9

Perhaps more than most medical specialties, the field of radiology is faced with economic pressures. The acquisition of imaging devices, the use of those devices for screening or diagnosis, the use of radiotherapy in oncology, the personnel dedicated to operation, maintenance of radiotherapy and imaging equipment, and the need to upgrade and expand use are common considerations in the life of a radiologist.1

Once reserved for specialized consultation in rare cases, MRI has become routine in the diagnosis of a wide range of disorders. There were an estimated almost 2 million MRI exams performed in Canada in 2014 to 2015.² This is 20 times more than the number performed in 1990. One might wonder if this increased use has resulted in a concomitant reduction in the use of alternative imaging approaches such as computed tomography (CT). In fact, 5.3 million CT exams were performed on Canadian patients in 2014–2015, 1.8 times more than the number of exams performed in 1990.2 In the same year, the rate of MRI exams performed ranged from a low of 33 per 1,000 persons in British Columbia to a high of 55 per 1,000 persons in New Brunswick. CT exams ranged from 91 per 1,000 persons in Alberta to 174 per 1,000 persons in New Brunswick.²

These statistics prompt a number of questions for the budding radiologist-health economist. Does the relatively high rate of imaging in New Brunswick result in better health outcomes for residents of this province compared to residents of other provinces? What is the opportunity cost of imaging in New Brunswick? Is less money being spent on other health services resulting in poorer health outcomes in other clinical areas? Are radiologists in Canada performing an MRI when a less expensive CT might suffice? Are they ordering excessive imaging exams? Are choices between different imaging approaches made with evidence of value for money?

All of these questions are relevant to the field of economic evaluation. Indeed, questions of economic efficiency apply not only to radiation therapy and diagnostic radiology, but to population screening such as mammography to identify breast tumors or bone density screening to identify women at risk for osteoporosis. With an aging population, the demand for radiation therapy for cancer treatment and imaging for screening and diagnosis is expected to increase, underscoring the critical importance of making the wisest allocation decisions possible to ensure the maximum health benefits are achieved for the money spent.

In summary, the goals of economic evaluation are:

• To increase efficiency in the selection of treatments/diagnostic strategies

• To promote optimal use of available therapies

• To understand the relationship between health care system policies that invest in health interventions, services and programs, and patient health outcomes

• To enhance the health status of target populations

• To achieve a net welfare gain in society through improved health of the population

It is equally important to understand that economic evaluation is not a form of cost containment or cost reduction nor is it a specific health policy. Health economic evaluation involves the use of rigorous methods to produce evidence to facilitate decisions regarding investment in health care resources. The ultimate goal of health economic evaluation is to inform health care decision making so that the greatest health benefits can be achieved for any given investment.

Concepts

Defining Economic Evaluation

An economic evaluation is defined as a comparison of two or more interventions in terms of both costs and health consequences (Fig. 9.1).

In the comparison of interventions, the inputs are the costs associated with health care resources consumed. Health care resources may include physician consultations, imaging tests, technician time, surgical procedures, and use of medical devices. These are weighed against the outputs, which are the changes in health status that are a consequence of using the health interventions of interest. Measuring the volume of resources, such as the number of imaging tests performed, is never the endpoint in an economic evaluation. Health resources are consumption items that must always be weighed against the output or health consequences, which are the observed changes in health status.

The interventions being evaluated are often a new or emerging treatment that is compared to standard care, or they may be an array of possible approaches compared to standard or usual care. Sometimes an intervention represents an entirely novel treatment for a condition for which no treatment previously existed. In this case the appropriate “usual care” comparator may simply be periodic monitoring and may result in poor health outcomes. It is important to remember that there must always be a comparator stipulated even when standard care does not exist. A “do nothing” approach will still result in resource use (costs) and declining health status (con-sequences) as the patient deteriorates. These costs and consequences must be measured for comparison to the alternative intervention.

Fig. 9.1 Balancing costs against health benefits. Economic evaluation is a comparison of two or more alternative treatments in terms of costs and consequences.

There are four different types of analytic approaches to economic evaluation. These four approaches are listed and defined in Table 9.1.

The four types of economic evaluations are cost-effectiveness analysis (CEA), cost–utility analysis (CUA), cost–benefit analysis (CBA), and cost–minimization analysis (CMA). When comparing two or more interventions, one must collect all of the resources used (costs) and all of the health status changes (consequences) for all the interventions being compared, regardless of the analytic approach. All four approaches measure health care resource use in terms of monetary units, such as dollars. The four approaches differ in terms of how health outcomes are specified.

The most common type of economic evaluation, CEA, measures health outcomes in natural health units. Examples include how many people live or die, the number of years of additional life achieved for each intervention, or the number of adverse events or complications observed for each intervention. The precise choice of health status measure will usually vary between medical specialties and often varies from study to study. This can complicate comparisons between studies within a therapeutic area and will make it difficult, if not impossible, to make allocation decisions across therapeutic areas. Should one spend more money on mammography (number of cases of breast cancer detected) or treatment of depression (number of patients achieving good mental health)? CEA is explained in greater detail in the cost-effectiveness analysis section of this chapter.

Table 9.1 Types of economic evaluations

The problem of disparate outcomes is solved by using a common metric. In CUA the quality-adjusted life year (QALY) is used. The QALY is a composite measure that considers not only the life years achieved for any given intervention, but also the health-related quality of life that the patient experiences during those life years. Thus, while a CEA might indicate that a liver transplant and medical treatment of advanced breast cancer may result in the same number of life years for the two patient groups, only a CUA can reveal that the surgical patients also have a much better quality of life. The preference for the higher quality of life state, or utility for the health state, is used as weight to adjust the observed life expectancy. Knowing that there is an improvement in quality-adjusted life years, not just the number of life years, is an important consideration when making an allocation decision. A significant advantage of the QALY is that it allows comparisons not only across different studies, but across very different patient groups and therapeutic areas. For that reason it is considered a universal outcome measure and is recommended by economic evaluation guidelines.3,4 Despite the allure of the universal metric, CUAs are often difficult to carry out as available health-related quality of life measures may not have the appropriate performance characteristics or sensitivity to detect changes in health states in some patient populations.

The terms “cost–benefit” or “cost–benefit analysis” are often used by lay persons to vaguely denote any type of economic evaluation. This is an unfortunate misuse of the term, since CBA has a precise meaning in economic evaluation. Only a CBA attempts to convert health status changes into dollar terms. More than that, a well-executed CBA will consider nonhealth benefits, such as better workforce participation resulting from improved health, as well as benefits to other individuals (spillover effects), such as improvements in the quality of life of the patient’s family members. Because the benefits measured are so wide ranging, of the four analytic approaches, only a CBA can claim to be founded in welfare economics.⁵ That being said, the methods required to convert health effects into dollars are complex and often difficult to undertake. For that reason, there are few CBAs in health care.

The final approach listed is CMA. In a cost-minimization analysis only the costs are compared between two or more interventions. This is not because the outcomes are not considered relevant or important to the research question, but because existing high quality, reliable evidence demonstrates that the two interventions may result in comparable levels of effectiveness. One cannot proceed with a CMA based on the assumption of equivalent health status improvement—there must in fact be evidence of the case. In reality, few studies are labeled as CMAs because although the observed difference in health status between two interventions may be below the threshold deemed clinically important, the difference may nevertheless have relevance for the weighing of costs against health benefits. CMA remains the least common of the analytic approaches.

The fundamental defining concept that an economic evaluation compares two or more interventions in terms of costs and health consequences is often illustrated with the use of the cost-effectiveness plane. As seen in Fig. 9.2, costs are plotted on the y-axis and health effects on the x-axis. The plane is divided into four quadrants, which can be labeled for convenience as the compass points northeast, southeast, southwest, and northwest.

If compared to standard care a new intervention is more costly and not as effective (lies in the northwest quadrant), then it is dominated by standard care and is clearly not a worthy investment. If the intervention saves money and results in better health outcomes compared to standard care (lies in the southeast quadrant), then it is surely worth adopting. The remaining two quadrants are the ones for which economic evaluation is most critical: when the intervention is less costly but also less effective (southwest quadrant) or when the intervention is more costly but also is more effective than standard care (northeast quadrant). In a majority of cases, detailed economic evaluations are carried out because the novel intervention is more costly but promises better effectiveness compared to standard care. The salient questions are: How much more costly? How much more effective? Is the added cost worth the added health benefit? This is the crux of economic evaluation. The basic methods are further delineated in the sections to come.

Fig. 9.2 The cost-effectiveness plane. Compared to standard care, a new intervention can be more costly but more effective (northeast quadrant), less costly and more effective (southeast quadrant), less costly and less effective (southwest quadrant), or more costly and less effective (northwest quadrant).

The example in Box 9.1 illustrates how to define a research question in an economic evaluation with an example from diagnostic radiology.⁶

Costing

An economic evaluation weighs the additional costs of a new intervention against any added benefits compared to standard care. While “benefits” may be represented in terms of a single effectiveness variable such as life years or quality-adjusted life years, the cost consists of many cost item variables, all of which must be properly combined to accurately represent the costs of an intervention. The costing process is comprised of several steps and requires extensive data collection.7 These steps are:

Although not comprehensive, Table 9.2 shows examples of typical direct health care cost items that could be considered in an economic evaluation of a radiological intervention.

Few health care systems cover 100% of all the health care costs for patients. Depending on how the health plan benefits are structured, patients may incur some out-of-pocket costs to receive care regardless of whether the health care plan is publicly or privately financed. Many health plans require patients to make copayments for physician services, emergency services, ambulance services, in-patient stays, and medications. Copayments may be fixed or a percent of the total cost, depending on the cost item and the health plan benefits. In addition, patients routinely pay out-of-pocket for items that are not covered by their health plans, such as over-the-counter medications, medical appliances, and transportation to access or receive care. Table 9.3 indicates the typical direct patient cost items that might be considered in an economic evaluation of a radiological intervention.

Table 9.2 Examples of direct health care cost items

Table 9.3 Examples of direct patient cost items

Out-of-pocket costs can be considerable and pose a financial burden for some patients and their families over a short or long interval.8 A large out-of-pocket expenditure burden may force patients to make choices that affect the health services used by other family members. It is thus important that these cost items not be overlooked.

The third category of cost items are the indirect costs, also referred to as productivity costs. In addition to paying out-of-pocket for some or all health care, patients, and often members of their family, miss time from work and other activities due to their illness or the need to receive care. These time losses too can be substantial. Perhaps the greatest cost stemming from time loss is due to premature death. Long-term disability due to illness also results in long-term productivity losses. Failing to account for these significant costs can lead to misclassifying an intervention that is cost saving as one that is cost-incurring. Table 9.4 lists some of the types of time losses that result in productivity costs.

It is important to note that often the patient’s family members and informal caregivers will incur indirect costs for the care of the patient. This is particularly true for pediatric and other dependent patient populations. The indirect costs of family members and other informal caregivers should be measured and can be ascribed to the patient for the purpose of the analysis.

Table 9.4 Examples of indirect cost items

In addition to the above cost categories, there are also intangible costs.5 These are the psychosocial costs, such as decrements in quality of life or well-being, that are difficult to measure but nevertheless play an important role in understanding the burden on the patient. In the field of radiology, intangible psychosocial costs may consist of heightened anxiety and concern stemming from waiting for test results and from receiving positive test results.

Cost Item Measurement

For every cost item listed in the tables, there are actually two variables that are relevant: the utilization quantity and the price. It becomes clear that data collection for costing is a significant task. Cost item measurement consists of collecting the necessary data related to both utilization quantity and price for each item. Some cost items themselves may be comprised of multiple items. As seen in Table 9.2, the cost of a hospital admission has been broken down into its component items. Institutions that employ case costing management information systems may be able to supply a bundled hospital admission cost for the type of patients being studied. In these situations it is important to ascertain exactly what items are included in the bundle. Oftentimes missing items must be added in, such as inpatient physician consultation fees.

After identifying all the cost items that may be relevant to an economic evaluation, the next step is to identify the data sources that will supply information about utilization volume and price. Utilization volume for each item is usually determined at the patient level. Some prices may also be determined at the patient level, but typically prices for health care goods and services are fixed, especially in publicly financed health care systems. Fixed prices may vary over time or across jurisdictions.

It is not unusual to use multiple data sources for any given study. It is also often useful to consider multiple data sources for each of the variables related to volume and price—as long as the data are accurate, reliable, and are appropriate for the research question.

Cost item data for economic evaluations can be collected retrospectively, prospectively, or both. Researchers usually begin by determining what patient-level resource use data can be obtained from existing sources. Common sources for retrospective data collection of cost item utilization volume include patient charts and health plan administrative databases. The latter can serve as valuable sources for stable estimates of health resource use such as frequency of admissions, number of doctor visits, and number of medication prescriptions within a given study period. Other sources for retrospective data collection include patient registries and existing clinical study databases.

When resource use data are not readily available from existing sources, or if the data quality is low, a prospective study may be undertaken for some or all necessary cost items. An example of a cost item that is rarely documented and might need to be collected prospectively is time losses associated with patient care. Prospective data collection may include simple surveys or questionnaires. When the bulk of cost (and outcomes) data are to be collected prospectively for the experimental intervention and the standard care comparators, then epidemiologic principles of study design should be followed (Chapters 5 and 6). Ideal study designs include randomized controlled trials (RCTs). Nonexperimental observational study designs, such as cohort studies, are sometimes better suited, such as when the intervention is deployed at the institutional rather than the patient level. Such would be the case in an economic evaluation comparing, for example, the use of positron emission tomography (PET)-CT for diagnosis of Hodgkin lymphoma at one hospital or in one region compared to standard CT at another. It is also expedient to consider “piggy-backing” an economic evaluation onto a planned RCT. Adding a health resource use questionnaire onto a single or multicenter RCT that is already collecting outcomes data can be a highly efficient means of data collection. However, if the RCT has limited external validity because of patient inclusion and exclusion criteria, then the health economic data that are collected may not be meaningful for decision-makers who must consider cost-effectiveness for heterogeneous patient populations. In addition, a rigid RCT protocol may introduce protocol-driven health resource costs, or may interfere with the natural pattern of health resource use. In these cases estimates from RCTs may have to be disregarded, or adjusted if possible.

Some resource use variables, such as treatments for rare or latent adverse events (e.g., adult-onset leukemia caused by radiotherapy in childhood), may not be available from either short-term prospective or retrospective data sources. Health economic researchers must therefore sometimes rely on expert opinion.9 In these cases unbiased methods using structured questionnaires should be implemented. Soliciting information from multiple experts or using a formal Delphi panel⁹ also helps to mitigate bias, although the data cannot be assumed to be accurate if they are not empirical.

Typically price variables can be obtained from existing sources. Public health plan fee schedules can be used to cost physician services for the types of procedures and assessments relevant for the analysis. In Ontario, Canada, for example, fees can be easily obtained from the online Ontario Health Insurance (OHIP) schedule of Benefits and Fees.¹⁰ Laboratory procedure and technical fees may also be obtained from published fee schedules. Wage, tariff, and fee schedules are also often available for other professional groups such as chiropractors, complementary practitioners, and physical therapists. Institutions may provide salary information for nurses and imaging technicians. While institutions may agree to release these values for the analysis, compensation rules may require that the actual values be suppressed in descriptions of methods and technical reports. Many public drug plan formularies provide current prices for outpatient prescription medications. The prices of prescription drugs listed on the Ontario Drug Benefit Program formulary are readily available.¹¹ As described above, assigning a price to an inpatient stay is a complex task. The price varies as a function of case complexity, length of stay, and the types of cost items that are bundled into the case cost. High-quality case costs for the most responsible diagnosis associated with an inpatient admission are increasingly available as more and more institutions ascribe to management information systems that follow a uniform case costing protocol. An average bundled in-patient stay cost (i.e., case cost) based on an average length of stay can be retrieved for patients in Ontario for any given diagnosis who are admitted to one of the hospitals participating in the Ontario Case Costing Initiative.¹²

Similarly to resource use data, not all prices can be found retrospectively. Patients often pay out-of-pocket for drug dispensing, medical devices, home care, or other costs. If the items used can be identified, then it may be possible to consult wholesale price lists. Alternatively, one might include questions about out-of-pocket expenditures in prospective data collection instruments.

In collecting and assigning prices to cost items, one must be aware of the difference between prices and charges. The best value for prices is the one that represents the true value, or opportunity cost of the good or service, before any mark-ups or taxes are added. In a publicly financed, not-for-profit health care system, wage tariffs for physicians and regulated prices (before mark-ups) for medications represent reasonable approximations of the true costs. In the United States, hospital charges for specific types of admissions may be easily obtained, but are not an accurate representation of the monetary value of the admission, since they include a profit mark-up.

Omitting indirect or time costs from an economic evaluation can bias the result if these costs are significant, such as in the case of parents caring for ill children. Time losses can also be difficult to recall, particularly for many of the items listed in Table 9.4. It is preferable to collect these data prospectively from study participants, with repeated assessments as needed. When this is not possible, sometimes simplifying assumptions are applied. For example for children admitted to hospital, one may assume that a parent will incur one day of time loss for every day of inpatient care. Just as prices represent the value of goods and services, in the human capital approach, wages represent the value of time.¹³ Some studies include demographics surveys that ask respondents about their earned income, usually by asking them to select from a list of ranges. These data are sensitive, and respondents may not be willing to divulge this information. Unless the respondents are a balanced representation of the target population for the study, then using respondents’ reported salaries to value time losses will limit the generalizability of the findings. Many studies are subject to volunteer bias, with study respondents often representing a more educated and higher income bracket. Using their wages to value time losses may affect the validity of the findings if time loss contributes significantly to total cost. An alternative approach is to value the reported time losses by average, national statistical wages. These data, stratified by age, sex, and occupation, may be available from national census data sources.

Cost Valuation

Once all sources for prices and resource use volumes are identified, these are listed in a costing table. Table 9.5 is an example of a typical costing table used in economic evaluations.

When stochastic data have been collected for each study patient i, cost valuation begins at the level of the individual. The cost of each item is valued by multiplying price (p) by patient-level resource use, or quantity (q_i). For indirect costs, the unit price is represented by an hourly or daily wage and the quantity of resource use by the patient and caregiver’s hours or days lost. Once the cost of all items is valued, a vector of total patient costs can be determined as follows:

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related posts:

Obtaining Ethics Approval, Data Management, and Budget Formulation Steps for Using Research as a Success Tool in Academia Sample Size Estimation Descriptive Statistics Linear and Logistic Regression Measurements: Validity, Reliability, and Responsiveness

Stay updated, free articles. Join our Telegram channel

Join