros Microsoft Excel (Microsoft Corporation, Redmond, WA, USA) [15]. In the trial-based approach, an economic evaluation of a new treatment strategy is performed alongside a clinical trial. The high-quality data acquirement and randomization are major benefits of this type of analysis. On the other hand, patients included in trials might not always reflect the real world, leading to limited generalizability. A further limitation of this approach is the strict (and often relatively short) time horizon, which may be insufficient to fully capture the costs or benefits of a given treatment strategy. Finally, in the hybrid approach, the concepts of trial-based and model-based CEAs are blended together such that the high-quality data of a trial are combined with external data to analyse long-term costs and effects. The ability to extend the results of a clinical trial to reflect a lifetime horizon is an important benefit of this approach as cost–effectiveness analyses should generally consider a lifetime horizon. An interesting example of the importance of an adequate time horizon on the results of a cost–effectiveness analysis can be found in the Synergy between percutaneous coronary intervention (PCI) with TAXUS and Cardiac Surgery (SYNTAX) trial, a randomized trial comparing coronary artery bypass grafting (CABG) versus PCI using drug-eluting stents (DESs) for revascularization of patients with 3-vessel and left main coronary artery disease [13, 16]. The initial 1-year cost–effectiveness analysis, which was a trial-based CEA, showed that PCI was economically more attractive than CABG, as total costs of CABG were higher ($5693) than DES-PCI, while QALYs gained were higher for DES-PCI when compared with CABG (0.82 QALYs and 0.80 QALYs, respectively), making DES-PCI economically dominant. However, a hybrid-based CEA based on 5-year trial data with lifetime extrapolations showed spectacularly different results; although CABG was more expensive than DES-PCI, it was also more effective with a favourable ICER. INTERPRETATION OF COST–EFFECTIVENESS ANALYSIS An important topic to consider in CEAs is the analytic perspective. In 1996, the US Panel on Cost–Effectiveness in Health and Medicine recommended the use of a ‘societal perspective’ [5]. CEAs with a societal perspective strive to answer if a new treatment strategy is economically attractive for the society and not only for individual parties. This perspective is the most comprehensive and aims to incorporate all costs (direct, indirect and avoided or induced) associated with the use of a new treatment strategy. The healthcare perspective is another frequently used perspective. In this perspective, the current and future healthcare costs (paid by any party including health insurances, patients, etc.) are considered, but other costs (e.g. sick leave) are not taken into account. Although papers often report to use the societal perspective in CEAs, the actual perspective being used is the healthcare perspective. In total, 40.6% of the CEAs stated that they used the societal perspective, while the reviewers found this to be true for 29.3%, whereas the use of the healthcare perspective was reported to be 32.8%, while this was actually 68.6% [17]. Cost–effectiveness plane Once a CEA has been performed, the results can be visualized in the cost–effectiveness plane (Fig. 4). In this graph, the lower right quadrant reflects therapies that are ‘economically dominant’, meaning that they are both more effective and are associated with lower costs when compared with the reference intervention (Fig. 4, letter B). In contrast, the upper left quadrant indicates those therapies that are ‘economically dominated’ (Fig. 4, letter D)—i.e. treatments that are both more costly and less effective than the reference strategy. When a therapy falls into 1 of these 2 quadrants, policy decisions are straightforward; economically dominant strategies are always preferred, and economically dominant strategies are never preferred. Figure 4: View largeDownload slide Cost–effectiveness plane. The upper left quadrant reflects interventions that are more expensive and less effective, whereas the lower right quadrant reflects studies with higher effectiveness combined at lower costs. Most new treatment strategies will fit in the upper right quadrant (more effective and more expensive), where ICERs are considered with relation to cost–effectiveness thresholds: new treatment strategies that fall into the red-shaded part of the quadrant are not considered economically attractive, whereas new treatment strategies that fall in the green-shaded part are considered economically attractive. Treatment strategies are indicated as A–D. A, increased effectiveness, with increased cost, below cost–effectiveness threshold and, therefore, considered economically attractive. B, cost saving and increased effectiveness: dominant treatment strategy. C, cost-saving strategy: marginally less effective but also significantly less expensive. D, less effective while more expensive: dominated. ICER: incremental cost–effectiveness ratio. Figure 4: View largeDownload slide Cost–effectiveness plane. The upper left quadrant reflects interventions that are more expensive and less effective, whereas the lower right quadrant reflects studies with higher effectiveness combined at lower costs. Most new treatment strategies will fit in the upper right quadrant (more effective and more expensive), where ICERs are considered with relation to cost–effectiveness thresholds: new treatment strategies that fall into the red-shaded part of the quadrant are not considered economically attractive, whereas new treatment strategies that fall in the green-shaded part are considered economically attractive. Treatment strategies are indicated as A–D. A, increased effectiveness, with increased cost, below cost–effectiveness threshold and, therefore, considered economically attractive. B, cost saving and increased effectiveness: dominant treatment strategy. C, cost-saving strategy: marginally less effective but also significantly less expensive. D, less effective while more expensive: dominated. ICER: incremental cost–effectiveness ratio. However, new interventions are often in neither of these 2 categories and will often require further analysis. If a new treatment strategy occupies the lower left quadrant, its effectiveness is less than the reference strategy. From an economic perspective, such a strategy might be considered if it is marginally less clinically effective than the reference, while the economic savings are substantial (Fig. 4, letter C); however, this situation is exceptional in many Western countries. Most new technologies, however, will fit in the upper right quadrant. These technologies, which are more effective, but also more expensive, must then be considered with relation to a cost–effectiveness threshold. A CE threshold is defined as the amount of money that is considered acceptable to gain 1 QALY. Thresholds are not uniformly defined but are often related to the economic wealth of a country. For instance, the acceptable thresholds for the ICER in the USA are substantially higher than those in Brazil. The threshold is formed by a balance between the ability-to-pay and willingness-to-pay with societal and political factors playing a role [18]. Cost–effectiveness thresholds In the USA, the cost–effectiveness threshold is traditionally set at $50 000/QALY. This value is derived from the 1970s, where the cost per QALY for patients with end-stage renal failure who received dialysis was approximately $50 000 [19]. More recently, a statement by the American College of Cardiology and American Heart Association on cost–effectiveness considered an ICER of <$50 000 to represent a high value and an ICER between $50 000 and $150 000 per QALY to represent a intermediate value [20]. The National Institute for Health and Care Excellence in the UK considers an ICER of <£20 000 per QALY to be economically attractive [21]. If the ICER is higher, an advisory board will carefully consider whether a new treatment strategy is supported or not. The Dutch Council for Public Health and Care proposes a CEA threshold that depends on the disease burden of a patient and ranges between €10 000 and €80 000 per QALY [22]. Sensitivity analyses Both trial-based and model-based CEAs harbour a number of uncertain parameters. To assess this uncertainty, sensitivity analyses are performed, which can be broken down in univariable and probabilistic sensitivity analyses. A univariable sensitivity analysis involves varying 1 parameter that carries uncertainty across a range of plausible values to assess the impact of that particular parameter on the results. In a probabilistic sensitivity analysis, all variables in the model are varied, allowing the assessment of uncertainty in all parameters simultaneously. For trial-based CEAs, a bootstrap method is generally used to assess the uncertainty in all parameters. Patients are randomly selected from a random dataset (with replacement), and the analysis is replicated on this new dataset. This procedure is then repeated a large number of times (e.g. 1000), leading to 1000 estimates of the ICER (e.g. Fig. 3) [23]. Both the probabilistic sensitivity analysis and bootstrap method procedures provide insight in the uncertainty of costs and effectiveness outcomes. The fraction of replicates of the resulting ICERs fitting into the different quadrants of the CE plane and the proportion of ICERs below or above a certain willingness-to-pay threshold can be calculated. Finally, as willingness-to-pay thresholds are not set in stone, the data from bootstrap replication can be used to produce a cost–effectiveness acceptability curve, which demonstrates the probability that a given treatment is economically attractive as a function of the cost–effectiveness threshold (Fig. 5) [24]. Figure 5: View largeDownload slide Cost–effectiveness acceptability curve. The y-axis represents the percentage of bootstraps that had an acceptable incremental cost–effectiveness ratio, and the x-axis represents the costs in Euros (€) per QALY gained. The 3 curves have different cost–effectiveness probabilities at the same costs in euros per QALY. QALY: quality-adjusted life-years. Figure 5: View largeDownload slide Cost–effectiveness acceptability curve. The y-axis represents the percentage of bootstraps that had an acceptable incremental cost–effectiveness ratio, and the x-axis represents the costs in Euros (€) per QALY gained. The 3 curves have different cost–effectiveness probabilities at the same costs in euros per QALY. QALY: quality-adjusted life-years. REAL-WORLD EXAMPLE: TRANSCATHETER VERSUS SURGICAL AORTIC VALVE REPLACEMENT One area of considerable interest for economic analyses has been the use of TAVI as an alternative to surgical aortic valve replacement (SAVR) to treat patients with severe aortic stenosis. Reasons for this interest include the large population of patients who may be candidates for either procedure and the substantial cost of the transcatheter valve prosthesis. The CoreValve US High-Risk Pivotal trial enrolled 795 patients who were randomized to TAVI or SAVR. For the CEA, detailed resource utilization (e.g. indirect costs) and hospital fees were available from randomization until 12 months later or death. QoL data were available for 2 years (1, 6, 12 and 24 months) by means of the EuroQol 5 dimensions questionnaire. The data collected during the trial were then used to predict survival, QALYs and costs (from a healthcare system perspective) over a lifetime horizon. The study showed that TAVI, when compared with SAVR, resulted in a 0.32-year increase in quality-adjusted life expectancy, with an increase in lifetime costs of approximately $18 000, resulting in an ICER of just above $55 000 [25]. Interestingly, a more recent CEA based on an intermediate-risk population showed TAVI to be cost saving when compared with SAVR, with a decrease in costs between $7900 and $9700 per QALY depending on the TAVI valve chosen [26]. These results indicate that TAVI could be considered an economically dominant treatment strategy when compared with SAVR—at least for the intermediate-risk population. A few factors could have contributed to the fact that TAVI was economically dominant in the intermediate-risk population: a 6.5-day reduction in length of stay for TAVI patients, compared to SAVR patients. This reduction in length of stay is more pronounced than in the High-Risk Pivotal trial where a difference of 4.4 days was observed when compared with SAVR [25]. Furthermore, a significant drop in rehospitalizations for TAVI and >50% reduction in rehabilitation days were observed. All 3 factors contribute to lower costs after TAVI, thereby closing the cost difference gap found in the High-Risk Pivotal trial. LIMITATIONS AND FUTURE PERSPECTIVES CEAs are an increasingly important form of research to aid and facilitate the sustainability of healthcare systems. Although CEAs are widely accepted, some have questioned the methodology. For instance, the willingness to pay threshold (often set at $50 000/QALY) is frequently questioned. However, we argue that the willingness-to-pay threshold should be seen as a guidance and not as a strict cut-point above which a treatment strategy is deemed to be too expensive. Moreover, the applicability of CEAs is questioned, as it would ignore that what is considered good on average does not necessarily have to be appropriate for individuals [27]. Others accept the inherent limitations of CEAs and see its value if used appropriately [28]. Finally, the quality of CEAs greatly depends on the quality and generalizability of the data used as input. Despite its limitations, CEAs offer us the best opportunity to evaluate not only the possible greater effectiveness of new treatment strategies but also its value for money. The development of new costly treatment strategies will continue, whereas healthcare budgets increasingly threaten other parts of the economy. This stresses the important role CEAs not only have now but also in the future. Conflict of interest: Christiaan F.J. Antonides has no conflict of interest. David J. Cohen discloses a financial relationship with Edwards Lifesciences, Medtronic, and Boston Scientific. Ruben L.J. Osnabrugge has no conflict of interest. REFERENCES 1 Health Expenditure and Financing: Current Expenditure on Health (All Functions) Measured by the Share of Gross Domestic Product (2000-2016). Organisation for Economic Co-operation and Development (OECD ), 2017 . 2 Bodenheimer T. High and rising health care costs. Part 1: seeking an explanation . Ann Intern Med 2005 ; 142 : 847 – 54 . Google Scholar CrossRef Search ADS PubMed 3 Bodenheimer T. High and rising health care costs. Part 2: technologic innovation . Ann Intern Med 2005 ; 142 : 932 – 7 . Google Scholar CrossRef Search ADS PubMed 4 Osnabrugge RL , Mylotte D , Head SJ , Van Mieghem NM , Nkomo VT , LeReun CM et al. Aortic stenosis in the elderly: disease prevalence and number of candidates for transcatheter aortic valve replacement: a meta-analysis and modeling study . J Am Coll Cardiol 2013 ; 62 : 1002 – 12 . Google Scholar CrossRef Search ADS PubMed 5 Weinstein MC , Siegel JE , Gold MR , Kamlet MS , Russell LB. Recommendations of the panel on cost-effectiveness in health and medicine . JAMA 1996 ; 276 : 1253 – 8 . Google Scholar CrossRef Search ADS PubMed 6 Oppe M , Devlin NJ , Szende A. EQ-5D Value Sets: Inventory, Comparative Review and User Guide . Dordrecht, The Netherlands : Springer , 2007 . Google Scholar CrossRef Search ADS 7 Chapman RH , Berger M , Weinstein MC , Weeks JC , Goldie S , Neumann PJ. When does quality-adjusting life-years matter in cost-effectiveness analysis? Health Econ 2004 ; 13 : 429 – 36 . Google Scholar CrossRef Search ADS PubMed 8 Cohen DJ , Reynolds MR. Interpreting the results of cost-effectiveness studies . J Am Coll Cardiol 2008 ; 52 : 2119 – 26 . Google Scholar CrossRef Search ADS PubMed 9 Cost-Effectiveness Analysis Registry (CEA-Registry) . http://healtheconomics.tuftsmedicalcenter.org/cear4/SearchingtheCEARegistry/SearchtheCEARegistry.aspx (12 December 2017, date last accessed). 10 Jain R , Grabner M , Onukwugha E. Sensitivity analysis in cost-effectiveness studies: from guidelines to practice . Pharmacoeconomics 2011 ; 29 : 297 – 314 . Google Scholar CrossRef Search ADS PubMed 11 Strong M , Pilgrim H , Oakley J , Chilcott J. Structural uncertainty in health economic decision models. ScHARR Occasional Paper, ISBN: 1 900752 27 1, 2009 . 12 Sonnenberg FA , Beck JR. Markov models in medical decision making: a practical guide . Med Decis Making 1993 ; 13 : 322 – 38 . Google Scholar CrossRef Search ADS PubMed 13 Cohen DJ , Osnabrugge RL , Magnuson EA , Wang K , Li H , Chinnakondepalli K et al. Cost-effectiveness of percutaneous coronary intervention with drug-eluting stents versus bypass surgery for patients with 3-vessel or left main coronary artery disease: final results from the Synergy Between Percutaneous Coronary Intervention With TAXUS and Cardiac Surgery (SYNTAX) trial . Circulation 2014 ; 130 : 1146 – 57 . Google Scholar CrossRef Search ADS PubMed 14 Williams C , Lewsey JD , Briggs AH , Mackay DF. Cost-effectiveness analysis in R using a multi-state modeling survival analysis framework: a tutorial . Med Decis Making 2017 ; 37 : 340 – 52 . Google Scholar CrossRef Search ADS PubMed 15 Watt M , Mealing S , Eaton J , Piazza N , Moat N , Brasseur P et al. Cost-effectiveness of transcatheter aortic valve replacement in patients ineligible for conventional aortic valve replacement . Heart 2012 ; 98 : 370 – 6 . Google Scholar CrossRef Search ADS PubMed 16 Cohen DJ , Lavelle TA , Van Hout B , Li H , Lei Y , Robertus K et al. Economic outcomes of percutaneous coronary intervention with drug-eluting stents versus bypass surgery for patients with left main or three-vessel coronary artery disease: one-year results from the SYNTAX trial . Catheter Cardiovasc Interv 2012 ; 79 : 198 – 209 . Google Scholar CrossRef Search ADS PubMed 17 Neumann PJ. Costing and perspective in published cost-effectiveness analysis . Med Care 2009 ; 47 : S28 – 32 . Google Scholar CrossRef Search ADS PubMed 18 Schwarzer R , Rochau U , Saverno K , Jahn B , Bornschein B , Muehlberger N et al. Systematic overview of cost-effectiveness thresholds in ten countries across four continents . J Comp Eff Res 2015 ; 4 : 485 – 504 . Google Scholar CrossRef Search ADS PubMed 19 Neumann PJ , Cohen JT , Weinstein MC. Updating cost-effectiveness—the curious resilience of the $50,000-per-QALY threshold . N Engl J Med 2014 ; 371 : 796 – 7 . Google Scholar CrossRef Search ADS PubMed 20 Anderson JL , Heidenreich PA , Barnett PG , Creager MA , Fonarow GC , Gibbons RJ et al. ACC/AHA statement on cost/value methodology in clinical practice guidelines and performance measures: a report of the American College of Cardiology/American Heart Association Task Force on Performance Measures and Task Force on Practice Guidelines . J Am Coll Cardiol 2014 ; 63 : 2304 – 22 . Google Scholar CrossRef Search ADS PubMed 21 McCabe C , Claxton K , Culyer AJ. The NICE cost-effectiveness threshold: what it is and what that means . Pharmacoeconomics 2008 ; 26 : 733 – 44 . Google Scholar CrossRef Search ADS PubMed 22 Cost-effectiveness in practice. National Health Care Institute, The Netherlands, 26 June 2015. 23 Campbell MK , Torgerson DJ. Bootstrapping: estimating confidence intervals for cost-effectiveness ratios . QJM 1999 ; 92 : 177 – 82 . Google Scholar CrossRef Search ADS PubMed 24 Al MJ. Cost-effectiveness acceptability curves revisited . Pharmacoeconomics 2013 ; 31 : 93 – 100 . Google Scholar CrossRef Search ADS PubMed 25 Reynolds MR , Lei Y , Wang K , Chinnakondepalli K , Vilain KA , Magnuson EA et al. Cost-effectiveness of transcatheter aortic valve replacement with a self-expanding prosthesis versus surgical aortic valve replacement . J Am Coll Cardiol 2016 ; 67 : 29 – 38 . Google Scholar CrossRef Search ADS PubMed 26 Cohen DJ. Cost-effectiveness of transcatheter vs. surgical aortic valve replacement in intermediate risk patients. Results from the PARTNER 2A and SAPIEN 3 intermediate risk trials. Presented at TCT 2017, Denver, Colorado, USA, 2017 . 27 Diamond GA , Kaul S. Cost, effectiveness, and cost-effectiveness . Circ Cardiovasc Qual Outcomes 2009 ; 2 : 49 . Google Scholar CrossRef Search ADS PubMed 28 Weintraub WS , Cohen DJ. The limits of cost-effectiveness analysis . Circ Cardiovasc Qual Outcomes 2009 ; 2 : 55. Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

European Journal of Cardio-Thoracic Surgery – Oxford University Press

Published: May 2, 2018

There are no references for this article.