The methods associated with measuring outcomes (the right-hand side of the equation) will be discussed in this section. ❹ As shown inTable 6–1, there are four ways to measure outcomes: CMA, CBA, CEA, and CUA. Each type of outcome measurement is associated with a different type of pharmacoeconomic analysis. The advantages and disadvantages of each type of analysis will be discussed in this section.
For a CMA, costs are measured in dollars, and outcomes are assumed to be equivalent. One example of a CMA is the measurement and comparison of costs for two therapeutically equivalent products, such as glipizide and glyburide.5 Another example is the measurement and comparison of using prostaglandin E2 on an inpatient versus an outpatient basis.6 In both cases, all the outcomes (e.g., efficacy, incidence of adverse drug interactions) are expected to be equal, but the costs are not. Some researchers contend that a CMA is not a true pharmacoeconomic study, because costs are measured, but outcomes are not. Others say that the strength of a CMA depends on the evidence that the outcomes are the same. This evidence can be based on previous studies, publications, FDA data, or expert opinion. The advantage of this type of study is that it is relatively simple compared to the other types of analyses because outcomes need not be measured. The disadvantage of this type of analysis is that it can only be used when outcomes are assumed to be identical.
A hospital needs to decide if it should add a new intravenous antibiotic to the formulary, which is therapeutically equivalent to the current antibiotic used in the institution and has the same adverse event profile. The advantage of the new antibiotic is that it only has to be administered once per day versus three times a day for the comparison antibiotic. Because the outcomes are expected to be nearly identical, and the objective is to assess the costs to the hospital (e.g., the hospital perspective), only direct medical costs need to be estimated and compared. The direct medical costs include the daily costs of each medication, the pharmacy personnel time used in the preparation of each dose, and the nursing personnel time used in the administration of each dose. Even if the cost of the new medication is a little higher than the cost of the current antibiotic, the lower cost of preparing and administering the new drug (once a day versus three times per day) may offset this difference. Direct nonmedical, indirect, and intangible costs are not expected to differ between these two alternatives and they need not be included if the perspective is that of the hospital, so these costs are not included in the comparison.
Mithani and Brown7 examined once-daily intravenous administration of an aminoglycoside versus the conventional every 8-hour administration (Table 6–2). The drug acquisition cost was in Canadian dollars ($Can) 43.70 for every 8 hours dosing, and $Can 55.39 for the single dose administration. Not including laboratory drug level measurements, the costs of the intravenous bag ($Can 29.32), preparation ($Can 13.81), and administration ($Can 67.63) were $Can 110.76 for the three-times daily administration versus $Can 42.23 (intravenous bag $Can 10.90, preparation $Can 6.20, and administration $Can 25.13) for the single daily dose. With essentially equivalent clinical outcomes, the once-daily administration of the aminoglycoside minimized hospital costs ($Can 97.62 versus $Can 154.46).
EXAMPLE OF COST MINIMIZATION
||Download (.pdf) TABLE 6–2.
EXAMPLE OF COST MINIMIZATION
|Type of Cost ||Every 8 Hours ||Once Daily |
|Drug acquisition cost ||$43.70 ||$55.39 |
|Minibag cost ||$29.32 ||$10.90 |
|Preparation cost ||$13.81 ||$6.20 |
|Administration costs ||$67.63 ||$25.13 |
|Total cost ||$154.46 ||$97.62 |
A CBA measures both inputs and outcomes in monetary terms. One advantage to using a CBA is that alternatives with different outcomes can be compared, because each outcome is converted to the same unit (dollars). For example, the costs (inputs) of providing a pharmacokinetic service versus a diabetes clinic can be compared with the cost savings (outcomes) associated with each service, even though different types of outcomes are expected for each alternative. Many CBAs are performed to determine how institutions can best spend their resources to produce monetary benefits. For example, a study conducted at Walter Reed Army Medical Center looked at costs and savings associated with the addition of a pharmacist to its medical teams.8 Discounting of both the costs of the treatment or services and the benefits or cost savings is needed if they extend for more than a year. Comparing costs and benefits (outcomes in monetary terms) is accomplished by using one of two methods. One method divides the estimated benefits by the estimated costs to produce a benefit-to-cost ratio. If this ratio is more than 1, the choice is cost beneficial. The other method is to subtract the costs from the benefits to produce a net benefit calculation. If this difference is positive, the choice is cost beneficial. The example at the end of this section will use both methods for illustrative purposes.
Another more complex use of CBA consists of measuring clinical outcomes (e.g., avoidance of death, reduction of blood pressure, and reduction of pain) and placing a dollar value on these clinical outcomes. This type of CBA is not often seen in the pharmacy literature, but will be discussed here briefly. This use of the method still offers the advantage that alternatives with different types of outcomes can be assessed, but a disadvantage is that it is difficult to put a monetary value on pain, suffering, and human life. ❺ There are two common methods that economists use to estimate a value for these health-related consequences: the human capital (HC) approach and the willingness-to-pay (WTP) approach. The HC approach assumes that the values of health benefits are equal to the economic productivity that they permit. The cost of disease is the cost of the lost productivity due to the disease. A person’s expected income before taxes and/or an inputted value for nonmarket activities (e.g., housework and child care) is used as an estimate of the value of any health benefits for that person. The HC approach was used when calculating the costs and benefits of administering a meningococcal vaccine to college students. The value of the future productivity of a college student was estimated at $1 million in this study.9 There are disadvantages to using this method. People’s earnings may not reflect their true value to society, and this method lacks a solid literature of research to back this notion. The WTP method estimates the value of health benefits by estimating how much people would pay to reduce their chance of an adverse health outcome. For example, if a group of people is willing to pay, on average, $100 to reduce their chance of dying from 1:1000 to 1:2000, theoretically a life would be worth $200,000 [$100/(0.001−0.0005)]. Problems with this method include the issue that what people say they are willing to pay may not correspond to what they actually would pay, and it is debatable if people can meaningfully answer questions about a 0.0005 reduction in outcomes.
An independent pharmacy owner is considering the provision of a new clinical pharmacy service. The objective of the analysis is to estimate the costs and monetary benefits of two possible services over the next 3 years (Table 6–3). Clinical Service A would cost $50,000 in start-up and operating costs during the first year, and $20,000 in years two and three. Clinical Service A would provide an added revenue of $40,000 each of the 3 years, Clinical Service B would cost $40,000 in start-up and operating costs the first year and $30,000 for years two and three. Clinical Service B would provide added revenue of $45,000 for each of the 3 years. Table 6–3 illustrates the comparison of both options using the perspective of the independent pharmacy with no discounting and when a discount rate of 5% is used. Although both services are estimated to be cost beneficial, Clinical Service B has both a higher benefit-to-cost ratio and a higher net benefit when compared to Clinical Service A.
CBA EXAMPLE CALCULATIONS
||Download (.pdf) TABLE 6–3.
CBA EXAMPLE CALCULATIONS
|Year 1 Dollars (No Discounting in Year 1) ||Year 2 Dollars (Discounted Dollars) ||Year 3 Dollars (Discounted Dollars) ||Total Dollars (Discounted Dollars) ||Benefit-to-Cost Ratio Dollars (Discounted Dollars) ||Net Benefit Dollars (Discounted Dollars) |
|Costs of A ||$50,000 ||$20,000 ||$20,000 ||$90,000 ||$120,000/$90,000 = 1.33:1 ||$120,000 − $90,000 = $30,000 |
| ||($50,000) ||($19,048) ||($18,140) ||($87,188) ||($114,376/87,188 = 1.31:1) ||($114,376 − 87,188 = 27,188) |
|Benefits of A ||$40,000 ||$40,000 ||$40,000 ||$120,000 || || |
| ||($40,000) ||($38,095) ||($36,281) ||($114,376) || || |
|Costs of B ||$40,000 ||$30,000 ||$30,000 ||$100,000 ||$135,000/100,000 = 1.35:1 ||$135,000 − 100,000 = 35,000 |
| ||($40,000) ||($28,571) ||($27,211) ||($95,782) ||($128,673/95,782 = 1.34:1) ||($128,673 − 95,782 = 32,891) |
|Benefits of B ||$45,000 ||$45,000 ||$45,000 ||$135,000 || || |
| ||($45,000) ||($42,857) ||($40,816) ||($128,673) || || |
This is the most common type of pharmacoeconomic analysis found in the pharmacy literature. A CEA measures costs in dollars and outcomes in natural health units such as cures, lives saved, or blood pressure. An advantage of using a CEA is that health units are common outcomes practitioners can readily understand and these outcomes do not need to be converted to monetary values. On the other hand, the alternatives used in the comparison must have outcomes that are measured in the same units, such as lives saved with each of two treatments. If more than one natural unit outcome is important when conducting the comparison, a cost-effectiveness ratio should be calculated for each type of outcome. Outcomes cannot be collapsed into one unit measure in CEAs as they can with CBAs (outcome = dollars) or CUAs (outcome = quality-adjusted life years [QALYs]). Because CEA is the most common type of pharmacoeconomic study in the pharmacy literature, many examples are available. Bloom and others10 compared two medical treatments for gastroesophageal reflux disease (GERD), using both healed ulcers confirmed by endoscopy and symptom-free days as the outcomes measured. Law and others11 assessed two antidiabetic medications by comparing the percentage of patients who achieved good glycemic control as the outcome measure.
A cost-effectiveness grid (Table 6–4) can be used to illustrate the definition of cost-effectiveness. In order to determine if a therapy or service is cost-effective, both the costs and effectiveness must be considered. Think of comparing a new drug with the current standard treatment. If the new treatment is (1) both more effective and less costly (cell G), (2) more effective at the same price (cell H), or (3) has the same effectiveness at a lower price (cell D), the new therapy is considered cost-effective. On the other hand, if the new drug is (1) less effective and more costly (cell C), (2) has the same effectiveness but costs more (cell F), or (3) has lower effectiveness for the same costs (cell B), then the new product is not cost-effective. There are three other possibilities: the new drug is (1) more expensive and more effective (cell I)—a very common finding, (2) less expensive but less effective (cell A), or (3) has the same price and the same effectiveness as the standard product (cell E). For the middle cell E, other factors may be considered to determine which medication might be best. For the other two cells, an incremental cost-effectiveness ratio (ICER) is calculated to determine the extra cost for each extra unit of outcome. It is left up to the readers to determine if they think the new product is cost-effective, based on a value judgment. The underlying subjectivity as to whether the added benefit is worth the added cost is a disadvantage of CEA.
||Download (.pdf) TABLE 6–4.
|Cost-Effectiveness ||Lower Cost ||Same Cost ||Higher Cost |
|Lower effectiveness ||A ||B ||C |
|Same effectiveness ||D ||E ||F |
|Higher effectiveness ||G ||H ||I |
An MCO is trying to decide whether to add a new cholesterol-lowering agent to its preferred formulary. The new product has a greater effect on lowering cholesterol than the current preferred agent, but a daily dose of the new medication is also more expensive. Using the perspective of the MCO (e.g., direct medical costs of the product to the MCO), the results will be presented in three ways in Tables 6–5 to 6–7 to illustrate the various ways that costs and effectiveness are presented in the literature. Table 6–5 presents the simple listing of the costs and benefits of the two alternatives. Sometimes for each alternative, the costs and various outcomes are listed but no ratios are conducted—this is termed a cost-consequence analysis (CCA).
LISTING OF COSTS AND OUTCOMES
||Download (.pdf) TABLE 6–5.
LISTING OF COSTS AND OUTCOMES
|Alternative ||Costs for 12 Months of Medication ||Lowering of LDL in 12 Months (mg/dL) |
|Current preferred medication ||$1000 ||25 |
|New medication ||$1500 ||30 |
The second method of presenting results includes calculating the average cost- effectiveness ratio (CER) for each alternative. Table 6–6 shows the cost-effectiveness ratio for the two alternatives. The CER is the ratio of resources used per unit of clinical benefit and implies that this calculation has been made in relation to doing nothing or no treatment. In this case, the current medication costs $40 for every 1 mg/dL decrease in LDL while the new medication under consideration costs $50 for the same decrease. In clinical practice, the question is infrequently: “Should we treat the patient or not?” or “What are the costs and outcomes of this intervention versus no intervention?” More often the question is: “How does one treatment compare with another treatment in costs and outcomes?” To answer this more common question, an incremental cost-effectiveness ratio (ICER) is calculated. The ICER is the ratio of the difference in costs divided by the difference in outcomes. Most economists agree that an ICER (the extra cost for each added unit of benefit) is the more appropriate way to present CEA results. Table 6–7 shows the incremental cost-effectiveness (the extra cost of producing one extra unit) of the new medication compared to the current medication. For the new medication, it costs an additional $100 for every additional decrease in LDL of 1 mg/dL. The formulary committee would need to decide if this increase in cost is worth the increase in benefit (improved clinical outcome). In this example, the costs and benefits of the medications are estimated for only 1 year; discounting is not needed. If incremental calculations produce negative numbers, this indicates that one treatment is both more effective and less expensive, or dominant, compared to the other option. The magnitude of the negative ratio is difficult to interpret, so it is suggested that authors instead indicate which treatment is the dominant one. As mentioned earlier, when one of the alternatives is both more expensive and more effective than another, the ICER is used to determine the magnitude of added cost for each unit in health improvement (see CEA grid, cell I, Table 6–4).
||Download (.pdf) TABLE 6–6.
|Alternative ||Costs for 12 Months of Medication ||Lowering of LDL in 12 Months ||Average Cost per Reduction in LDL |
|Current preferred medication ||$1000 ||25 mg/dL ||$40 per mg/dL |
|New medic ation ||$1500 ||30 mg/dL ||$50 per mg/dL |
INCREMENTAL COST-EFFECTIVENESS RATIO
||Download (.pdf) TABLE 6–7.
INCREMENTAL COST-EFFECTIVENESS RATIO
|Alternative ||Costs for 12 Months of Medication ||Lowering of LDL in 12 Months ||Incremental Cost per Marginal Reduction in LDL |
|Current preferred medication ||$1000 ||25 mg/dL ||($1500 − $1000)/(30 mg/dL − 25 mg/dL) = $100 per mg/dL |
|New medication ||$1500 ||30 mg/dL || |
Clinicians must then wrestle with this type of information—it becomes a clinical call. Many economists will argue that this uncertainty is why cost-effectiveness may not be the preferred method of pharmacoeconomic analysis.
❻ A CUA takes patient preferences, also referred to as utilities, into account when measuring health consequences.12 The most common unit used in conducting CUAs is QALYs. A QALY is a health-utility measure combining quality and quantity of life, as determined by some valuations process. The advantage of using this method is that different types of health outcomes can be compared using one common unit (QALYs) without placing a monetary value on these health outcomes (like CBA). The disadvantage of this method is that it is difficult to determine an accurate QALY value. This is an outcome measure that is not well understood or embraced by many providers and decision makers. Therefore, this method is not commonly seen in the pharmacy literature. One reason researchers are working to establish methods for measuring QALYs is the belief that 1 year of life (a natural unit outcome that can be used in CEAs) in one health state should not be given the same weight as 1 year of life in another health state. For example, if two treatments both add 10 years of life, but one provides an added 10 years of being in a healthy state and the other adds 10 years of being in a disabled health state, the outcomes of the two treatments should not be considered equal. Adjusting for the quality of those extra years is warranted. When calculating QALYs, 1 year of life in perfect health has a score of 1 QALY. If health-related quality of life (HR-QOL) is diminished by disease or treatment, 1 year of life in this state is less than 1 QALY. This unit allows comparisons of morbidity and mortality. By convention, perfect health is assigned 1 per year and death is assigned 0 per year, but how are scores between these two determined? Different techniques for determining scales of measurement for QALY are discussed below.
There are three common methods for determining QALY scores: rating scales (RS), standard gamble (SG), and time trade-off (TTO). A rating scale consists of a line on a page, somewhat like a thermometer, with perfect health at the top (100) and death at bottom (0). Different disease states are described to subjects, and they are asked to place the different disease states somewhere on the scale indicating preferences relative to all diseases described. As an example, if they place a disease state at 70 on the scale, the disease state is given a score of 0.7 QALYs.
The second method for determining patient preference (or utility) scores is the standard gamble method. For this method, each subject is offered two alternatives. Alternative one is treatment with two possible outcomes: either the return to normal health or immediate death. Alternative two is the certain outcome of a chronic disease state for life. The probability (p) of dying is varied until the subject is indifferent between alternative one and alternative two. As an example, a person considers two options: a kidney transplant with a 20% probability of dying during the operation (alternative one) or dialysis for the rest of his life (alternative two). If this percent is his or her point of indifference (he or she would not have the operation if the chances of dying during the operation were any higher than 20%), the QALY is calculated as 1 − p or 0.8 QALY.
The third technique for measuring health preferences is the TTO method. Again, the subject is offered two alternatives. Alternative one is a certain disease state for a specific length of time t, the life expectancy for a person with the disease, then death. Alternative two is being healthy for time x, which is less than t. Time x is varied until the respondent is indifferent between the two alternatives. The proportion of the number of years of life a person is willing to give up (t − x) to have his or her remaining years (x) of life in a healthy state is used to assess his or her QALY estimate. For example, a person with a life expectancy of 50 years is given two options: being blind for 50 years or being completely healthy (including being able to see) for 25 years. If the person is indifferent between these two options (he or she would rather be blind than give up any more years of life), the QALY for this disease state (blindness) would be 0.5. Table 6–8 contains examples of disease states and QALY estimates for each disease state listed.
SELECTED QALY ESTIMATES
||Download (.pdf) TABLE 6–8.
SELECTED QALY ESTIMATES
|Disease State ||QALY Estimate |
|Complete health ||1.00 |
|Moderate angina ||0.83 |
|Breast cancer: removed breast, unconcerned ||0.80 |
|Severe angina ||0.53 |
|Cancer spread, constant pain, tired, not expected to live long ||0.16 |
|Death ||0.00 |
As one might surmise, QALY measurement is not regarded as being as precise or scientific as natural health unit measurements (such as blood pressure and cholesterol levels) used in CEAs. Some issues in the measurement of QALYs are debated in the literature. One issue concerns whose viewpoint is the most valid. An advantage of having patients with the disease of interest determine health state scores is that these patients may understand the effects of the disease better than the general population, whereas, some believe these patients would provide a biased view of their disease compared with other diseases they have not experienced. Some contend that health care professionals could provide good estimates because they understand various diseases and others argue that these professionals may not rate discomfort and disability as seriously as patients or the general population.
Another issue that has been addressed regarding patient preference or utility-score measures is the debate over which is the best measure. Utility scores calculated using one method might differ from those using another. Finally, utility measures have been criticized for not being sensitive to small, but clinically meaningful, changes in health status.
An article by Kennedy and associates13 assessed the costs and utilities associated with two common chemotherapy regimens (vindesine and cisplatin [VP], and cyclophosphamide, doxorubicin, and cisplatin [CAP]) and compared the results with the costs and utilities of using best supportive care (BSC) in patients with nonsmall cell lung cancer. The perspective was that of the health care system or the payer. Using the TTO method, treatment utility scores were estimated by personnel of the oncology ward. Although the chemotherapy regimens provide a longer survival (VP = 214 days, CAP = 165 days) than BSC (112 days), the quality-of-life TTO score was higher for BSC (0.61) compared with the chemotherapy regimens (0.34). When survival time is multiplied by the TTO scores, the use of BSC results in an estimated 0.19 QALYs, which is similar to VP (0.19 QALYs), but higher than CAP (0.15 QALY). The costs to the health care system for the three options are about $5000 for BSC, $10,000 for VF, and $7000 for CAP (the authors reported median costs instead of average costs due to the abnormality of the cost data). Cost-utility ratios are calculated similarly to cost-effectiveness ratios, except that the outcome unit is QALYs. Therefore, the cost-utility ratio is about $26,000/QALY for BSC and about $44,000 to $52,000/QALY for the chemotherapy regimens. Because BSC is at least as effective, as measured by QALYs, and is less expensive than the other two options, a marginal (or incremental) cost-utility ratio does not need to be calculated. Marginal cost-utility ratios only need to be calculated to estimate the added cost for an added benefit, not when the added benefit comes at a lower cost.