- Email: [email protected]
Introduction to health economics and decision-making: Is economics relevant for the frontline clinician?
Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
Contents lists available at ScienceDirect
Best Practice & Research Clinical Gastroenterology
1
Introduction to health economics and decisionmaking: Is economics relevant for the frontline clinician? Ron Goeree, MA, Professor a, b, *, Vakaramoko Diaby, PhD, Post-doc fellow a, b a
Programs for Assessment of Technology in Health (PATH) Research Institute, St Joseph’s Healthcare Hamilton, Hamilton, Ontario, Canada b Department of Clinical Epidemiology & Biostatistics, McMaster University, Hamilton, Ontario, Canada
a b s t r a c t Keywords: Health economics Economic evaluation Clinical decision-making Decision-making criteria
In a climate of escalating demands for new health care services and significant constraints on new resources, the disciplines of health economics and health technology assessment (HTA) have increasingly been turned to as explicit evidence-based frameworks to help make tough health care access and reimbursement decisions. Health economics is the discipline of economics concerned with the efficient allocation of health care resources, essentially trying to maximize health benefits to society contingent upon available resources. HTA is a broader field drawing upon several disciplines, but which relies heavily upon the tools of health economics and economic evaluation. Traditionally, health economics and economic evaluation have been widely used at the political (macro) and local (meso) decision-making levels, and have progressively had an important role even at informing individual clinical decisions (micro level). The aim of this paper is to introduce readers to health economics and discuss its relevance to frontline clinicians. Particularly, the content of the paper will facilitate clinicians’ understanding of the link between economics and their medical practice, and how clinical decision-making reflects on health care resource allocation. Ó 2013 Elsevier Ltd. All rights reserved.
* Corresponding author. PATH Research Institute, 25 Main St. W., Suite 2000, Hamilton, Ontario L8P 1H1, Canada. Tel.: þ1 (905) 523 7284x5268; fax: þ1 (905) 522 0568. E-mail address: [email protected] (R. Goeree). 1521-6918/$ – see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.bpg.2013.08.016
832
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
Background A combination of the rise of health care expenditures in many jurisdictions (see Fig. 1), everincreasing technological innovation in health care (especially in drugs and devices), patients’ insatiable needs and demands, the emergence of new diseases and demographic changes [1] have all come together to force decision- and policy-makers, at all levels of the health care system, to rethink reimbursement and provision decisions. With limited capacity of the health care system to handle the accumulation of these factors, health care decision-makers have recognized the need for prioritizing between competing health care uses and the important role a transparent, structured and evidencebased process can play in making decisions. General economic principles offer a theoretical foundation for dealing with resource allocation in an environment of scarcity of resources [2]. The underlying rationale is that, in a free market, resource allocation is regulated by price, ability to pay [3] and perfect information, such that market forces will provide the optimal allocation of health care resources. However, in health care significant market failures exist that prevent a freely functioning market. For example, health care professionals have significantly more knowledge about the potential benefits and risks of health care treatment (i.e. asymmetry of information) and as a result patients place trust in the advice and recommendations provided to them and this can influence patient preferences and decisions [3,4]. This is different than in many other markets like buying a new car where consumers can better judge the benefits to them of car attributes or enhancements and know how much they are willing to pay for these features. Similarly, obstacles to a freely functioning health care insurance market exist such that a number of individuals may not be able to afford adequate health insurance. These, and other market failures in health care, hinder a free market optimal allocation of resources in the health care sector. Therefore, general economic theory is not a viable option for supporting decision-making in health care as it fails to address market failures and society’s ethical attitudes toward health and the provision of health care. The field of health economics (HE) proposes to maximize benefits stemming from the use of health technologies against available resources [3]. As HE recognizes, and attempts to deal with, the various forms of market failure in health care, the use of HE has been legitimated as a framework to guide decisions around health care resource allocation. HE can broadly be defined as a field of economics concerned with the way health care resources are allocated among competing alternatives, at the societal level [5]. HE encompasses many techniques and tools among which economic evaluation (EE) is the most relevant to health care decision-makers. For this reason, the main focus of this paper will be on EE and its role in health care decision-making. According to Drummond et al, EE is ‘the comparative analysis of alternative courses of action both in terms of their costs and consequences’ [6]. It is a widely adopted tool for the assessment of the value of health technologies. This is evidenced by its formal application as part of reimbursement decisionmaking, at the macro and meso levels (political and administrative levels), as well as the abundant production of health EEs in the medical literature. However, concerns have been raised regarding the
Fig. 1. Total expenditure on health care in six countries, members of the Organisation for Economic Co-operation and Development (OECD), as a proportion of gross domestic product. Source: Adapted from OECD Health Data 2012 (http://stats.oecd.org/Index.aspx? DataSetCode¼SHA).
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
833
relevance and ethics of using EE to inform clinical decisions at the micro level [7]. It has been argued that the relevance of EE to decision-makers differs according to the decision-making level [8,9]. For example, many consider it unethical to include ‘costs’ in clinical decision-making as they believe that saving, prolonging or improving patients’ quality of life is priceless. From an economic point of view, health economists tend to consider ‘opportunity costs’, and argue that every decision, even at the clinical decision level, means that there is something else which cannot be done and it is the benefits foregone from this alternative use of time or resources which is the real cost of any decision or action [10]. The argument is that disregarding these alternative uses (which would be accounted for in an EE), would in fact itself be considered unethical from a society point of view. As such, although the role of HE and EE in clinical decision-making might be controversial, this debate stresses the need to explore further the way HE and EE might impact frontline clinician practice. In this paper, we intend to introduce readers to general principles of HE and decision-making. The emphasis will be on EE, presenting the major tools and techniques it covers, however central to this discussion is the relevance of HE more generally to clinical decision-making. The paper is outlined as follows. The first section provides an introduction to HE in general and EE in particular. This introduction will include a discussion of the different types of EE, as well as a discussion about the different approaches used to implement these evaluations. The second section is devoted to linking EE and decision-making, while the third section discusses the potential role of EE for frontline clinicians. And finally, the paper ends with practice points and recommendations in terms of a research agenda. Introduction to health economics As defined in the introduction, HE is a branch of economics that focuses on the field of health care. HE addresses a number of different topics, and there are many papers and books devoted to these topics. One useful summary of the topics was provided by Williams [11], where he distinguished between the following HE topics: U U U U U U U U
Definition of health and estimation of its value Determinants of health other than health care Determinants of the demand of health care Determinants of the supply of health care (behaviour of doctors and health care providers) Economic evaluation Economic equilibrium–market equilibrium Health system analysis Planning, budgeting, and monitoring of health care.
Discussing all the topics above is beyond the scope of the paper. Among these topics, EE methods have received particular attention as these methods have become a requirement in most jurisdictions for supporting publicly financed health technology reimbursement decision-making since the early 1990’s [12]. Consequently, from an HE perspective, EE is considered to be most relevant to clinicians and decision-makers. The next section introduces the different types of EE and the approaches used to implement them. Types of economic evaluations This section is not meant to give a comprehensive presentation of EE methods as many textbooks are devoted to this subject, including the textbook by Drummond et al [6]. Rather, the objective is to briefly cover some key aspects of the different types of EEs. EE attempts to estimate and present the expected costs and outcomes of alternative courses of action. It is generally agreed that the different types of EEs can be subdivided into four categories (see Fig. 2), these are: ‘cost-minimization analysis (CMA)’; ‘cost-effectiveness analysis (CEA)’; ‘cost-utility analysis (CUA)’; and ‘cost-benefit analysis (CBA)’. These methods all compare competing options both in terms of costs and outcomes, with the main difference being the way outcomes are measured, valued and included in the analysis (see Fig. 2).
834
Types of economic evaluations
Cost effectiveness analysis (CEA)
True CEA
Evaluation: Comparison of interventions in terms of costs only as they do not differ in terms of benefits (rarely the case) Costs: Measured and valued monetarily (dollars, euros, etc…)
Cost benefit analysis (CBA)
Cost-utility analysis (CUA)
Evaluation: Comparison of interventions in terms of costs and benefits
Evaluation: Comparison of interventions in terms of costs and benefits
Evaluation: Comparison of interventions in terms of costs and benefits
Costs: Measured and valued monetarily (dollars, euros, etc…)
Costs: Measured and valued monetarily (dollars, euros, etc…)
Costs: Measured and valued monetarily (dollars, euros, etc…)
Benefits: Expressed and valued in natural units (cases detected, events averted, life years gained)
Benefits: Expressed in natural units and converted in Quality Adjusted Life Years (QALYs)
Benefits: Expressed in natural units and valued monetarily
CMA: Cost minimization analysis; CEA: Cost effectiveness analysis; CUA: Cost utility analysis; CBA: Cost benefit analysis Fig. 2. Types of economic evaluations.
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
Cost minimization analysis (CMA)
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
835
Cost-minimization analysis (CMA) A CMA compares alternatives in terms of expected costs and outcomes, but proponents of CMA claim that the outcomes of the alternatives are identical and therefore the focus is on differences in costs only. This type of EE stems from a frequentist interpretation of outcomes, where it is concluded that if there is no statistically significant difference in outcomes (usually at a 5% level of significance), then the conclusion would be that there is no difference between treatments in terms of outcomes. However, this type of approach has met with strong criticism from researchers as the premise of interventions being exactly equivalent in terms of effects is rarely valid, especially from a Bayesian statistical point of view. For example, Briggs and O’Brien postulated that the equivalence assumption is based on an incorrect interpretation of hypothesis tests used to determine whether (or not) the difference between interventions’ efficacy is statistically significant [13]. In addition, since researchers who conduct an EE with an assessment of both costs and outcomes intend to conduct either a CEA, CUA or CBA, several opponents have also argued that a categorization of CMA makes no sense. Nevertheless, this type of EE is still used by some researchers. An example of CMA is shown in the study by Kleus et al 2009 [14]. In this study, the authors aimed to compare two operative techniques: small incision versus laparoscopic cholecystectomy, in a blinded randomized trial. As the authors reported that analysis of the clinical evidence did not favour one alternative over its comparator, they compared both alternatives in terms of costs only. They concluded that the small incision operative technique represents the greatest value for money, whatever the perspective adopted. However as noted above, the use of CMA in the manner conducted by the authors is not encouraged by most health economists. For a full economic evaluation, the authors should have reported more fully on the outcomes, regardless of the statistical significance of the clinical trial findings. Cost-effectiveness analysis (CEA) In early days of EE, CEA was the most common type of EE used in the health care field. With this method, alternatives are compared in terms of costs and outcomes and outcomes are measured and valued in natural units collected in clinical trials or observational studies. Examples of outcome measures used in CEA might include number of events (acute myocardial infarctions or strokes), number of cases detected, symptom free days (SFDs), progression free survival, or overall survival usually expressed as life years (LYs). Since the outcome measure used in a CEA is a natural outcome from clinical trials, conducting a CEA alongside a clinical trial, using the trial’s primary outcome measure for the EE, is common practice. In a CEA, expected costs and outcomes (however measured) are calculated and compared. If one treatment has a higher expected cost and has lower expected outcomes, then that treatment is determined to be dominated by the alternative treatment. However, if there is a trade-off between one treatment costing more but also providing improved outcomes, then the results of a CEA are presented in terms of an incremental cost-effectiveness ratio (ICER), which is literally the differences in mean expected costs divided by the difference in mean expected outcomes. The ICER provides a measure of the expected cost needed to gain a unit of effect (e.g. cost per year of life gained or cost per stroke averted). An example of a CEA was conducted on an early intervention with budesonide in the management of mild persistent asthma symptoms [15]. The comparator was usual care, and the outcome measure was symptom free days (complete 24-h periods with no asthma symptoms). According to a payer and societal perspectives, the authors respectively estimated an ICER of $11.30 and $3.70 per SFD gained. Another example of a CEA was an evaluation of drug eluting stents versus bare metal stents for coronary artery disease where the authors found that the cost-effectiveness of drug eluting stents varied depending on patient risk profiles (e.g. $14,000–$279,000 per revascularization procedure avoided) [16]. Although convenient to apply because the outcome measure used in CEA typically matches outcome measures used in clinical trials, two of the main challenges for decision-makers in using CEA is that the outcome measures can be intermediate in nature like a change in a patient risk factor with questionable impact on more final patient outcomes, and that it is difficult for decision-makers to compare CEA outcomes across diseases and interventions when making health care resource allocation decisions. For example, how would a decision-maker compare a $/stroke avoided versus
836
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
$/revascularization avoided versus $/life year gained for making society health care decisions? This problem of having an intermediate outcome is not unique to CEA and also applies to the assessment of clinical outcomes. As decision-makers request more final outcome measures for clinical assessments, this will also help for EE as the outcome measure used in CEA can be more final in nature. Cost-utility analysis (CUA) In a CUA, expected costs and outcomes for each intervention are calculated, with the outcome measure expressed as quality adjusted life years (QALYs) [6]. Although a more composite outcome measure because it combines quality of length and length of life into a single measure, some consider the QALY simply to be another outcome measures, and as such many consider a CUA to be a variant of a CEA. Still others consider a CUA to be unique in that the QALY can combine various outcomes or impacts (e.g. length of life and quality of length from various health states or events) from a disease or intervention into a single measure of effect which can then facilitate decision-maker comparisons across diseases or interventions. The QALY is calculated by combining weights for the quality of life of being in different health states (i.e. perfect health ¼ 1; death ¼ 0) with the length of time spent in each health state, to obtain an overall measure of length of life weighted by quality of life. Different instruments can be used to determine the quality of life weights (or utility value) [6], but once obtained the results of a CUA are checked for dominance first and if there is a trade-off between higher costs and greater QALYs with an intervention, then the results are presented in terms of incremental cost-utility ratio (ICUR). As CUA allows for the comparison of interventions with different outcomes using a common denominator, it represents a more useful analysis for decision-makers trying to compare across programs and services. CUA has become a standard requirement for EE in many jurisdictions around the world, however, methodological controversy and concern regarding the use of QALYs as a generic measure of health benefits remain [17]. An example of a CUA is presented in the study by Diaby et al [18], where the authors compared hormone replacement therapy (HRT) with a synthetic steroid for the management of postmenopausal women suffering climacteric symptoms. The authors reported an ICUR of $9198 per QALY gained with HRT. Another example of a CUA is from a three-year prospective study comparing medical management to laparoscopic surgery in patients with chronic gastroesophageal reflux disease [19]. In this study the authors reported an ICUR of $29,000 per QALY gained with surgery compared to medical management. As the outcome measure is the same across these two divergent studies, decisionmakers can use the results of these evaluations and combine it with other factors (e.g. ethical and political issues) to better assist them in making resource allocation decisions. Cost-benefit analysis (CBA) In a CBA costs are measured and valued the same way as in CEA or CUA, however, outcomes are measured in natural units and then valued in monetary terms. Since both costs and outcomes are measured in the same monetary terms, it can be determined whether the net monetary value of an intervention is positive or negative. Different techniques are available to value benefits monetarily and include the human capital approach, the stated preference methods and the willingness to pay approach. The details about these methods are provided in the textbooks by Drummond et al [6] and McIntosh et al [20]. A common misconception when using or interpreting CBA, is that if the CBA is greater than zero (i.e. monetary value of benefits outweighs the costs), then this means the intervention should be funded. This misconception ignores the fact that decision-makers cannot afford to fund everything in society (even if CBA > $0), and prioritization and other factors like ethical issues and societal values still need to be considered in decision-making processes. Although CBA is a more flexible tool for incorporating a broader range of potential impacts (both positive and negative) from an intervention [21], there have been methodological concerns raised regarding its application to health care. Major issues with implementing CBA in health care include the complexity of measuring health benefits monetarily and the ethical issues arising from valuing them. For these reasons, the conduct of CBA to inform health care resource allocation is very limited compared to CEA and CUA. It should be noted that a large proportion of studies in the literature claiming to be CBAs are actually either cost analyses or cost-consequence analyses [22].
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
837
An illustration of CBA is provided by the study of Ney et al [23]. In their study, the authors sought to determine whether it was worthwhile implementing an intraoperative neurophysiological monitoring (IOM) for spinal surgeries. The results of the study suggest that, compared to ‘no monitoring’, IOM is implementing because the mean lifetime cost of IOM was estimated at $35,325 compared to $58,514 for unmonitored spinal surgeries. Approaches to implementing an economic evaluation Distinct from the types of EEs discussed in the previous sections, there are two main approaches to conducting an EE: trial-based and modelling-based EEs. These approaches are described in more detail below. Trial-based economic evaluations This approach consists of adding on an economic component to a clinical trial and using both the clinical trial data and economic add-on information to conduct the EE [24]. With this approach, the data proposed to be collected for the clinical trial is reviewed and additional information that would be important to obtain from patients during the prospective phase of the clinical trial for the economic analysis are collected. This information usually consists of health care resource utilization such as hospitalizations, doctor visits, drugs, tests and procedures and the use of other health care programs and services. Depending on the needs of the analysis, other information such as patient out-of-pocket costs or lost productivity might also be collected. In addition to costing information, it may be necessary to consider an alternative or additional outcome measure for the EE. For example, if the primary outcome measure is a more final outcome measure such as mortality/survival, a CEA might be conducted based on a cost per life year saved measure and this might be sufficient for the economic analysis. However, if the primary clinical outcome measure is too intermediate or too specific to the disease, it might be necessary to consider adding on a more broader and final outcome measure for the EE. For example, many reimbursement agencies require a CUA using QALYs as the measure of outcomes, and in that case it might be necessary to collect quality of life information directly on patients during the trial. Conducting a trial-based EE has many advantages. Since information is collected directly on patients in the trial, minimal assumptions are required and the results have high internal generalizability. However, the use clinical trials as the sole source of evidence may contradict with some requirements that must be met for a proper implementation of EEs. Briggs et al alluded to these requirements as being synthesis, incorporation of all relevant comparators, and appropriate temporal framework [25]. Synthesis refers to the imperative to include all relevant data sources in the EE as opposed to a unique source of evidence available through a single clinical trial. Clinical trials also have the disadvantage that they typically compare only two alternatives, while the relevant policy question might involve a comparison of all relevant treatment alternatives. And finally, clinical trials are often shorter term in duration, and may not be adequate for determining the long term costs and outcomes of alternative course of action (especially for chronic diseases). For these reasons, decision analytic modelling techniques are often used as a substitute for, or complement to, trial-based economic analyses. Before discussing the role of decision analytic modelling in EEs, there are two other requirements that should be described which are common to both EE approaches: the perspective of the study; and the quantification of uncertainty. The perspective of the study represents the viewpoint against which costs and effects are identified and valued. Many perspectives can be adopted in an EE, the most commonly adopted perspectives being the third-party payer (e.g. reimbursement agency) and the societal perspectives. It is recommended to present the results of an EE in light of different perspectives so that decision-makers have a better sense of the implications of the results according to broader or narrower perspectives. Another important consideration is uncertainty. As uncertainty is inherent in the collection of data that are used in any evaluation, it is critical to determine how uncertainty could influence the EE results. This is usually addressed using a series of sensitivity analyses where alternative assumptions or values are considered. In this regard, different techniques to handle uncertainty have been developed through years and are discussed in textbooks by Drummond et al [6] and Briggs et al [25].
838
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
An example of trial-based EE is presented in the study by Hoeijenbos et al [26]. The authors conducted a cost-utility analysis to compare the implementation of an evidence-based physiotherapy guideline for low back pain. Based on a randomized trial of 500 patients with back pain, the authors concluded that the evidence-based physiotherapy guideline was not cost-effective compared to the standard guideline dissemination method as no differences were found between both costs and effects of these strategies. Another example of a trial-based EE is the three-year prospective CUA study comparing 104 patients randomized to medical management or laparoscopic surgery in patients with chronic gastroesophageal reflux disease [19]. In this study, the authors reported that surgery was fairly cost-effective with an ICUR of $29,000 per QALY gained. Modelling-based economic evaluations The HE field has witnessed an increasing use of decision analytic models as a tool for informing decision-making under uncertainty. In recent years, there has been a flourishing production of textbooks providing guidance on how to design and use decision analytic models in health EEs [6,25,27]. Decision analytic models can be subdivided into two main categories: cohort and patient-level simulation models (see Fig. 3). Cohort models typically depict the possible consequences that might occur to a group of individuals sharing the same characteristics [25]. The patient experience modelled in this case is that of the average patient. The most common sub-types of cohort models are decision trees and Markov models. Nonetheless, it is not unusual to come across hybrid models, meaning models consisting of a mix of a decision tree for the short term duration of the model and a Markov model for longer term cost and outcome projections. Decision trees represent the simplest form of decision analytic models [28]. A full description of decision trees is presented in the book by Drummond et al [6] and only key features of decision trees are summarized here. Decision trees are built from left to right, and start with a decision node (a square) representing a choice between competing interventions. Following the decision node appear chance nodes (circles), representing points beyond which upcoming events are uncertain. This uncertainty is expressed by assigning probabilities (summing up to one at each chance node) to the occurrence of possible events, which are mutually exclusive. As constructed, a decision tree is analogous to possible clinical pathways that clinicians typically consider for their patients. However, instead of just presenting the possible pathways, in decision analytic modelling the expected costs and outcomes of each treatment alternative are calculated based on the costs and outcomes of each clinical pathway weighted by the probability of going down each pathway. The major advantage of using decision trees is that since pathways are clearly laid out, the model and calculations are transparent for less experience modellers. The major disadvantage of decision trees is that the tree can get very large (i.e. bushy) the more comparators included in the analysis, the larger the number of health states patients can experience and the longer the model is run (e.g. long term chronic disease). Markov models are similar to decision trees in that patients are modelled as being in defined health states over time, but are different in that the clinical pathways for patients are only partially displayed in the model diagram, and key information about how patients progress in the model are dictated though the use of equations, formulas and tables which are largely hidden in the background. This makes Markov models more manageable for complex and chronic diseases, but also makes them less transparent to less experienced modellers. Markov models tend to be used more for chronic diseases where the goal is to model disease progression or risk of events over longer periods of time [28]. The model is constructed based on a finite number of mutually exclusive health or Markov states, over a series of equal time periods referred to as Markov cycles. At the end of each cycle, patients may transition from one state to another or remain in the same health state based on transition probabilities which are estimated from trial data or from longer term observation study data. For each cycle, each health state is associated with a cost and effect/outcome measure. To obtain overall expected costs and outcomes for each treatment alternative, these costs and effects are multiplied by the estimated time patients will spend in each state [25]. Even though it is generally accepted that Markov models are well suited for modelling diseases with on-going risks [28], they have limited ability to structure very complex conditions [25]. This is due to the so called Markov assumption [25,28], which means that
Patient-level models
Cohort models
Decision trees
Simplest form of cohort models; Parsimonious model in terms of data needed to conduct analyses
Markov models
Microsimulation
More sophisticated cohort model as they model chronic and recurrent conditions
Integrated system of equations used to model clinical conditions at the patient level;
Can be used with decision trees
Discrete Event Simulation (DES)
System analysis modeling Modeling competition for resources
Allow for building patient history
*: United Kingdom Prospective Diabetes Study model [27]
Agent-based models
Modeling of biological and physiological processes;
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
Decision analytic models
Fig. 3. Types of decision analytic models.
839
840
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
there are limitations with standard Markov models to account for the past medical history of a patient. Markov models can be made more complex by adding in additional health states (e.g. number of times experiencing an event) to reflect the past history of the patient. An example of a Markov model was the CEA of drug eluting stents versus bare metal stents for coronary artery disease [16], where the trial data was used to estimate short term outcomes and a Markov model was used to estimate longer term costs and outcomes. In this study the authors conducted both a CEA and CUA using trial data for the short term costs and outcomes and a Markov model for estimating longer term costs and outcomes. As shown in Fig. 3, the other main modelling category is patient-level simulation models. Patientlevel simulation models are more complex and more data demanding than cohort models. There are three main sub-types of patient-level simulation models: microsimulation models; discrete event simulation (DES) models; and agent-based models. Microsimulation models allow for variability between patients [25] as patients are considered individually in the model. Each patient travels through the model one at the time, with the model being able to accommodate patients’ individual characteristics and their disease history/progression. An example of a microsimulation model is the United Kingdom Prospective Diabetes (UKPDS) model. It was developed by Clarke et al [29] and further adapted and used to model other conditions. The model consists of an integrated system of parametric risk equations allowing for estimation of seven diabetes complications and their associated costs and outcomes. DES models have the further advantage that they can consider resources and resource constraints. For example, typical cohort models ignore capacity constraints and assume resources can simply be increased. As such these models are ideal for modelling bottlenecks in the system or what might be the implications of increasing health care use beyond capacity (e.g. emergency room visits, surgical suite use). DES models are typically used to model wait times and queuing patterns. For example, the study by Lim M et al [30] provides a good example of a DES. In this study, the authors sought to call decisionmakers’ attention on the importance of accurately modelling physician interactions and roles in which they provide care to patients. The study showed that ignoring constraints and interactions could significantly underestimate estimates of physician time utilization. Finally, agent-based models are sophisticated models in which agents are conscious of themselves and their environment [31]. Agents in these models can be patients, objects or even groups. They have the advantage to model interactions between different agents or entities. In health care, these models are useful to model disease transmission, epidemics or any problems that involve physiological or biological processes. Examples of agent-based modelling in health care can be found in the literature [8,9]. It should be emphasized that in all types of economic models, that the results of the model are dependent upon assumptions made the quality of data inputs used in the model construction. For example, an economic model based on inputs from one clinical trial may produce different results than a model based on inputs from another clinical trial. For this reason, it is encouraged to carefully consider the inputs used in economic models and wherever possible use clinical data that are synthesized from comprehensive literature reviews and meta-analyses. Economic evaluation and decision-making Although geographic variations do exist, health care decision-making regarding the adoption and reimbursement of technologies has traditional been based on four main criteria. These criteria are identified in a literature review [32] comparing the selection processes of reimbursable drugs in public drug benefit plans across high- , middle- and low-income countries. These criteria are safety, efficacy, value for money (cost-effectiveness) and budgetary impact (affordability). Many see these criteria as hurdles, where new technologies must be proven first to be safe, then effective, then cost-effective and then finally affordable. Following considerations of safety and effectiveness, it should be noted that EEs are not meant to be used in a formulaic way. Regardless of the type of EE used (e.g. CMA, CEA, CUA, CBA), if a treatment dominates another (i.e. improved outcomes and lower cost), then that treatment is determined to represent both good value for money and would be affordable from the perspective taken for the
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
841
analysis. It should be noted, however, that because of different budget responsibilities (i.e. budget silos), that even though a treatment may be cost saving from a societal perspective overall, some of the cost savings (or cost offsets) may benefit different ‘payers’ in a health care system, and as such a dominant treatment may still be considered unaffordable to a particular funding agency. For example, this typical occurs with the assessment of new drugs, where the drug results in additional costs to drug program branch and the cost savings (e.g. reduced doctor visits or hospitalizations) are realized by a different budgetary ‘silo’ of the government. In the case of a trade-off between increased outcomes but at a higher cost, then there is always a judgment required regarding both value for money and affordability. In other words, there is no ‘threshold’ where the results of an EE would automatically be considered worthwhile. In a CMA, where outcomes were determined to be identical (rarely if ever the case), there is judgment required regarding the extra cost and extra budgetary impact of the intervention. In a CBA, even if the monetary valuation of benefits exceeds costs, there is still judgment required regarding affordability. And in a CEA or CUA, there is judgment required to determine for example whether the cost per life year gained, cost per stroke avoided or cost per QALY gained represented good value for money and whether the estimated budgetary impact is affordable. Instead of presenting EE results as a cost-effectiveness ratio, some researchers have proposed the use of a measure called the net monetary benefit (NMB) [24], but this measure requires the implicit use of a specific decision-making threshold which many argue should be discouraged in EE. Where judgments are required for determining whether a treatment represents good value for money, some agencies have developed rigid ‘decision rules’ and other agencies have adopted informal guidelines. In Canada for example, Laupacis et al proposed grades of recommendations for the adoption and appropriate use of new technologies based on $20,000 and $100,000 per QALY thresholds [33]. In the United Kingdom (UK), more rigid thresholds have been used (i.e. £20,000– £30,000 per QALY), but empirical and statistical research have shown that the real threshold used by the National Institute for Health and Care Excellence (NICE) in making funding decisions may be in excess of £30,000 [34,35]. As mentioned in the beginning of this section, EE is an important but not sufficient dimension of decision-making regarding the provision or funding of new technologies. In recent years, many jurisdictions have been challenged with the need to expand the four traditional criteria. Some additional criteria proposed include patient preferences, social and ethical values. Until now, the full incorporation of these criteria into existing decision-making processes is in early stages. In this regard, there has been a growing interest into alternative decision-making frameworks, among which multicriteria decision analysis (MCDA) has obtained particular attention. MCDA is both a process and set of methods that explicitly and simultaneously take into consideration multiple decision-making attributes that may conflict to a certain extent [36]. Some jurisdictions, including Canada [37], the United Kingdom [38], and Germany [39] have explored the formal use of this type of decision-making framework. However, research into alternative decision-making frameworks, using EE among other dimensions, is still in its infancy and more work in this area is needed to fully inform decision-making in health care. Is economics relevant for the frontline clinicians? HE, EE in particular, represents a systematic and sound framework for allocating health care resources against a context of limited budget and increasing demand. Its role in informing decisionmaking at the political and local levels has been demonstrated through the past years and is well understood by all stakeholders of decision-making processes. However, the role of EE at the micro level remains a source of debate. Two major views regarding this contentious subject emerge from the literature. The first view supports the statement that using EE in clinical practice conflicts with physician–patient relationship [40–42], and may be considered unethical. Indeed, Merino argues that EEs help identify alternatives that are good value for money at a population level while frontline clinicians are confronted with making decisions at the patient-level [41]. Under the Hippocratic Oath, frontline clinicians seek the best care for their individual patient, regardless of the impact of their decision on the remaining patients seeking care and costs borne by the society as a whole. Additionally,
842
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
the perspectives considered, the preferences and values measured in EEs do not meet the best interests or do not reflect the values of the patients. Following this rationale, making individual clinical decisions in light of the results of ‘population-based’ EEs may be misleading. Ideas opposing the first view have been presented by many authors including Williams [43,44], Carlsen and Norheim [45] and Lessard et al [46]. Lessard et al acknowledge the difficult decisions the frontline clinicians are faced with on a daily basis. They are prone to pressure of time, management of multiple goals while complying with principles of the Hippocratic Oath. However, this situation should not prevent the medical practitioners to make use of evidence produced in EEs. Lessard et al [46] support the idea ‘the real cost of any health decision is the health benefits achievable in some other patient which have been forgone by committing the resources in question to the first patient’ [45]. In this sense, a simple decision to prescribe a new drug or a lab test affects the allocation of other health care resources. Thus, understanding the economic component of decision-making may help practitioners comprehend why a new treatment promoted by a sales representative is not available for a desperate patient in need of new alternatives. On another note, decision analytic models in health care was first used for clinical decision-making and promoted by medical doctors such as Beck and Pauker [47] and Sonnenberg and Beck [28]. This reinforces the idea that economics, in health care for instance, is not about ‘cost per se’, but rather is about opportunity cost. The latter applies both to costs and benefits. The broader picture is that using an economic framework helps ensuring that health resources are allocated in an efficient way, especially in medical practice. As clinicians continue to expand their role beyond provider of care to decision-making bodies and committees, it is imperative that they have a broader understanding of HE, EE and decision-making processes in order assist policy and decision-makers in making the best use of our available health care resources. In this regard, continued education is needed for frontline clinicians about basic economic principles like scarcity and limitation of resources and that in a constrained health care system every decision made by the clinician for every patient they treat has an opportunity cost in terms of what cannot be done for another patient. Not only should training in HE and EE become a mandatory component of education for all health care professionals, but HE and EE should also become a more important component of continuing education (e.g. conferences, journals), re-certification programs and practice guideline development.
Practice points Decision-makers including clinicians are challenged with increasing demands from patients, due to medical (therapeutic) innovation, and scarcity of resources. Inevitably, tough choices have to be made among competing alternatives. HE and EE methods in particular encompasses a set of methods that provides a sound framework to guide resource allocation in health care, including clinical decisions. As clinical practices are settings where these resource allocation decisions occur frequently, it is important for frontline clinicians to understand the concept of opportunity cost inherent in decision-making and incorporate it in their decisions.
Research agenda Decision-making at the clinical level should be investigated to have a grasp of clinicians’ attitude and perceptions towards the role of HE, and in particular EE, plays in their daily practice. Gaps in education and training should be identified and appropriate educational venues need to be developed and implemented.
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
843
Conflict of interest statement The authors declare they have no conflict of interest.
Acknowledgements No funding source to acknowledge.
References [1] Mechanic D, Rochefort AD. Comparative medical systems. Annu Rev Sociol 1996;22:239–70. [2] Sussex J, Towse A, Devlin N. Operationalizing value-based pricing of medicines: a taxonomy of approaches. Pharmacoeconomics 2013;31:1–10. [3] Kernick D. An introduction to the basic principles of health economics for those involved in the development and delivery of headache care. Cephalalgia 2005;25:709–14. [4] Arrow K. Uncertainty and the welfare economics of medical care. Am Econ Rev 1963;53:941–73. [5] Kernick DP. Introduction to health economics for the medical practitioner. Postgrad Med J 2003;79:147–50. [6] Drummond MF, Sculpher MJ, Torrance GW, O’Brien BJ, Stoddart GL. Methods for the economic evaluation of health care programmes. Oxford: Oxford University Press; 2005. [7] Cunningham SJ. An introduction to economic evaluation of health care. J Orthod 2001;28:246–50. [8] Burke DS, Epstein JM, Cummings DA, Parker JI, Cline KC, Singa RM, et al. Individual-based computational modeling of smallpox epidemic control strategies. Acad Emerg Med 2006;13:1142–9. [9] Halloran ME, Ferguson NM, Eubank S, Longini Jr IM, Cummings DA, Lewis B, et al. Modeling targeted layered containment of an influenza pandemic in the United States. Proc Natl Acad Sci U S A 2008;105:4639–44. [10] Tuck J. How is marginal analysis used in economics; 2013 [Ref Type: Online Source]. [11] Williams A. Health economics: the cheerful face of a dismal science. London: Health and Economics; 1987. [12] Hutton J. ‘Health economics’ and the evolution of economic evaluation of health technologies. Health Econ 2012;21:13–8. [13] Briggs AH, O’Brien BJ. The death of cost-minimization analysis? Health Econ 2001;10:179–84. [14] Keus F, de JT, Gooszen HG, Buskens E, van Laarhoven CJ. Cost-minimization analysis in a blind randomized trial on smallincision versus laparoscopic cholecystectomy from a societal perspective: sick leave outweighs efforts in hospital savings. Trials 2009;10:80. [15] Sullivan SD, Buxton M, Andersson LF, Lamm CJ, Liljas B, Chen YZ, et al. Cost-effectiveness analysis of early intervention with budesonide in mild persistent asthma. J Allergy Clin Immunol 2003;112:1229–36. [16] Goeree R, Bowen JM, Blackhouse G, Lazzam C, Cohen E, Chiu M, et al. Economic evaluation of drug-eluting stents compared to bare metal stents using a large prospective study in Ontario. Int J Technol Assess Health Care 2009;25:196–207. [17] Weinstein MC. A qaly is a qaly – or is it? J Health Econ 1988;7:289–90. [18] Diaby V, Perreault S, Lachaine J. Economic impact of Tibolone compared with continuous-combined hormone replacement therapy in the management of climacteric symptoms in postmenopausal women. Maturitas 2007;58:138–49. [19] Goeree R, Hopkins R, Marshall JK, Armstrong D, Ungar W, Goldsmith C, et al. Cost-utility of laparoscopic nissen fundoplication versus proton pump inhibitors for chronic and controlled gastroesophageal reflux disease. A 3-year prospective randomized controlled trial and economic evaluation. Value Health 2011;14:263–73. [20] McIntosh E, Clarke P, Frew AJ, Louviere JJ. Applied methods of cost-benefit analysis in health care. Oxford: Oxford University Press; 2010. [21] McIntosh E. Using discrete choice experiments within a cost-benefit analysis framework: some considerations. Pharmacoeconomics 2006;24:855–68. [22] Zarnke KB, Levine MA, O’Brien BJ. Cost-benefit analyses in the health-care literature: don’t judge a study by its label. J Clin Epidemiol 1997;50:813–22. [23] Ney JP, van der Goes DN, Watanabe JH. Cost-benefit analysis: intraoperative neurophysiological monitoring in spinal surgeries. J Clin Neurophysiol 2013;30:280–6. [24] Glick HA, Doshi JA, Sonnad SS, Polsky D. Economic evaluation in clinical trials. Oxford: Oxford University Press; 2007. [25] Briggs A, Claxton K, Sculpher M. Decision modelling for health economic evaluation. Oxford: Oxford University Press; 2006. [26] Hoeijenbos M, Bekkering T, Lamers L, Hendriks E, van TM, Koopmanschap M. Cost-effectiveness of an active implementation strategy for the Dutch physiotherapy guideline for low back pain. Health Policy 2005;75:85–98. [27] Alemi F, Gustafson D. Decision analysis for healthcare managers. Health Administration Press; 2006. [28] Sonnenberg FA, Beck JR. Markov models in medical decision making: a practical guide. Med Decis Making 1993;13:322– 38. [29] Clarke PM, Gray AM, Briggs A, Farmer AJ, Fenn P, Stevens RJ, et al. A model to estimate the lifetime health outcomes of patients with type 2 diabetes: the United Kingdom Prospective Diabetes Study (UKPDS) Outcomes Model (UKPDS no. 68). Diabetologia 2004;47:1747–59. [30] Lim ME, Worster A, Goeree R, Tarride JE. Simulating an emergency department: the importance of modeling the interactions between physicians and delegates in a discrete event simulation. BMC Med Inform Decis Mak 2013;13:59. [31] Smith KJ, Robert MS. Decision modeling techniques. Pharmacoeconomics: from theory to practice. CRC Press; 2009. p. 17– 36. [32] Diaby V, Laurier C, Lachaine J. A proposed framework for formulary listing in low-income countries: incorporating key features from established drug benefit plans. Pharm Med 2011;25:71–82.
844
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
[33] Laupacis A, Feeny D, Detsky AS, Tugwell PX. How attractive does a new technology have to be to warrant adoption and utilization? Tentative guidelines for using clinical and economic evaluations. CMAJ 1992;146:473–81. [34] Devlin N, Parkin D. Does NICE have a cost-effectiveness threshold and what other factors influence its decisions? A binary choice analysis. Health Econ 2004;13:437–52. [35] Rawlins MD, Culyer AJ. National Institute for Clinical Excellence and its value judgments. BMJ 2004;329:224–7. [36] Baltussen R, Niessen L. Priority setting of health interventions: the need for multi-criteria decision analysis. Cost Eff Resour Alloc 2006;4:14. [37] Husereau D, Boucher M, Noorani H. Priority setting for health technology assessment at CADTH. Int J Technol Assess Health Care 2010;26:341–7. [38] Thokala P, Duenas A. Multiple criteria decision analysis for health technology assessment. Value Health 2012;15:1172–81. [39] Danner M, Hummel JM, Volz F, van Manen JG, Wiegard B, Dintsios CM, et al. Integrating patients’ views into health technology assessment: analytic hierarchy process (AHP) as a method to elicit patient preferences. Int J Technol Assess Health Care 2011;27:369–75. [40] Kluge EH. Resource allocation in healthcare: implications of models of medicine as a profession. MedGenMed 2007;9:57. [41] Merino JG. Clinicians and the economic evaluation of health. Salud Pública de México 2002;44:153–7. [42] Weinstein MC. Should physicians be gatekeepers of medical resources? J Med Ethics 2001;27:268–74. [43] Williams A. The role of health economics in clinical decision-making: is it ethical? Respir Med 1991;85(Suppl. B):3–5. [44] Williams A. Cost-effectiveness analysis: is it ethical? J Med Ethics 1992;18:7–11. [45] Carlsen B, Norheim OF. “Saying no is no easy matter” a qualitative study of competing concerns in rationing decisions in general practice. BMC Health Serv Res 2005;5:70. [46] Lessard C, Contandriopoulos AP, Beaulieu MD. The role (or not) of economic evaluation at the micro level: can Bourdieu’s theory provide a way forward for clinical decision-making? Soc Sci Med 2010;70:1948–56. [47] Beck JR, Pauker SG. The Markov process in medical prognosis. Med Decis Making 1983;3:419–58.
Contents lists available at ScienceDirect
Best Practice & Research Clinical Gastroenterology
1
Introduction to health economics and decisionmaking: Is economics relevant for the frontline clinician? Ron Goeree, MA, Professor a, b, *, Vakaramoko Diaby, PhD, Post-doc fellow a, b a
Programs for Assessment of Technology in Health (PATH) Research Institute, St Joseph’s Healthcare Hamilton, Hamilton, Ontario, Canada b Department of Clinical Epidemiology & Biostatistics, McMaster University, Hamilton, Ontario, Canada
a b s t r a c t Keywords: Health economics Economic evaluation Clinical decision-making Decision-making criteria
In a climate of escalating demands for new health care services and significant constraints on new resources, the disciplines of health economics and health technology assessment (HTA) have increasingly been turned to as explicit evidence-based frameworks to help make tough health care access and reimbursement decisions. Health economics is the discipline of economics concerned with the efficient allocation of health care resources, essentially trying to maximize health benefits to society contingent upon available resources. HTA is a broader field drawing upon several disciplines, but which relies heavily upon the tools of health economics and economic evaluation. Traditionally, health economics and economic evaluation have been widely used at the political (macro) and local (meso) decision-making levels, and have progressively had an important role even at informing individual clinical decisions (micro level). The aim of this paper is to introduce readers to health economics and discuss its relevance to frontline clinicians. Particularly, the content of the paper will facilitate clinicians’ understanding of the link between economics and their medical practice, and how clinical decision-making reflects on health care resource allocation. Ó 2013 Elsevier Ltd. All rights reserved.
* Corresponding author. PATH Research Institute, 25 Main St. W., Suite 2000, Hamilton, Ontario L8P 1H1, Canada. Tel.: þ1 (905) 523 7284x5268; fax: þ1 (905) 522 0568. E-mail address: [email protected] (R. Goeree). 1521-6918/$ – see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.bpg.2013.08.016
832
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
Background A combination of the rise of health care expenditures in many jurisdictions (see Fig. 1), everincreasing technological innovation in health care (especially in drugs and devices), patients’ insatiable needs and demands, the emergence of new diseases and demographic changes [1] have all come together to force decision- and policy-makers, at all levels of the health care system, to rethink reimbursement and provision decisions. With limited capacity of the health care system to handle the accumulation of these factors, health care decision-makers have recognized the need for prioritizing between competing health care uses and the important role a transparent, structured and evidencebased process can play in making decisions. General economic principles offer a theoretical foundation for dealing with resource allocation in an environment of scarcity of resources [2]. The underlying rationale is that, in a free market, resource allocation is regulated by price, ability to pay [3] and perfect information, such that market forces will provide the optimal allocation of health care resources. However, in health care significant market failures exist that prevent a freely functioning market. For example, health care professionals have significantly more knowledge about the potential benefits and risks of health care treatment (i.e. asymmetry of information) and as a result patients place trust in the advice and recommendations provided to them and this can influence patient preferences and decisions [3,4]. This is different than in many other markets like buying a new car where consumers can better judge the benefits to them of car attributes or enhancements and know how much they are willing to pay for these features. Similarly, obstacles to a freely functioning health care insurance market exist such that a number of individuals may not be able to afford adequate health insurance. These, and other market failures in health care, hinder a free market optimal allocation of resources in the health care sector. Therefore, general economic theory is not a viable option for supporting decision-making in health care as it fails to address market failures and society’s ethical attitudes toward health and the provision of health care. The field of health economics (HE) proposes to maximize benefits stemming from the use of health technologies against available resources [3]. As HE recognizes, and attempts to deal with, the various forms of market failure in health care, the use of HE has been legitimated as a framework to guide decisions around health care resource allocation. HE can broadly be defined as a field of economics concerned with the way health care resources are allocated among competing alternatives, at the societal level [5]. HE encompasses many techniques and tools among which economic evaluation (EE) is the most relevant to health care decision-makers. For this reason, the main focus of this paper will be on EE and its role in health care decision-making. According to Drummond et al, EE is ‘the comparative analysis of alternative courses of action both in terms of their costs and consequences’ [6]. It is a widely adopted tool for the assessment of the value of health technologies. This is evidenced by its formal application as part of reimbursement decisionmaking, at the macro and meso levels (political and administrative levels), as well as the abundant production of health EEs in the medical literature. However, concerns have been raised regarding the
Fig. 1. Total expenditure on health care in six countries, members of the Organisation for Economic Co-operation and Development (OECD), as a proportion of gross domestic product. Source: Adapted from OECD Health Data 2012 (http://stats.oecd.org/Index.aspx? DataSetCode¼SHA).
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
833
relevance and ethics of using EE to inform clinical decisions at the micro level [7]. It has been argued that the relevance of EE to decision-makers differs according to the decision-making level [8,9]. For example, many consider it unethical to include ‘costs’ in clinical decision-making as they believe that saving, prolonging or improving patients’ quality of life is priceless. From an economic point of view, health economists tend to consider ‘opportunity costs’, and argue that every decision, even at the clinical decision level, means that there is something else which cannot be done and it is the benefits foregone from this alternative use of time or resources which is the real cost of any decision or action [10]. The argument is that disregarding these alternative uses (which would be accounted for in an EE), would in fact itself be considered unethical from a society point of view. As such, although the role of HE and EE in clinical decision-making might be controversial, this debate stresses the need to explore further the way HE and EE might impact frontline clinician practice. In this paper, we intend to introduce readers to general principles of HE and decision-making. The emphasis will be on EE, presenting the major tools and techniques it covers, however central to this discussion is the relevance of HE more generally to clinical decision-making. The paper is outlined as follows. The first section provides an introduction to HE in general and EE in particular. This introduction will include a discussion of the different types of EE, as well as a discussion about the different approaches used to implement these evaluations. The second section is devoted to linking EE and decision-making, while the third section discusses the potential role of EE for frontline clinicians. And finally, the paper ends with practice points and recommendations in terms of a research agenda. Introduction to health economics As defined in the introduction, HE is a branch of economics that focuses on the field of health care. HE addresses a number of different topics, and there are many papers and books devoted to these topics. One useful summary of the topics was provided by Williams [11], where he distinguished between the following HE topics: U U U U U U U U
Definition of health and estimation of its value Determinants of health other than health care Determinants of the demand of health care Determinants of the supply of health care (behaviour of doctors and health care providers) Economic evaluation Economic equilibrium–market equilibrium Health system analysis Planning, budgeting, and monitoring of health care.
Discussing all the topics above is beyond the scope of the paper. Among these topics, EE methods have received particular attention as these methods have become a requirement in most jurisdictions for supporting publicly financed health technology reimbursement decision-making since the early 1990’s [12]. Consequently, from an HE perspective, EE is considered to be most relevant to clinicians and decision-makers. The next section introduces the different types of EE and the approaches used to implement them. Types of economic evaluations This section is not meant to give a comprehensive presentation of EE methods as many textbooks are devoted to this subject, including the textbook by Drummond et al [6]. Rather, the objective is to briefly cover some key aspects of the different types of EEs. EE attempts to estimate and present the expected costs and outcomes of alternative courses of action. It is generally agreed that the different types of EEs can be subdivided into four categories (see Fig. 2), these are: ‘cost-minimization analysis (CMA)’; ‘cost-effectiveness analysis (CEA)’; ‘cost-utility analysis (CUA)’; and ‘cost-benefit analysis (CBA)’. These methods all compare competing options both in terms of costs and outcomes, with the main difference being the way outcomes are measured, valued and included in the analysis (see Fig. 2).
834
Types of economic evaluations
Cost effectiveness analysis (CEA)
True CEA
Evaluation: Comparison of interventions in terms of costs only as they do not differ in terms of benefits (rarely the case) Costs: Measured and valued monetarily (dollars, euros, etc…)
Cost benefit analysis (CBA)
Cost-utility analysis (CUA)
Evaluation: Comparison of interventions in terms of costs and benefits
Evaluation: Comparison of interventions in terms of costs and benefits
Evaluation: Comparison of interventions in terms of costs and benefits
Costs: Measured and valued monetarily (dollars, euros, etc…)
Costs: Measured and valued monetarily (dollars, euros, etc…)
Costs: Measured and valued monetarily (dollars, euros, etc…)
Benefits: Expressed and valued in natural units (cases detected, events averted, life years gained)
Benefits: Expressed in natural units and converted in Quality Adjusted Life Years (QALYs)
Benefits: Expressed in natural units and valued monetarily
CMA: Cost minimization analysis; CEA: Cost effectiveness analysis; CUA: Cost utility analysis; CBA: Cost benefit analysis Fig. 2. Types of economic evaluations.
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
Cost minimization analysis (CMA)
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
835
Cost-minimization analysis (CMA) A CMA compares alternatives in terms of expected costs and outcomes, but proponents of CMA claim that the outcomes of the alternatives are identical and therefore the focus is on differences in costs only. This type of EE stems from a frequentist interpretation of outcomes, where it is concluded that if there is no statistically significant difference in outcomes (usually at a 5% level of significance), then the conclusion would be that there is no difference between treatments in terms of outcomes. However, this type of approach has met with strong criticism from researchers as the premise of interventions being exactly equivalent in terms of effects is rarely valid, especially from a Bayesian statistical point of view. For example, Briggs and O’Brien postulated that the equivalence assumption is based on an incorrect interpretation of hypothesis tests used to determine whether (or not) the difference between interventions’ efficacy is statistically significant [13]. In addition, since researchers who conduct an EE with an assessment of both costs and outcomes intend to conduct either a CEA, CUA or CBA, several opponents have also argued that a categorization of CMA makes no sense. Nevertheless, this type of EE is still used by some researchers. An example of CMA is shown in the study by Kleus et al 2009 [14]. In this study, the authors aimed to compare two operative techniques: small incision versus laparoscopic cholecystectomy, in a blinded randomized trial. As the authors reported that analysis of the clinical evidence did not favour one alternative over its comparator, they compared both alternatives in terms of costs only. They concluded that the small incision operative technique represents the greatest value for money, whatever the perspective adopted. However as noted above, the use of CMA in the manner conducted by the authors is not encouraged by most health economists. For a full economic evaluation, the authors should have reported more fully on the outcomes, regardless of the statistical significance of the clinical trial findings. Cost-effectiveness analysis (CEA) In early days of EE, CEA was the most common type of EE used in the health care field. With this method, alternatives are compared in terms of costs and outcomes and outcomes are measured and valued in natural units collected in clinical trials or observational studies. Examples of outcome measures used in CEA might include number of events (acute myocardial infarctions or strokes), number of cases detected, symptom free days (SFDs), progression free survival, or overall survival usually expressed as life years (LYs). Since the outcome measure used in a CEA is a natural outcome from clinical trials, conducting a CEA alongside a clinical trial, using the trial’s primary outcome measure for the EE, is common practice. In a CEA, expected costs and outcomes (however measured) are calculated and compared. If one treatment has a higher expected cost and has lower expected outcomes, then that treatment is determined to be dominated by the alternative treatment. However, if there is a trade-off between one treatment costing more but also providing improved outcomes, then the results of a CEA are presented in terms of an incremental cost-effectiveness ratio (ICER), which is literally the differences in mean expected costs divided by the difference in mean expected outcomes. The ICER provides a measure of the expected cost needed to gain a unit of effect (e.g. cost per year of life gained or cost per stroke averted). An example of a CEA was conducted on an early intervention with budesonide in the management of mild persistent asthma symptoms [15]. The comparator was usual care, and the outcome measure was symptom free days (complete 24-h periods with no asthma symptoms). According to a payer and societal perspectives, the authors respectively estimated an ICER of $11.30 and $3.70 per SFD gained. Another example of a CEA was an evaluation of drug eluting stents versus bare metal stents for coronary artery disease where the authors found that the cost-effectiveness of drug eluting stents varied depending on patient risk profiles (e.g. $14,000–$279,000 per revascularization procedure avoided) [16]. Although convenient to apply because the outcome measure used in CEA typically matches outcome measures used in clinical trials, two of the main challenges for decision-makers in using CEA is that the outcome measures can be intermediate in nature like a change in a patient risk factor with questionable impact on more final patient outcomes, and that it is difficult for decision-makers to compare CEA outcomes across diseases and interventions when making health care resource allocation decisions. For example, how would a decision-maker compare a $/stroke avoided versus
836
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
$/revascularization avoided versus $/life year gained for making society health care decisions? This problem of having an intermediate outcome is not unique to CEA and also applies to the assessment of clinical outcomes. As decision-makers request more final outcome measures for clinical assessments, this will also help for EE as the outcome measure used in CEA can be more final in nature. Cost-utility analysis (CUA) In a CUA, expected costs and outcomes for each intervention are calculated, with the outcome measure expressed as quality adjusted life years (QALYs) [6]. Although a more composite outcome measure because it combines quality of length and length of life into a single measure, some consider the QALY simply to be another outcome measures, and as such many consider a CUA to be a variant of a CEA. Still others consider a CUA to be unique in that the QALY can combine various outcomes or impacts (e.g. length of life and quality of length from various health states or events) from a disease or intervention into a single measure of effect which can then facilitate decision-maker comparisons across diseases or interventions. The QALY is calculated by combining weights for the quality of life of being in different health states (i.e. perfect health ¼ 1; death ¼ 0) with the length of time spent in each health state, to obtain an overall measure of length of life weighted by quality of life. Different instruments can be used to determine the quality of life weights (or utility value) [6], but once obtained the results of a CUA are checked for dominance first and if there is a trade-off between higher costs and greater QALYs with an intervention, then the results are presented in terms of incremental cost-utility ratio (ICUR). As CUA allows for the comparison of interventions with different outcomes using a common denominator, it represents a more useful analysis for decision-makers trying to compare across programs and services. CUA has become a standard requirement for EE in many jurisdictions around the world, however, methodological controversy and concern regarding the use of QALYs as a generic measure of health benefits remain [17]. An example of a CUA is presented in the study by Diaby et al [18], where the authors compared hormone replacement therapy (HRT) with a synthetic steroid for the management of postmenopausal women suffering climacteric symptoms. The authors reported an ICUR of $9198 per QALY gained with HRT. Another example of a CUA is from a three-year prospective study comparing medical management to laparoscopic surgery in patients with chronic gastroesophageal reflux disease [19]. In this study the authors reported an ICUR of $29,000 per QALY gained with surgery compared to medical management. As the outcome measure is the same across these two divergent studies, decisionmakers can use the results of these evaluations and combine it with other factors (e.g. ethical and political issues) to better assist them in making resource allocation decisions. Cost-benefit analysis (CBA) In a CBA costs are measured and valued the same way as in CEA or CUA, however, outcomes are measured in natural units and then valued in monetary terms. Since both costs and outcomes are measured in the same monetary terms, it can be determined whether the net monetary value of an intervention is positive or negative. Different techniques are available to value benefits monetarily and include the human capital approach, the stated preference methods and the willingness to pay approach. The details about these methods are provided in the textbooks by Drummond et al [6] and McIntosh et al [20]. A common misconception when using or interpreting CBA, is that if the CBA is greater than zero (i.e. monetary value of benefits outweighs the costs), then this means the intervention should be funded. This misconception ignores the fact that decision-makers cannot afford to fund everything in society (even if CBA > $0), and prioritization and other factors like ethical issues and societal values still need to be considered in decision-making processes. Although CBA is a more flexible tool for incorporating a broader range of potential impacts (both positive and negative) from an intervention [21], there have been methodological concerns raised regarding its application to health care. Major issues with implementing CBA in health care include the complexity of measuring health benefits monetarily and the ethical issues arising from valuing them. For these reasons, the conduct of CBA to inform health care resource allocation is very limited compared to CEA and CUA. It should be noted that a large proportion of studies in the literature claiming to be CBAs are actually either cost analyses or cost-consequence analyses [22].
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
837
An illustration of CBA is provided by the study of Ney et al [23]. In their study, the authors sought to determine whether it was worthwhile implementing an intraoperative neurophysiological monitoring (IOM) for spinal surgeries. The results of the study suggest that, compared to ‘no monitoring’, IOM is implementing because the mean lifetime cost of IOM was estimated at $35,325 compared to $58,514 for unmonitored spinal surgeries. Approaches to implementing an economic evaluation Distinct from the types of EEs discussed in the previous sections, there are two main approaches to conducting an EE: trial-based and modelling-based EEs. These approaches are described in more detail below. Trial-based economic evaluations This approach consists of adding on an economic component to a clinical trial and using both the clinical trial data and economic add-on information to conduct the EE [24]. With this approach, the data proposed to be collected for the clinical trial is reviewed and additional information that would be important to obtain from patients during the prospective phase of the clinical trial for the economic analysis are collected. This information usually consists of health care resource utilization such as hospitalizations, doctor visits, drugs, tests and procedures and the use of other health care programs and services. Depending on the needs of the analysis, other information such as patient out-of-pocket costs or lost productivity might also be collected. In addition to costing information, it may be necessary to consider an alternative or additional outcome measure for the EE. For example, if the primary outcome measure is a more final outcome measure such as mortality/survival, a CEA might be conducted based on a cost per life year saved measure and this might be sufficient for the economic analysis. However, if the primary clinical outcome measure is too intermediate or too specific to the disease, it might be necessary to consider adding on a more broader and final outcome measure for the EE. For example, many reimbursement agencies require a CUA using QALYs as the measure of outcomes, and in that case it might be necessary to collect quality of life information directly on patients during the trial. Conducting a trial-based EE has many advantages. Since information is collected directly on patients in the trial, minimal assumptions are required and the results have high internal generalizability. However, the use clinical trials as the sole source of evidence may contradict with some requirements that must be met for a proper implementation of EEs. Briggs et al alluded to these requirements as being synthesis, incorporation of all relevant comparators, and appropriate temporal framework [25]. Synthesis refers to the imperative to include all relevant data sources in the EE as opposed to a unique source of evidence available through a single clinical trial. Clinical trials also have the disadvantage that they typically compare only two alternatives, while the relevant policy question might involve a comparison of all relevant treatment alternatives. And finally, clinical trials are often shorter term in duration, and may not be adequate for determining the long term costs and outcomes of alternative course of action (especially for chronic diseases). For these reasons, decision analytic modelling techniques are often used as a substitute for, or complement to, trial-based economic analyses. Before discussing the role of decision analytic modelling in EEs, there are two other requirements that should be described which are common to both EE approaches: the perspective of the study; and the quantification of uncertainty. The perspective of the study represents the viewpoint against which costs and effects are identified and valued. Many perspectives can be adopted in an EE, the most commonly adopted perspectives being the third-party payer (e.g. reimbursement agency) and the societal perspectives. It is recommended to present the results of an EE in light of different perspectives so that decision-makers have a better sense of the implications of the results according to broader or narrower perspectives. Another important consideration is uncertainty. As uncertainty is inherent in the collection of data that are used in any evaluation, it is critical to determine how uncertainty could influence the EE results. This is usually addressed using a series of sensitivity analyses where alternative assumptions or values are considered. In this regard, different techniques to handle uncertainty have been developed through years and are discussed in textbooks by Drummond et al [6] and Briggs et al [25].
838
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
An example of trial-based EE is presented in the study by Hoeijenbos et al [26]. The authors conducted a cost-utility analysis to compare the implementation of an evidence-based physiotherapy guideline for low back pain. Based on a randomized trial of 500 patients with back pain, the authors concluded that the evidence-based physiotherapy guideline was not cost-effective compared to the standard guideline dissemination method as no differences were found between both costs and effects of these strategies. Another example of a trial-based EE is the three-year prospective CUA study comparing 104 patients randomized to medical management or laparoscopic surgery in patients with chronic gastroesophageal reflux disease [19]. In this study, the authors reported that surgery was fairly cost-effective with an ICUR of $29,000 per QALY gained. Modelling-based economic evaluations The HE field has witnessed an increasing use of decision analytic models as a tool for informing decision-making under uncertainty. In recent years, there has been a flourishing production of textbooks providing guidance on how to design and use decision analytic models in health EEs [6,25,27]. Decision analytic models can be subdivided into two main categories: cohort and patient-level simulation models (see Fig. 3). Cohort models typically depict the possible consequences that might occur to a group of individuals sharing the same characteristics [25]. The patient experience modelled in this case is that of the average patient. The most common sub-types of cohort models are decision trees and Markov models. Nonetheless, it is not unusual to come across hybrid models, meaning models consisting of a mix of a decision tree for the short term duration of the model and a Markov model for longer term cost and outcome projections. Decision trees represent the simplest form of decision analytic models [28]. A full description of decision trees is presented in the book by Drummond et al [6] and only key features of decision trees are summarized here. Decision trees are built from left to right, and start with a decision node (a square) representing a choice between competing interventions. Following the decision node appear chance nodes (circles), representing points beyond which upcoming events are uncertain. This uncertainty is expressed by assigning probabilities (summing up to one at each chance node) to the occurrence of possible events, which are mutually exclusive. As constructed, a decision tree is analogous to possible clinical pathways that clinicians typically consider for their patients. However, instead of just presenting the possible pathways, in decision analytic modelling the expected costs and outcomes of each treatment alternative are calculated based on the costs and outcomes of each clinical pathway weighted by the probability of going down each pathway. The major advantage of using decision trees is that since pathways are clearly laid out, the model and calculations are transparent for less experience modellers. The major disadvantage of decision trees is that the tree can get very large (i.e. bushy) the more comparators included in the analysis, the larger the number of health states patients can experience and the longer the model is run (e.g. long term chronic disease). Markov models are similar to decision trees in that patients are modelled as being in defined health states over time, but are different in that the clinical pathways for patients are only partially displayed in the model diagram, and key information about how patients progress in the model are dictated though the use of equations, formulas and tables which are largely hidden in the background. This makes Markov models more manageable for complex and chronic diseases, but also makes them less transparent to less experienced modellers. Markov models tend to be used more for chronic diseases where the goal is to model disease progression or risk of events over longer periods of time [28]. The model is constructed based on a finite number of mutually exclusive health or Markov states, over a series of equal time periods referred to as Markov cycles. At the end of each cycle, patients may transition from one state to another or remain in the same health state based on transition probabilities which are estimated from trial data or from longer term observation study data. For each cycle, each health state is associated with a cost and effect/outcome measure. To obtain overall expected costs and outcomes for each treatment alternative, these costs and effects are multiplied by the estimated time patients will spend in each state [25]. Even though it is generally accepted that Markov models are well suited for modelling diseases with on-going risks [28], they have limited ability to structure very complex conditions [25]. This is due to the so called Markov assumption [25,28], which means that
Patient-level models
Cohort models
Decision trees
Simplest form of cohort models; Parsimonious model in terms of data needed to conduct analyses
Markov models
Microsimulation
More sophisticated cohort model as they model chronic and recurrent conditions
Integrated system of equations used to model clinical conditions at the patient level;
Can be used with decision trees
Discrete Event Simulation (DES)
System analysis modeling Modeling competition for resources
Allow for building patient history
*: United Kingdom Prospective Diabetes Study model [27]
Agent-based models
Modeling of biological and physiological processes;
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
Decision analytic models
Fig. 3. Types of decision analytic models.
839
840
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
there are limitations with standard Markov models to account for the past medical history of a patient. Markov models can be made more complex by adding in additional health states (e.g. number of times experiencing an event) to reflect the past history of the patient. An example of a Markov model was the CEA of drug eluting stents versus bare metal stents for coronary artery disease [16], where the trial data was used to estimate short term outcomes and a Markov model was used to estimate longer term costs and outcomes. In this study the authors conducted both a CEA and CUA using trial data for the short term costs and outcomes and a Markov model for estimating longer term costs and outcomes. As shown in Fig. 3, the other main modelling category is patient-level simulation models. Patientlevel simulation models are more complex and more data demanding than cohort models. There are three main sub-types of patient-level simulation models: microsimulation models; discrete event simulation (DES) models; and agent-based models. Microsimulation models allow for variability between patients [25] as patients are considered individually in the model. Each patient travels through the model one at the time, with the model being able to accommodate patients’ individual characteristics and their disease history/progression. An example of a microsimulation model is the United Kingdom Prospective Diabetes (UKPDS) model. It was developed by Clarke et al [29] and further adapted and used to model other conditions. The model consists of an integrated system of parametric risk equations allowing for estimation of seven diabetes complications and their associated costs and outcomes. DES models have the further advantage that they can consider resources and resource constraints. For example, typical cohort models ignore capacity constraints and assume resources can simply be increased. As such these models are ideal for modelling bottlenecks in the system or what might be the implications of increasing health care use beyond capacity (e.g. emergency room visits, surgical suite use). DES models are typically used to model wait times and queuing patterns. For example, the study by Lim M et al [30] provides a good example of a DES. In this study, the authors sought to call decisionmakers’ attention on the importance of accurately modelling physician interactions and roles in which they provide care to patients. The study showed that ignoring constraints and interactions could significantly underestimate estimates of physician time utilization. Finally, agent-based models are sophisticated models in which agents are conscious of themselves and their environment [31]. Agents in these models can be patients, objects or even groups. They have the advantage to model interactions between different agents or entities. In health care, these models are useful to model disease transmission, epidemics or any problems that involve physiological or biological processes. Examples of agent-based modelling in health care can be found in the literature [8,9]. It should be emphasized that in all types of economic models, that the results of the model are dependent upon assumptions made the quality of data inputs used in the model construction. For example, an economic model based on inputs from one clinical trial may produce different results than a model based on inputs from another clinical trial. For this reason, it is encouraged to carefully consider the inputs used in economic models and wherever possible use clinical data that are synthesized from comprehensive literature reviews and meta-analyses. Economic evaluation and decision-making Although geographic variations do exist, health care decision-making regarding the adoption and reimbursement of technologies has traditional been based on four main criteria. These criteria are identified in a literature review [32] comparing the selection processes of reimbursable drugs in public drug benefit plans across high- , middle- and low-income countries. These criteria are safety, efficacy, value for money (cost-effectiveness) and budgetary impact (affordability). Many see these criteria as hurdles, where new technologies must be proven first to be safe, then effective, then cost-effective and then finally affordable. Following considerations of safety and effectiveness, it should be noted that EEs are not meant to be used in a formulaic way. Regardless of the type of EE used (e.g. CMA, CEA, CUA, CBA), if a treatment dominates another (i.e. improved outcomes and lower cost), then that treatment is determined to represent both good value for money and would be affordable from the perspective taken for the
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
841
analysis. It should be noted, however, that because of different budget responsibilities (i.e. budget silos), that even though a treatment may be cost saving from a societal perspective overall, some of the cost savings (or cost offsets) may benefit different ‘payers’ in a health care system, and as such a dominant treatment may still be considered unaffordable to a particular funding agency. For example, this typical occurs with the assessment of new drugs, where the drug results in additional costs to drug program branch and the cost savings (e.g. reduced doctor visits or hospitalizations) are realized by a different budgetary ‘silo’ of the government. In the case of a trade-off between increased outcomes but at a higher cost, then there is always a judgment required regarding both value for money and affordability. In other words, there is no ‘threshold’ where the results of an EE would automatically be considered worthwhile. In a CMA, where outcomes were determined to be identical (rarely if ever the case), there is judgment required regarding the extra cost and extra budgetary impact of the intervention. In a CBA, even if the monetary valuation of benefits exceeds costs, there is still judgment required regarding affordability. And in a CEA or CUA, there is judgment required to determine for example whether the cost per life year gained, cost per stroke avoided or cost per QALY gained represented good value for money and whether the estimated budgetary impact is affordable. Instead of presenting EE results as a cost-effectiveness ratio, some researchers have proposed the use of a measure called the net monetary benefit (NMB) [24], but this measure requires the implicit use of a specific decision-making threshold which many argue should be discouraged in EE. Where judgments are required for determining whether a treatment represents good value for money, some agencies have developed rigid ‘decision rules’ and other agencies have adopted informal guidelines. In Canada for example, Laupacis et al proposed grades of recommendations for the adoption and appropriate use of new technologies based on $20,000 and $100,000 per QALY thresholds [33]. In the United Kingdom (UK), more rigid thresholds have been used (i.e. £20,000– £30,000 per QALY), but empirical and statistical research have shown that the real threshold used by the National Institute for Health and Care Excellence (NICE) in making funding decisions may be in excess of £30,000 [34,35]. As mentioned in the beginning of this section, EE is an important but not sufficient dimension of decision-making regarding the provision or funding of new technologies. In recent years, many jurisdictions have been challenged with the need to expand the four traditional criteria. Some additional criteria proposed include patient preferences, social and ethical values. Until now, the full incorporation of these criteria into existing decision-making processes is in early stages. In this regard, there has been a growing interest into alternative decision-making frameworks, among which multicriteria decision analysis (MCDA) has obtained particular attention. MCDA is both a process and set of methods that explicitly and simultaneously take into consideration multiple decision-making attributes that may conflict to a certain extent [36]. Some jurisdictions, including Canada [37], the United Kingdom [38], and Germany [39] have explored the formal use of this type of decision-making framework. However, research into alternative decision-making frameworks, using EE among other dimensions, is still in its infancy and more work in this area is needed to fully inform decision-making in health care. Is economics relevant for the frontline clinicians? HE, EE in particular, represents a systematic and sound framework for allocating health care resources against a context of limited budget and increasing demand. Its role in informing decisionmaking at the political and local levels has been demonstrated through the past years and is well understood by all stakeholders of decision-making processes. However, the role of EE at the micro level remains a source of debate. Two major views regarding this contentious subject emerge from the literature. The first view supports the statement that using EE in clinical practice conflicts with physician–patient relationship [40–42], and may be considered unethical. Indeed, Merino argues that EEs help identify alternatives that are good value for money at a population level while frontline clinicians are confronted with making decisions at the patient-level [41]. Under the Hippocratic Oath, frontline clinicians seek the best care for their individual patient, regardless of the impact of their decision on the remaining patients seeking care and costs borne by the society as a whole. Additionally,
842
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
the perspectives considered, the preferences and values measured in EEs do not meet the best interests or do not reflect the values of the patients. Following this rationale, making individual clinical decisions in light of the results of ‘population-based’ EEs may be misleading. Ideas opposing the first view have been presented by many authors including Williams [43,44], Carlsen and Norheim [45] and Lessard et al [46]. Lessard et al acknowledge the difficult decisions the frontline clinicians are faced with on a daily basis. They are prone to pressure of time, management of multiple goals while complying with principles of the Hippocratic Oath. However, this situation should not prevent the medical practitioners to make use of evidence produced in EEs. Lessard et al [46] support the idea ‘the real cost of any health decision is the health benefits achievable in some other patient which have been forgone by committing the resources in question to the first patient’ [45]. In this sense, a simple decision to prescribe a new drug or a lab test affects the allocation of other health care resources. Thus, understanding the economic component of decision-making may help practitioners comprehend why a new treatment promoted by a sales representative is not available for a desperate patient in need of new alternatives. On another note, decision analytic models in health care was first used for clinical decision-making and promoted by medical doctors such as Beck and Pauker [47] and Sonnenberg and Beck [28]. This reinforces the idea that economics, in health care for instance, is not about ‘cost per se’, but rather is about opportunity cost. The latter applies both to costs and benefits. The broader picture is that using an economic framework helps ensuring that health resources are allocated in an efficient way, especially in medical practice. As clinicians continue to expand their role beyond provider of care to decision-making bodies and committees, it is imperative that they have a broader understanding of HE, EE and decision-making processes in order assist policy and decision-makers in making the best use of our available health care resources. In this regard, continued education is needed for frontline clinicians about basic economic principles like scarcity and limitation of resources and that in a constrained health care system every decision made by the clinician for every patient they treat has an opportunity cost in terms of what cannot be done for another patient. Not only should training in HE and EE become a mandatory component of education for all health care professionals, but HE and EE should also become a more important component of continuing education (e.g. conferences, journals), re-certification programs and practice guideline development.
Practice points Decision-makers including clinicians are challenged with increasing demands from patients, due to medical (therapeutic) innovation, and scarcity of resources. Inevitably, tough choices have to be made among competing alternatives. HE and EE methods in particular encompasses a set of methods that provides a sound framework to guide resource allocation in health care, including clinical decisions. As clinical practices are settings where these resource allocation decisions occur frequently, it is important for frontline clinicians to understand the concept of opportunity cost inherent in decision-making and incorporate it in their decisions.
Research agenda Decision-making at the clinical level should be investigated to have a grasp of clinicians’ attitude and perceptions towards the role of HE, and in particular EE, plays in their daily practice. Gaps in education and training should be identified and appropriate educational venues need to be developed and implemented.
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
843
Conflict of interest statement The authors declare they have no conflict of interest.
Acknowledgements No funding source to acknowledge.
References [1] Mechanic D, Rochefort AD. Comparative medical systems. Annu Rev Sociol 1996;22:239–70. [2] Sussex J, Towse A, Devlin N. Operationalizing value-based pricing of medicines: a taxonomy of approaches. Pharmacoeconomics 2013;31:1–10. [3] Kernick D. An introduction to the basic principles of health economics for those involved in the development and delivery of headache care. Cephalalgia 2005;25:709–14. [4] Arrow K. Uncertainty and the welfare economics of medical care. Am Econ Rev 1963;53:941–73. [5] Kernick DP. Introduction to health economics for the medical practitioner. Postgrad Med J 2003;79:147–50. [6] Drummond MF, Sculpher MJ, Torrance GW, O’Brien BJ, Stoddart GL. Methods for the economic evaluation of health care programmes. Oxford: Oxford University Press; 2005. [7] Cunningham SJ. An introduction to economic evaluation of health care. J Orthod 2001;28:246–50. [8] Burke DS, Epstein JM, Cummings DA, Parker JI, Cline KC, Singa RM, et al. Individual-based computational modeling of smallpox epidemic control strategies. Acad Emerg Med 2006;13:1142–9. [9] Halloran ME, Ferguson NM, Eubank S, Longini Jr IM, Cummings DA, Lewis B, et al. Modeling targeted layered containment of an influenza pandemic in the United States. Proc Natl Acad Sci U S A 2008;105:4639–44. [10] Tuck J. How is marginal analysis used in economics; 2013 [Ref Type: Online Source]. [11] Williams A. Health economics: the cheerful face of a dismal science. London: Health and Economics; 1987. [12] Hutton J. ‘Health economics’ and the evolution of economic evaluation of health technologies. Health Econ 2012;21:13–8. [13] Briggs AH, O’Brien BJ. The death of cost-minimization analysis? Health Econ 2001;10:179–84. [14] Keus F, de JT, Gooszen HG, Buskens E, van Laarhoven CJ. Cost-minimization analysis in a blind randomized trial on smallincision versus laparoscopic cholecystectomy from a societal perspective: sick leave outweighs efforts in hospital savings. Trials 2009;10:80. [15] Sullivan SD, Buxton M, Andersson LF, Lamm CJ, Liljas B, Chen YZ, et al. Cost-effectiveness analysis of early intervention with budesonide in mild persistent asthma. J Allergy Clin Immunol 2003;112:1229–36. [16] Goeree R, Bowen JM, Blackhouse G, Lazzam C, Cohen E, Chiu M, et al. Economic evaluation of drug-eluting stents compared to bare metal stents using a large prospective study in Ontario. Int J Technol Assess Health Care 2009;25:196–207. [17] Weinstein MC. A qaly is a qaly – or is it? J Health Econ 1988;7:289–90. [18] Diaby V, Perreault S, Lachaine J. Economic impact of Tibolone compared with continuous-combined hormone replacement therapy in the management of climacteric symptoms in postmenopausal women. Maturitas 2007;58:138–49. [19] Goeree R, Hopkins R, Marshall JK, Armstrong D, Ungar W, Goldsmith C, et al. Cost-utility of laparoscopic nissen fundoplication versus proton pump inhibitors for chronic and controlled gastroesophageal reflux disease. A 3-year prospective randomized controlled trial and economic evaluation. Value Health 2011;14:263–73. [20] McIntosh E, Clarke P, Frew AJ, Louviere JJ. Applied methods of cost-benefit analysis in health care. Oxford: Oxford University Press; 2010. [21] McIntosh E. Using discrete choice experiments within a cost-benefit analysis framework: some considerations. Pharmacoeconomics 2006;24:855–68. [22] Zarnke KB, Levine MA, O’Brien BJ. Cost-benefit analyses in the health-care literature: don’t judge a study by its label. J Clin Epidemiol 1997;50:813–22. [23] Ney JP, van der Goes DN, Watanabe JH. Cost-benefit analysis: intraoperative neurophysiological monitoring in spinal surgeries. J Clin Neurophysiol 2013;30:280–6. [24] Glick HA, Doshi JA, Sonnad SS, Polsky D. Economic evaluation in clinical trials. Oxford: Oxford University Press; 2007. [25] Briggs A, Claxton K, Sculpher M. Decision modelling for health economic evaluation. Oxford: Oxford University Press; 2006. [26] Hoeijenbos M, Bekkering T, Lamers L, Hendriks E, van TM, Koopmanschap M. Cost-effectiveness of an active implementation strategy for the Dutch physiotherapy guideline for low back pain. Health Policy 2005;75:85–98. [27] Alemi F, Gustafson D. Decision analysis for healthcare managers. Health Administration Press; 2006. [28] Sonnenberg FA, Beck JR. Markov models in medical decision making: a practical guide. Med Decis Making 1993;13:322– 38. [29] Clarke PM, Gray AM, Briggs A, Farmer AJ, Fenn P, Stevens RJ, et al. A model to estimate the lifetime health outcomes of patients with type 2 diabetes: the United Kingdom Prospective Diabetes Study (UKPDS) Outcomes Model (UKPDS no. 68). Diabetologia 2004;47:1747–59. [30] Lim ME, Worster A, Goeree R, Tarride JE. Simulating an emergency department: the importance of modeling the interactions between physicians and delegates in a discrete event simulation. BMC Med Inform Decis Mak 2013;13:59. [31] Smith KJ, Robert MS. Decision modeling techniques. Pharmacoeconomics: from theory to practice. CRC Press; 2009. p. 17– 36. [32] Diaby V, Laurier C, Lachaine J. A proposed framework for formulary listing in low-income countries: incorporating key features from established drug benefit plans. Pharm Med 2011;25:71–82.
844
R. Goeree, V. Diaby / Best Practice & Research Clinical Gastroenterology 27 (2013) 831–844
[33] Laupacis A, Feeny D, Detsky AS, Tugwell PX. How attractive does a new technology have to be to warrant adoption and utilization? Tentative guidelines for using clinical and economic evaluations. CMAJ 1992;146:473–81. [34] Devlin N, Parkin D. Does NICE have a cost-effectiveness threshold and what other factors influence its decisions? A binary choice analysis. Health Econ 2004;13:437–52. [35] Rawlins MD, Culyer AJ. National Institute for Clinical Excellence and its value judgments. BMJ 2004;329:224–7. [36] Baltussen R, Niessen L. Priority setting of health interventions: the need for multi-criteria decision analysis. Cost Eff Resour Alloc 2006;4:14. [37] Husereau D, Boucher M, Noorani H. Priority setting for health technology assessment at CADTH. Int J Technol Assess Health Care 2010;26:341–7. [38] Thokala P, Duenas A. Multiple criteria decision analysis for health technology assessment. Value Health 2012;15:1172–81. [39] Danner M, Hummel JM, Volz F, van Manen JG, Wiegard B, Dintsios CM, et al. Integrating patients’ views into health technology assessment: analytic hierarchy process (AHP) as a method to elicit patient preferences. Int J Technol Assess Health Care 2011;27:369–75. [40] Kluge EH. Resource allocation in healthcare: implications of models of medicine as a profession. MedGenMed 2007;9:57. [41] Merino JG. Clinicians and the economic evaluation of health. Salud Pública de México 2002;44:153–7. [42] Weinstein MC. Should physicians be gatekeepers of medical resources? J Med Ethics 2001;27:268–74. [43] Williams A. The role of health economics in clinical decision-making: is it ethical? Respir Med 1991;85(Suppl. B):3–5. [44] Williams A. Cost-effectiveness analysis: is it ethical? J Med Ethics 1992;18:7–11. [45] Carlsen B, Norheim OF. “Saying no is no easy matter” a qualitative study of competing concerns in rationing decisions in general practice. BMC Health Serv Res 2005;5:70. [46] Lessard C, Contandriopoulos AP, Beaulieu MD. The role (or not) of economic evaluation at the micro level: can Bourdieu’s theory provide a way forward for clinical decision-making? Soc Sci Med 2010;70:1948–56. [47] Beck JR, Pauker SG. The Markov process in medical prognosis. Med Decis Making 1983;3:419–58.