Introduction to Economic Evaluation
Health Economics, Pharmacoeconomics and Economic
Health expenditure in the United States surpassed $3 trillion in 2014, representing over 15% of the country's GDP, and is projected to reach $3.6 trillion by the end of the year, potentially accounting for 33% of GDP by 2050 Similarly, in the United Kingdom, healthcare spending constitutes about 17% of government expenditure Across developed economies, healthcare is a significant factor in spending, investment, and employment, highlighting the importance of a healthy population for the overall economic well-being of a nation and its citizens.
Concerns exist regarding whether healthcare should be viewed as an economic good, given the finite and scarce nature of healthcare resources, including a limited supply of medical professionals (Morris, 2006; Santerre & Neun, 2009) While some argue that prioritizing treatment for certain populations could enhance overall healthcare access, others contend that healthcare is a fundamental human right and should not be treated merely as a consumer good This discussion, however, is beyond the scope of this book, which focuses solely on the quantitative methods for evaluating the comparative value of new or existing treatments, rather than the ethical considerations of healthcare rationing.
During economic turmoil, healthcare resource constraints force governments to critically evaluate public spending, including the medicines budget Policymakers face the challenge of managing limited budgets to effectively allocate healthcare resources for their citizens In this context, 'payers'—such as governments and health insurance providers—play a crucial role in determining the cost of healthcare provision.
In the context of economic evaluation, the design and analysis of clinical trials must consider that patients are becoming increasingly selective in their healthcare choices due to budget constraints Just as individuals face limitations in fulfilling all their desires due to scarce resources like money and time, healthcare systems also encounter similar challenges in addressing patient demands effectively.
Health economics is the study of the allocation of scarce healthcare resources, such as hospital beds, treatments, and GP time, which have associated costs While health itself is not considered an economic good, the resources required to provide healthcare are For instance, a doctor's visit may be valued at £100 per hour Consequently, health economics encompasses a broader scope beyond pharmaceuticals, influencing government decisions on balancing healthcare supply and demand within budget constraints.
In the pharmaceutical sector, pharmacoeconomics plays a crucial role in comparing new treatments by applying health economics principles to inform policy decisions regarding the supply and demand for medications, especially during clinical trials This field involves the analysis of drug therapy costs to healthcare systems and society, focusing on identifying, measuring, and comparing the costs and outcomes of pharmaceutical products and services (Rascati, 2013) The methodology employed in pharmacoeconomics is known as economic evaluation, which utilizes mathematical and statistical techniques to assess the costs and consequences of various treatment options (Drummond et al., 2005).
Over the past decade, economic evaluation has become a crucial aspect of clinical trials, particularly in Phase III studies, where the inclusion of health economic components has significantly risen Notably, submissions to reimbursement authorities, such as the Scottish Medicines Consortium (SMC) and the National Institute for Clinical Health Excellence (NICE) in the UK, have surged by over 45%, reflecting the growing importance of assessing the value of new treatments.
This book explores economic evaluation within pharmacoeconomics, focusing on the assessment of costs and benefits of treatments derived from prospectively collected clinical trial data Such evaluations are conducted by specialized committees or government agencies, varying by country; for instance, in the UK, this role is fulfilled by NICE Economic evaluation employs various methodologies to determine the value of pharmaceutical interventions and health strategies, with cost-effectiveness serving as the central concept in pharmacoeconomics.
A clinical trial with an economic evaluation requires collaboration among professionals in marketing, clinical research, biostatistics, health economics, and epidemiology to create a plan that demonstrates the new treatment's value for both patients and taxpayers, ensuring successful reimbursement Reimbursement refers to the price that a pharmaceutical company seeks from payers for its new drug treatment; for instance, a company may request £120 per tablet while a payer may counter with an offer of £95 per tablet, influenced by the value evidence provided by the company.
Figure 1.1 illustrates the collaborative functions involved in developing a value proposition, which is a strategic document detailing the approach to demonstrating the cost-effectiveness of a new treatment This proposition also influences the methods for collecting and analyzing clinical trial data.
A multidisciplinary team (MDT) collaborates effectively in designing a Phase III trial that includes an economic evaluation component Clinical research and statistics teams work together to create a robust study focused on key clinical endpoints Meanwhile, the market access team, along with support from health economics and marketing groups, plays a crucial role in ensuring the trial's success.
Clinical research (medical and trial design)
Statistics (trial design and analysis)
Market access and health economics (strategy and analysis)
Marketing (distribution/ sales/message delivery)
Description of a multidisciplinary team for designing a clinical trial with a health economic component.
In the design and analysis of clinical trials for economic evaluation, expert input is crucial for identifying key characteristics of new drugs that can effectively demonstrate their value and enhance market share This includes suggesting clinically and commercially significant secondary endpoints or analyses for comparison with existing treatments Collaboration between market access and clinical teams is essential to establish the necessary evidence for cost-effectiveness, including selecting appropriate comparators that align with country-specific requirements Additionally, statisticians and health economists should work together to agree on pre-specified analyses and modeling techniques, which will optimize the likelihood of successful early reimbursement if the clinical trial objectives are met, particularly if the new treatment proves superior to standard therapies.
Example 1.1: The Importance of Early Health Economic Input
Company Z has launched Drug X, an innovative opioid designed to treat non-malignant pain, with a notable advantage of causing less severe opioid-induced constipation (OIC) compared to standard opioids like Drug S While Drug X offers equivalent pain relief, the company struggled to justify a premium price, as reimbursement authorities concluded that it did not provide sufficient value when compared to the combination of Drug S and a low-cost laxative for managing OIC At that time, there was no trial data available to directly compare the effectiveness of Drug X against the Drug S plus laxative approach.
In Example 1.1, the clinical research and manufacturing teams prioritized creating a high-quality drug that combined laxative and pain relief benefits in a single tablet or capsule However, they neglected to address the value proposition related to reimbursement If the company had taken the reimbursement argument into account sooner, the outcome could have been significantly different.
Important Concepts in Economic Evaluation
Economic evaluations for future decision-making and policy formulation rely on foundational economic theories Analysts may conduct these evaluations without fully grasping their economic principles Similar to how researchers must understand pharmacokinetics, including key parameters like Cmax and area under the curve, analysts need a fundamental comprehension of economic concepts to effectively carry out evaluations.
Economic evaluation is essential for researchers and analysts, as it involves understanding key concepts and terms that influence the assessment process Gaining familiarity with these fundamental economic principles enhances the effectiveness of conducting evaluations, ultimately leading to more informed decision-making.
In the realm of healthcare, the average individual often lacks knowledge about the costs associated with surgeries or treatment plans, particularly in systems like the National Health Service (NHS) in the United Kingdom, where healthcare is perceived as free While many can easily assess the value of items like diamonds or footballs, the disparity in understanding healthcare costs is significant For instance, during a drought, a person might prioritize a basic necessity like water over luxury items, highlighting the varying perceptions of value based on scarcity and need.
Health products and services cannot be easily purchased off the shelf, making their valuation and pricing challenging This complexity applies to both new treatments and intricate procedures like surgery Economists propose that the value of a health item can be assessed based on how much consumers are willing to pay, provided certain market conditions are met Economic theories have been developed to illustrate how value is influenced by the interplay of demand and supply Ultimately, the allocation of health goods is dictated by the price consumers are willing to pay.
In the healthcare sector, the buyer of health products often differs from the consumer, with government institutions typically purchasing on behalf of patients This means that the distribution of healthcare goods is influenced by the prices that governments, funded by taxpayers, are willing to pay However, this dynamic can vary significantly depending on the healthcare system in each country.
Valuing health and products solely based on what individuals are willing to pay overlooks their broader societal impact and overall welfare Wealthier individuals are less affected by price increases compared to those with lower incomes, highlighting an inequity in the system To address these disparities, welfare economics was established to promote fairness and consider the well-being of all members of society.
Welfare economics aims to address scenarios where the allocation of health resources varies significantly, such as when treating one individual with Drug A costs the same as treating three individuals with Drug B Society may perceive the value of Treatment A as inadequate compared to Treatment B This branch of economics seeks to evaluate whether it's possible to assess value based on the pricing of goods, like the costs associated with Treatment A or B.
The design and analysis of clinical trials for economic evaluation aim to identify the most effective distribution of treatments A and B from a societal perspective, a concept known as Pareto optimality.
Paretian economics emphasizes that value is determined by the pricing mechanism of demand and supply, particularly in scenarios where the allocation of Treatment A benefits one patient without harming another For instance, if Treatment A costs £12,000 and consumes all available resources, it leaves other patients worse off, especially if Treatment B costs £4,000 Pareto optimality occurs when one individual is better off without negatively impacting others; however, a situation where the wealthy improve their circumstances while the less fortunate remain unchanged is also considered Pareto optimal, albeit not ideal.
Extra welfare economics has emerged as a significant development in economics, focusing on value within the health sector This concept critiques existing economic evaluation methods, arguing that they may not align with true welfare principles Notable discussions on the validity of current economic evaluation methods can be found in the works of Culyer et al (2011) and Aki et al (1998, 2001) For our purposes, it is essential to recognize that the economic evaluation techniques presented in this book serve as decision-making tools, aimed at identifying which treatments or health technologies provide the most value based on the willingness to pay (WTP) or the cost-effectiveness threshold.
Allocative efficiency involves utilizing evidence from economic evaluations to optimize the allocation of treatments within a specified medicines budget For instance, with a budget of £100,000 for two treatment options, Drug A priced at £2,000 per year and Drug B at £5,000 per year, decision-makers must determine the most effective way to allocate funds, which could result in purchasing 50 units of Drug A or 20 units of Drug B.
A and some of B Exactly what mixture of A and B to spend the £100,000 on is for the decision maker to choose The comparative value that A and B offer will influence this decision.
Chapter 2 outlines the tools utilized in economic evaluation, including cost-utility analysis (CUA), which focuses on achieving allocative efficiency in the health sector This approach aims to ensure that patients receive 'value for money' treatments by efficiently allocating medicines within a given budget constraint.
With a budget constraint of £100 million for cholesterol-lowering drugs, it is crucial to target resources towards specific patient groups, as not all individuals with high cholesterol will benefit equally from the treatment For instance, patients with more severe heart disease are likely to experience greater benefits from lipid-lowering medications compared to those with milder conditions Conducting subgroup analysis in clinical trials can help identify which demographics would derive the most value from the new treatment, ensuring efficient allocation of resources.
Technical efficiency refers to using the least amount of resources, such as the lowest dose or shortest treatment duration, to achieve a specific outcome, like a 20% reduction in lipid levels or the most cost-effective resolution of a peptic ulcer Cost-effectiveness analysis (CEA) is a key tool for evaluating this efficiency, but the term can be misleading as it also serves as a general descriptor for assessing efficiency In this context, economic evaluation is preferred to clarify the distinction between CEA as a specific analytical tool and its broader meaning in healthcare efficiency assessment.
After a pharmaceutical drug receives approval from regulatory authorities like the FDA or EMEA, the pharmaceutical company will then pursue pricing strategies for its newly licensed medication.
Health Economic Evaluation and Drug Development
The traditional drug development process, illustrated in Figure 1.2, highlights the evolving role of pharmacoeconomics in reimbursement and market access Historically, demonstrating the value of new treatments received minimal emphasis, with the focus primarily on conducting Phase I to Phase III trials and securing market authorization Pricing negotiations with individual countries occurred post-approval, without formal requests for evidence of value, as only efficacy was deemed significant, neglecting relative efficacy and costs Consequently, market access primarily involved negotiating a purchase price with government bodies.
The design and analysis of clinical trials for economic evaluation of new drugs typically occur soon after the drug's approval, relying heavily on Phase III clinical trial data submitted for market authorization to inform pricing decisions.
In Germany, the introduction of 'free pricing' allowed pharmaceutical companies to have significant flexibility in setting prices However, the AMNOG law enacted in 2011 limited this flexibility to a one-year period, requiring companies to assess the value for money of new treatments within that timeframe As a result, German payers are increasingly unwilling to cover the costs of expensive drugs that do not demonstrate clear value Oncology drugs, in particular, are affected by the decisions of the Institute for Quality and Efficiency in Health Care (IQWiG), as they are often costly and undergo rigorous evaluations for their value Previously, market access strategies did not prioritize value for money, resulting in less emphasis on presenting health economic arguments and data that highlighted uncertainties in value.
The evolving landscape of drug development necessitates clear evidence of the cost-benefit and value relationship A multidisciplinary team (MDT) is established to identify the requirements for clinical trial data that will support a value-added argument as early as possible While the economic evaluation analysis takes place after the completion of Phase III trial results, the design and planning stages are crucial for integrating value considerations from the outset.
Economic evaluation plays a crucial role in assessing both efficacy and value prior to market access The multidisciplinary team (MDT) fosters collaboration among experts in clinical research, biostatistics, health economics, and other fields to formalize evidence from clinical trials, ultimately demonstrating the product's value for market approval.
A strategic collaboration among researchers in drug development can yield significant outcomes by effectively showcasing the 'value-added' benefits of new medications, which are often essential for securing reimbursement through cost-effectiveness justifications.
Demonstrating efficacy in a randomized controlled trial (RCT) is not the sole criterion for reimbursement and market access; a comprehensive synthesis of existing data is essential to evaluate the value of a new treatment effectively If an innovator drug fails to show value for money to the payer, it may not achieve the desired price While efficacy may still exist without such synthesis, utilizing all available information is crucial to minimize uncertainty in the decision-making process.
Incorporating additional information into drug pricing can lead to uncertainty, particularly when that information is of low quality and varies widely This uncertainty may result in new drugs being priced similarly to inexpensive generics, adversely affecting reinvestment, shareholder returns, and profit margins for pharmaceutical companies Therefore, meticulous planning of value within clinical development programs is essential Utilizing early estimates of incremental cost-effectiveness ratios (ICER) can aid in determining market positioning for new drugs It is crucial to design clinical trials that efficiently balance both clinical and economic objectives, as premature requests for trial data to inform reimbursement strategies could risk unblinding and compromise the trial's primary endpoint.
The relationship between drug development, market authorisation, market access, and reimbursement is crucial for understanding sales dynamics Following licence approval, there is often a lag before reimbursement decisions are made, during which sales typically remain stagnant This stagnation occurs as reimbursement authorities evaluate the value of the new treatment However, once reimbursement agencies confirm that the treatment offers value for money, sales are likely to rise significantly, reflecting the increased acceptance and recommendation of the treatment Additionally, the agreed premium price can further impact profitability.
14 Design & Analysis of Clinical Trials for Economic Evaluation
The revenue loss between market approval and reimbursement decision can be significant, as illustrated by a case where debates over exploratory analyses during a clinical trial delayed reimbursement, leading to millions in lost revenue While regulatory agencies may focus on statistical significance, reimbursement authorities prioritize economic evaluation, making early consideration of pharmacoeconomics in the clinical development program essential Preparing a robust value argument can mitigate the risk of losing market access, even for drugs with questionable efficacy.
Efficacy, Effectiveness and Efficiency
The concept of efficacy is well understood in the context of an RCT Restricted inclusion criteria and very controlled monitoring of efficacy and
Time to reimbursement should be minimised Phase III should consider all plausible questions to minimise further studies to answer reimbursement queries
Relationship between drug licensing, market access and reimbursement (W: sales at some amount £W are flat, after which £K are achieved [after reimbursement] Consequently, the loss in revenue is £K−£W).
Economic evaluation often relies on Randomized Controlled Trials (RCTs), which are regarded as the 'gold standard' for internal validity due to their robust safety outcomes and well-accounted bias (Pocock, 1983) However, RCTs may fall short in external validity, as their findings are typically limited to the specific population studied, leading to challenges in generalizability (Sculpher, 2006).
Figure 1.4 shows the relationship between the drug development process and the activities related to an economic evaluation during each phase of drug development.
Pharmaceutical companies aim to conduct scientifically rigorous trials while navigating profitability and risk in drug development For instance, when developing a transdermal patch for pain relief, it is crucial to address manufacturing challenges and ensure the patch’s performance, such as dose delivery and adherence, alongside proving its superior efficacy compared to competitor products.
In the early and preclinical phases of drug development, the market landscape for new treatments is evaluated, with Phase I and II focusing on early evidence of efficacy to refine cost-effectiveness models Phases III and IV involve updating these models and preparing value arguments, while also assessing the impact on profitability and risk at launch and post-Phase III Importantly, health economic activities are integral throughout the drug development process, not just in Phase III, as early cost-effectiveness assessments can inform pricing strategies and enhance the overall value argument.
Effectiveness is assessed by testing new treatments in real-world conditions, contrasting with the controlled environments of traditional clinical trials A key goal of clinical trials focusing on cost-effectiveness is to evaluate the clinical impact of new treatments in practical settings This includes measuring efficacy in patients with co-morbidities, considering longer follow-up periods, less controlled dosing, and varying levels of patient compliance, which may not have been addressed in Phase III randomized controlled trials While the patient populations may be similar, it's crucial to explore questions like, "How well does the drug perform in everyday practice?"
To effectively evaluate the long-term performance of the new treatment, it is essential to conduct a well-designed randomized controlled trial (RCT) While a 12-month follow-up period post-treatment initiation may provide initial insights, extending this period to 24 months can yield valuable data on the treatment's effectiveness in real-world scenarios This extended duration allows for a more comprehensive understanding of the treatment's impact, especially when certain restrictive conditions, such as double blinding, are eased.
16 Design & Analysis of Clinical Trials for Economic Evaluation
Pr oc es s step Cu stomer analysis (e pidemiolo gy / guidelines ) Ac tivity (+ re view / refinement of pr ev ious ) Input s Ba sis for go/ la unc h de cisio n
Several market attractiveness scenarios can be analyzed to enhance profitability and manage risks effectively By optimizing regional profitability through model forecasts, businesses can align their strategies with market demands Additionally, reviewing forecasts alongside negotiation outcomes ensures informed decision-making and strategic adjustments for sustained growth.
Market analysis (c om pe titive landsc ap e) Op po rt unity as sessmen t
Initial pr od uc t profile Preliminar y test of value sc enario s wi th pa yers an d KO Ls Ri sk as sessmen t
Profitability mo del: manufacturin g co st s analysis an d initial sales & price scenario s
Effica cy & tolerability + burden of illness + cost - effica cy + patien t- re po rt ed outcome s inform the ne w pro duc t profile
Global budget impact mo de l
Pricing and po sitioning test ed wi th ke y stakeholders Pricing scenarios & price volume trade- off s
Price corridor and la unch se quenc e Dra ft global P&R & MA dossier Ag re e lo cal strate gy & adaptation of HE mo dels
Fi nalisation of global P&R & market access dossier Supp or t to lo ca l submissions Monitor P&
R en vir onmen t and changes’ impact on th e pro duc t
Value claims, evidence, and methodology requirements have been tested with stakeholders through market research feedback This information will inform modifications to the development plan, including protocols, clinical research frameworks (CRFs), and target patient demographics Additionally, early estimates from the Incremental Cost-Effectiveness Ratio (ICER) will guide decision-making.
Ta rget Di scover yP re-clinical/Ph as e IP ha se II Ph as e II IF ile & Ne goti ate
Ex plorator y and full discover y Ex plorator y de velopmen t Full de velopmen t P&R negotiation and la unc h • •
The relationship between the drug development process and health economic-related activities is crucial for understanding the overall impact of pharmaceuticals on healthcare systems This connection highlights how drug development not only influences medical outcomes but also affects economic factors such as cost-effectiveness and resource allocation in health care.
A long-term compliance measure is essential for understanding the practical application of new treatments, particularly in maintenance therapy for cancer trials, where costs can significantly inflate the Incremental Cost-Effectiveness Ratio (ICER).
In randomized controlled trials (RCTs), compliance is meticulously monitored, but high compliance rates may be misleading, with actual compliance potentially as low as 60%, contrary to the commonly cited 80% for per protocol analyses While the per protocol population—comprised of patients adhering closely to the study protocol—can influence efficacy outcomes, the financial implications of protocol violators are frequently overlooked Additionally, the intention to treat (ITT) population, which encompasses all randomized patients, may not effectively evaluate treatment effectiveness due to restrictions imposed by inclusion and exclusion criteria.
In clinical trials aimed at cost-effectiveness, all patients randomized are included in the analysis, regardless of their treatment adherence These trials focus on capturing resource utilization and the associated costs for each treatment group.
Table 1.2 illustrates the connection between study objectives and key features of various study types Randomized Controlled Trials (RCTs) are considered the 'gold standard' for confirming efficacy, focusing on primary efficacy and safety outcomes, with flexible time frames ranging from short to long term However, longer trials, particularly in cancer and cardiovascular research, can incur high costs In contrast, pharmacoeconomic studies are tailored to assess efficiency, utilizing a combination of data from RCTs and observational studies These studies gather outcomes such as resource use, quality of life, and compliance to evaluate efficiency Additionally, a hybrid approach may be employed to maximize the effectiveness of a single trial.
Relationship between Types of Study and Design Features
Study Type Clinical Trial Outcomes Research Pharmacoeconomic
Design RCT Observational and RCT RCT, observational and various others
Measures Efficacy and safety Patient-based outcomes Resource use and outcomes
Time frame Short term or long term Long term Long term
Based on Ideal clinical practice
(restricted population) Normal clinical practice
(wider population) Normal clinical practice
18 Design & Analysis of Clinical Trials for Economic Evaluation
When Is a Pharmacoeconomic Hypothesis Possible?
Table 1.3 illustrates the connection between the anticipated clinical benefits in a clinical trial and their potential implications for a pharmacoeconomic hypothesis that demonstrates value A pharmacoeconomic hypothesis can typically only be formulated if a plausible clinical advantage exists However, certain clinical trials, like equivalence trials, do not aim to establish a clinical benefit, as they assess treatments that are expected to perform similarly to existing standard treatments Bioequivalence trials fall under this category, and despite variations in administration methods (such as intravenous versus oral dosing), they are not suitable for economic evaluations due to their reliance on healthy volunteers and the absence of direct clinical benefit assessment.
A superiority trial demonstrates that a new treatment is significantly better than the standard treatment This is determined by the mean treatment difference, ΔA−S, which quantifies the average difference between the new treatment (A) and the standard treatment (S) For a treatment to be considered superior, the 95% confidence interval (CI) for ΔA−S must not include the value 0 For instance, a CI range of [1.2–6.2] indicates statistical significance, confirming that the new treatment is indeed superior to the comparator.
To establish a cost-effectiveness hypothesis, it is crucial for the value of ΔA−S to be sufficiently significant This value proposition may hinge on the observed disparities in mean costs between Treatments A and B, in relation to the average difference in outcomes.
Relationship between Clinical Objective and Plausible
Clinical Advantage Possible Pharmacoeconomic Hypotheses
Superior efficacy Saves life years
Averts disease Improved QoL/QALY gain Better side-effect profile Improved QoL
Change in half-life More convenient administration
Improved compliance Improved QoL/QALY gain Improved delivery Improved compliance
Economic evaluation in healthcare often involves assessing the significance of treatment effects alongside their costs For instance, a large trial with 2000 patients may show a statistically significant mean difference of 0.3 mmHg (95% CI: 0.1–0.5), indicating a measurable treatment benefit However, this small difference raises questions about its cost-effectiveness A significant treatment effect does not guarantee clinical value if the associated costs are high Conversely, even a substantial treatment benefit may not justify the costs incurred, highlighting the need for a thorough cost-effectiveness analysis in clinical decision-making.
Example 1.3: Statistical Significance versus Economic Relevance
In a comparison of treatment costs and effectiveness, a difference of 0.3 mmHg with an additional cost of £10,000 for Drug A results in a cost-effectiveness ratio of over £33,000 per unit reduction in blood pressure Conversely, if the effectiveness of Treatment A is improved to a 0.6 mmHg reduction but incurs an increased cost of £25,000, the cost-effectiveness ratio worsens to more than £41,000.
A non-inferiority trial assesses whether a new drug's average treatment benefit is not clinically worse than an existing treatment In this context, the new drug is deemed equivalent to the standard treatment if the lower or upper limit of the 95% confidence interval for any observed difference exceeds a predefined threshold, represented as -K or +K For instance, these thresholds could be -4 and +3.4 If the observed treatment effects are likely to fall within these limits, the new treatment can be classified as non-inferior The interpretation of what constitutes 'worse' or 'non-inferior' is ultimately subjective and depends on clinical judgment.
There are two particular cases, important for economic evaluation, which concern non-inferiority: (i) where ΔA−S > 0 and (ii) where ΔA−S < 0 In case (i),
Treatment A worse than S Treatment A superior to S
In a non-inferiority trial comparing Treatment A to Treatment S, the average improvement observed for Treatment A was 0.3 mmHg While this result does not demonstrate superiority over Treatment S, it confirms non-inferiority, as the lower 95% confidence limit of -2.8 mmHg remains above the established non-inferiority threshold of -4 mmHg.
In the design and analysis of clinical trials for economic evaluation, it is essential to assess the average treatment benefit, which can vary between different treatment options In scenario (i), Treatment A demonstrates a positive average benefit over Treatment S, indicating its superiority Conversely, in scenario (ii), Treatment A shows a lesser mean effect compared to Treatment S Although the difference (ΔA−S) in scenario (i) may not be substantial enough to declare Treatment A as superior, it still presents a slight clinical advantage on average This nuanced understanding is crucial for informed decision-making in clinical trial evaluations.
Example 1.4: Superiority and Noninferiority in the Context of Cost-Effectiveness
In a respiratory study comparing treatments A and S, the observed treatment difference in blood pressure is 0.3 mmHg, with a 95% confidence interval (CI) ranging from −2.8 to +3.4 mmHg To conclude superiority, the 95% CI must exclude zero and indicate a positive treatment benefit Although the average benefit is 0.3 mmHg, the presence of zero within the CI means superiority is not achieved However, the lower limit of the 95% CI at −2.8 mmHg exceeds the non-inferiority threshold of K = −4 mmHg, indicating that non-inferiority conditions are met If the lower limit had fallen below K = −4, Treatment A would not have satisfied the non-inferiority criteria.
In Example 1.4, the average treatment effect indicated a minor benefit, although it was not superior Despite this small and potentially clinically irrelevant advantage, it remains possible that the quality of life (QoL) experienced with Treatment A was better than that of Treatment S From a clinical standpoint, a new treatment providing only a slight benefit may be deemed clinically irrelevant, as the primary goal is to achieve non-inferior equivalence.
From an economic evaluation standpoint, even minor benefits can hold significance, as assessments focus on mean effects The 95% confidence interval for the mean difference is not deemed crucial for economic evaluation, rendering statistical inference irrelevant (Claxton, 1999) Uncertainty in point estimates is addressed through sensitivity analysis, as discussed in Chapter 7 Ultimately, economic evaluation calculates health benefits in relation to costs by considering these mean effects.
A new drug can be considered cost-effective even when aiming for non-inferior equivalence, especially if it offers additional benefits like improved safety and compliance The economic rationale for adopting a new treatment goes beyond efficacy, incorporating various factors This section highlights that even a clinically insignificant advantage, which narrowly misses demonstrating superiority in a non-inferiority trial, can still present opportunities for showcasing economic benefits.
1.5.1.3 Non-Inferior Equivalence Where Δ A−S < 0 (Case II)
In this situation, the objective is again non-inferiority; however, the new treat- ment is not only shown to be non-inferior to the standard treatment, but also, on
Economic evaluation introduces the concept that Treatment A has a marginally worse effect compared to Treatment S, indicated by a difference (ΔA−S) that is either zero or negative Clinically, this slight disadvantage of Treatment A is deemed irrelevant, emphasizing the importance of considering both statistical and clinical significance in treatment comparisons.
The concept of 'worseness' relies on the 95% lower confidence limit rather than the mean difference In this context, it suggests that the new treatment is unlikely to provide a clinical advantage Any value proposition is likely to hinge on offering equivalent treatment benefits at reduced costs or improved safety.
Example 1.5: Equivalent Effects in the Context of Cost-Effectiveness
The standard treatment for infection involves a twice-a-day regimen; however, a new once-a-day modified-release formulation offers a more convenient alternative This once-a-day option is expected to enhance patient compliance and reduce costs, justifying a premium price from the manufacturer due to its added value While the treatment effects of the once-a-day regimen may be similar to or potentially inferior to the twice-a-day approach, the lower associated costs suggest that the new formulation could be more efficient, assuming comparable efficacy.
An example of this situation might be a twice-a-day form of clarithromy- cin (an anti-infective drug) versus a modified (once-a-day) formulation.
Health Economic Evaluation Concepts
Incremental Cost-Effectiveness Ratio (ICER)
In economic evaluation, the main results are usually reported in one of two ways:
1 The incremental cost-effectiveness ratio (ICER)
2 The incremental net monetary benefit (INMB)
In this chapter, the relationship between ICER and INMB is explored In Chapter 1, the ICER was informally introduced as costs relative to benefits
We now formally present the ICER in the context of the cost-effectiveness plane, which is how the results of an economic evaluation are often reported and interpreted.
The ICER is defined as
The equation Δc = ΔμA−S represents the mean difference in costs between Treatments A and S, while Δe = ΔɛA−S indicates the mean difference in effects between these treatments In this context, μA and μS denote the mean costs of Treatment A and Treatment S, respectively, and ɛA and ɛS refer to the mean effects of Treatment A and Treatment S.
In Equation 2.1, the numerator, μA − μS, represents the incremental cost, while the denominator, ɛA − ɛS, denotes the incremental effectiveness Here, A refers to the new drug treatment, and S signifies the standard treatment This ratio is crucial for evaluating the cost-effectiveness of the two treatments.
The design and analysis of clinical trials for economic evaluation focus on the cost-effectiveness ratio (Δc/Δe) and the uncertainty surrounding it, which are crucial for assessing economic outcomes This ratio is visually represented on the cost-effectiveness plane, as illustrated in Figure 2.1.
In Figure 2.1, the x-axis illustrates the incremental effectiveness, represented as the mean difference in effects between treatments A and S (Δe) Positive Δe values indicate that the new drug is more effective, while negative values suggest lower effectiveness Effectiveness, as defined in Equation 2.1, may not equate to efficacy; it can encompass a combination of efficacy measures, such as survival time and quality of life (QoL), resulting in quality-adjusted life years (QALYs) Furthermore, Δe can be quantified in natural units, like the difference in blood pressure measured in millimeters of mercury (mmHg), or expressed proportionally on a scale from 0 to 1.
In Figure 2.1, the y-axis represents the mean difference in costs (Δc), expressed in pounds sterling For instance, if Treatment A (the new drug) is £2000 more expensive than Treatment S (the standard), Δc would be +£2000 Conversely, a negative Δc indicates that Treatment A is, on average, less costly than Treatment S.
In Quadrants 2 and 4, determining the cost-effectiveness of treatments is straightforward If the ICER value falls in Quadrant 2, the new treatment is both cheaper and more effective, representing the optimal situation for pharmaceutical companies Conversely, Quadrant 4 indicates a less favorable outcome where the new treatment is more expensive and less effective It is important to note that the ICER values can be influenced by adjusting parameters like pricing.
New more effective New less effective
4 Increased efficacy not worth increase
Reduced efficacy worth the reduction
Health economic evaluation concepts play a crucial role in assessing new treatments To improve the Incremental Cost-Effectiveness Ratio (ICER) and potentially shift it to a more favorable quadrant, strategies such as reducing the price or enhancing efficacy may be considered, even if it results in lower profits However, a new treatment with an ICER in Quadrant 4 is unlikely to be perceived as valuable, and simply adjusting the price does not mitigate the issue of its inferior efficacy.
Example 2.1: Interpreting the Cost-Effectiveness Plane
Referring to Figure 2.1, we note that in Quadrant 2, the point Q (2,
A recent analysis indicates that a new drug demonstrates superior effectiveness, providing an improved effect of 2 while being £10,000 cheaper than the standard treatment This results in an Incremental Cost-Effectiveness Ratio (ICER) of -£5,000 per unit of effect, such as Quality-Adjusted Life Years (QALYs) Consequently, the new treatment is considered to dominate the standard treatment Decision-making regarding the ICER primarily focuses on Quadrants 1 and 3, with particular emphasis on Quadrant 1, where the justification of value is often sought.
The line through the origin, known as the willingness to pay (WTP) or cost-effectiveness threshold, represents the maximum amount a payer is willing to spend on a new drug Incremental Cost-Effectiveness Ratios (ICER) calculated to the right of this line indicate that the new treatment is cost-effective For instance, with an incremental effect (Δe) of 2 and an incremental cost (Δc) of -£10,000, the ICER is calculated as -£5,000, positioning it in Quadrant 2 Conversely, if the new treatment demonstrates lower efficacy, resulting in a value of -2.8, the ICER would shift to £3,571, moving the analysis from Quadrant 2 to Quadrant 3.
Example 2.2: Changing the Cost-Effectiveness Threshold
Figure 2.2 illustrates two cost-effectiveness thresholds: λ (dashed line) at £30,000 and λ* (solid line) at £12,000 The new treatment, which differs from the one in Example 2.1, initially demonstrated a treatment benefit of 2 units but came at a higher cost of £28,000.
In Figure 2.2, point Z, initially positioned below the cost-effectiveness threshold of £30,000, shifts above the new threshold of £12,000 as λ changes The incremental cost-effectiveness ratio (ICER) is calculated as £28,000 divided by 2, resulting in £14,000, which is positioned above the new threshold Consequently, Treatment A is deemed not cost-effective since point Z exceeds the new cost-effectiveness line.
In general, in Quadrant 1, if Δc/Δe λ and as long as Δe > 0, the new treatment is cost-effective; values of Δe > 0 suggest a benefit with the new treatment In
28 Design & Analysis of Clinical Trials for Economic Evaluation
Quadrant 3, Δc/Δe is always ≥0 (for Δe ≠ 0), so the ratio is < λ, and the new treatment is considered cost-effective.
Incremental INMB
Historically, decisions regarding cost-effectiveness were based on the Incremental Cost-Effectiveness Ratio (ICER) and its associated uncertainty, typically represented by a 95% confidence interval (CI) However, assessing the uncertainty of the ICER presents several challenges One major issue is that Y, the ratio used in the calculation, can exhibit problematic statistical properties, complicating inference For instance, when Δe is zero or nearly zero, as seen in equivalence trials, the ICER can become extremely large or even infinite Additionally, negative ratios further complicate the interpretation of Y, particularly when comparing multiple treatments, as it requires ranking several ratios to determine dominant options.
The main approaches to addressing how to provide a measure of uncer- tainty around the ICER are
1 Taylor’s Expansion to estimate the variance of the ICER as a ratio of two quantities (involving a complicated equation)
New more effective New less effective
New less costly λ = WTP = £30,000 λ∗ = WTP = £12,000 +∆ c (£)
Cost-effectiveness plane: Changing WTP/CE threshold.
2 Fieller’s theorem, which can result in wider CIs for the ICER (dis- cussed in Appendix 2A.1)
3 Bootstrapping, a resampling method to estimate the statistical prop- erties of the ICER (see Example 2.5)
4 The INMB approach, which removes the difficulties of statistical inference
The interpretation of the Incremental Cost-Effectiveness Ratio (ICER) can be challenging, particularly when confidence intervals (CIs) are involved A negative ICER may result in a 95% CI that ranges from a negative lower limit to a positive upper limit, which is not uncommon in clinical trials focused on efficacy endpoints For instance, an analysis of forced expiratory volume (FEV1) might produce a 95% CI of (−2.5 to +3.4), indicating that there are no significant mean treatment differences for FEV1 In this context, the negative value does not complicate the clinical effect interpretation; it simply suggests that the treatment is less effective.
Cost-effectiveness ratios can be complex, as average point estimates often deviate from a value of 1, indicating no mean differences These ratios may even be negative, as previously mentioned For instance, a 95% confidence interval for an Incremental Cost-Effectiveness Ratio (ICER) of £2000, ranging from −£6000 to £1500, suggests that the new treatment could either be less expensive and more effective (−£6000) or more costly yet still more effective (£1500) The negative value of −£6000 does not clarify whether the numerator or denominator is negative, complicating interpretation, especially with multiple treatment comparisons In quadrant 2, the new treatment may be viewed as cheaper and more effective or as more expensive and more effective by £1500 per unit of effect.
It is possible to encounter two cost-effectiveness ratios with identical negative values, yet they may convey entirely different implications For instance, an ICER of −£200 could be positioned in either Quadrant 2 or Quadrant 3 In Quadrant 2, this ICER indicates that the new treatment is superior, being both more cost-effective and more effective, while in Quadrant 3, it suggests a different interpretation.
3, it could be interpreted as being more expensive with worse effective- ness One should therefore exercise caution when interpreting CIs for cost- effectiveness ratios.
The Incremental Net Monetary Benefit (INMB) is a favored method for showcasing economic evaluation results and the associated uncertainties According to Stinnett and Mullahy (1998), the INMB modifies the scale of the cost-effectiveness ratio, enhancing the clarity and utility of economic assessments.
30 Design & Analysis of Clinical Trials for Economic Evaluation thenΔà A−S < Δε A−S ì λ(simple algebraic manipulation) Δà A−S − Δε A−S ì λ 0, using earlier notation.
As long as the INMB is >0, the new treatment is considered cost-effective.
Example 2.3: Computing the Incremental Net Monetary Benefit
In a scenario where the difference in effects (Δe) is 0.9 and the difference in costs (Δc) is £5,000, with a willingness to pay (WTP) of λ = £30,000, the Incremental Net Monetary Benefit (INMB) is calculated as 0.9 × £30,000 − £5,000, resulting in an INMB of £22,000 This positive INMB indicates that the new treatment provides a net benefit valued at £22,000, suggesting that the Incremental Cost-Effectiveness Ratio (ICER) is below the WTP threshold Additionally, the calculation of Y, defined as ΔμA−S/ΔɛA−S = Δc/Δe, yields a value of £5,555, further supporting the treatment's cost-effectiveness.