❃ Introduction Cost control is a major element of the job for all nurse managers and executives. Basic knowledge of costs and their behavior is central to an understanding of the expenses incurred in the responsibility centers under a manager’s control. The skillful management of costs is an essential element of the financial success of any health care organization (HCO). However, cost management is complicated. Many cost terms commonly used are not well understood. Direct costs, indirect costs, average costs, fixed costs, variable costs, and marginal costs are concepts that managers use and must understand clearly. This chapter provides definitions of these and other key terms to aid nurse managers and executives in performing their jobs. The relationship between fixed and variable costs is the most fundamental cost concept. The total costs incurred by HCOs depend directly on the interplay between fixed and variable costs. These concepts depend on the volume, or number, of services provided. Volume ultimately becomes a critical aspect of almost every successful HCO and of every failing one. HCOs are not static; rather, they are constantly in a state of flux and change. Services are added; other services are deleted. Some areas of the organization expand; others contract. To a great extent, these changes are dictated by clinical factors, technologies, and the introduction of evidence-based practices. However, every change has an impact on organizational cost. In each case, managers should determine whether the financial impact of a proposed change will be favorable or unfavorable for the organization. Although an unfavorable financial impact may not mean that a service must be deleted, a good manager makes decisions based on as much information as possible. One approach, called marginal cost analysis, is critical to generating information for such decisions. Another important area of cost management is cost estimation. What will costs likely be in the coming year? That question requires careful management attention. This chapter provides several techniques to help in exploring that topic. The last major topic covered in this chapter is break-even analysis (BEA). BEA addresses the issue of the volume of patients required for a program or service to become financially self-sufficient. ❃ Basic cost concepts The most critical of cost concepts is referred to as cost behavior. Cost behavior is the way that costs change in reaction to events within the organization. If patient volume rises by 5%, what happens to costs? What if patient volume falls by 5%? How can one predict whether total costs will exceed revenues or remain less than revenues? Which factors are related to stable costs and which to rising costs? Cost behavior depends on the specific elements of cost in any responsibility center or organization. A responsibility center is any unit or department in an organization that incurs costs and for which a manager is assigned responsibility. Some types of costs are stable, changing little, if at all, even in response to significant changes in patient workload. Other costs are highly changeable, reacting directly to other changes in the organization. This section of the chapter lays out a framework on which to develop an understanding of how costs behave in HCOs. Definitions Cost measurement is more complex than one might expect. When someone asks what something costs, accountants have trouble responding with a direct answer. The reason is that the appropriate measure of cost depends substantially on the intended use for the cost information. Finding out what it cost to treat each patient last year is very different from calculating what it might cost to treat one more patient next year. The cost per patient when 100 patients are treated may be very different from the cost per patient when 500 patients are treated. To make sense in this complicated area, all managers (nurse managers, executives, and accountants) must rely heavily on a consistent set of definitions. The definitions provided here form the basis of a common language so that when a cost per patient is cited, all managers can interpret that information in the same way and communicate effectively with each other. Service unit A basic measure of the product or service being produced by the organization, such as discharged patients, patient days, home care visits, emergency department treatments, or hours of surgery. Cost information is often collected on a service unit basis. This is an important concept for managers because they need to know the cost per service unit as it pertains to their area of responsibility and to other areas of the organization. Within one HCO, various types of service units may exist. For example, the operating room (OR) may use hours of surgery, but a medical-surgical unit uses patient days. Direct costs can be defined in two ways: a. Costs that are incurred within the organizational unit for which the manager has responsibility are referred to as direct costs of the unit. b. Costs of resources used in the direct care of patients are referred to as the direct costs of patient care. Indirect costs can also be defined in two ways: c. Costs assigned to an organizational unit from elsewhere in the organization are indirect costs for the unit. d. Costs within a unit that are incurred for reasons other than direct patient care are indirect costs of patient care. Direct costs are sometimes difficult to understand because their definition depends on the object of the analysis. If one were interested in the direct cost of patient care (definition “b”) for a specific patient care unit, that cost would not include the cost of the unit manager’s time spent on administrative duties or the cost of clerical personnel in the unit. It likewise would not include the cost of a nurse’s time to verify that essential items are on the emergency cart. It would, however, include the cost of the actual clinical supplies used, as well as the cost of nursing time spent with a patient. From a broader systems perspective, however, all of these costs are direct costs. This perspective applies when one is contrasting all of the costs assigned to a unit or department. For the unit as a whole (definition “a” above), all nursing salaries and clerical salaries incurred within the unit are direct costs. Costs assigned to the unit from outside, such as marketing, laundry, or housekeeping, are indirect costs to the unit. Similarly, nursing professional development and education, as well as a portion of the salary of the chief nurse executive (CNE), are indirect costs to a specific unit. However, from the perspective of the nursing department as a whole, rather than that of one unit, the salary of the CNE is a direct cost. Laundry costs are considered indirect by both the manager for the unit and the CNE for the nursing department. The labor and other costs of the laundry department are clearly incurred neither directly within individual nursing units nor within the nursing department as a whole, even though nursing may control the amount of laundry it uses (definition “c”). However, that does not mean that laundry costs are considered indirect costs by everyone. The manager of the laundry department would consider all costs incurred in that department as direct costs. Perspective is therefore critical in the classification of a cost as direct or indirect. From a focus on patient care, indirect costs would include all costs that are not direct patient care costs (definition “d”). Full cost Total of all costs associated with or in an organizational unit or activity. This includes direct and indirect costs. For example, all of the costs of a hospital would represent its full or total costs. At the department or unit level, it is important to include not only direct costs but also all indirect costs. In this case full costs include an appropriate share of laundry, administration, billing, engineering, medical records, housekeeping, marketing, and so forth. The question of what is an appropriate share of costs to assign to each responsibility center is difficult and is addressed in Chapter 9. Average cost Full cost divided by the volume of service units. Many questions faced by managers require information on the cost per patient, the cost per treatment, or the cost per patient day. These all represent average costs. Each responsibility center may use its own service units and calculate its own average cost. Fixed costs Costs that do not change in total as the volume of service units changes. Variable costs Costs that vary directly with changes in the volume of service units. The salaries of the CNE and unit nurse managers do not change day by day as the census changes. Other staff salaries also do not change on a daily basis, such as those of the unit clerks, assistant nurse managers, schedulers, or staff educators. Therefore these costs are fixed costs. Supplies that are used for every patient are variable. The more patients, the more supplies used. The definition of the service unit measure is crucial in defining fixed and variable costs. For example, although most clinical supplies used in a hospital vary with the number of patient days, surgical supplies are more likely to vary with the number of surgical procedures, and clinic supplies will likely vary with the number of clinic visits. Relevant range Normal range of expected activity for the responsibility center or organization. Fixed costs are fixed over the relevant range. Suppose that a 30-bed nursing unit anticipated an average occupancy rate of 75%. It might be reasonable to expect the unit to have an 80% or even 85% occupancy rate. However, a rate of 160% would clearly be beyond the reasonable anticipated rate for that unit. To accommodate such a large number of patients, the hospital would have to add another nursing unit. Within the normal relevant range, the costs for the salary of the nurse manager would be fixed. However, if occupancy reached 160%, there might well be expansion to two units, with a manager for each. The fixed costs would rise because costs are fixed only within the relevant range. Marginal costs Extra costs incurred as a result of providing one more service unit, such as one extra patient day or one more type of service. Marginal cost information is critical in the decision-making process. Managers are often interested in how costs change when the number of patients cared for changes. If we treat one more patient, how much more money will we have to spend? This differs to some extent from variable costs. What if a piece of equipment is at its capacity? To treat one more patient, we will incur not only variable costs such as supplies but also the fixed costs from purchasing another piece of equipment. Marginal cost would include both variable and fixed costs that change from adding an extra patient or extra group of patients. Marginal costs are often used to analyze major changes, such as the addition of a whole new service. Suppose that a hospital does not currently provide liver transplantation, but hospital leaders are exploring what it might cost to add this program. In this case an evaluation of marginal costs would focus on what it would take to add the new liver transplantation program to the hospital’s range of services. Because no transplants are done currently, it is probable that the hospital would need equipment, supplies, and personnel to provide that service. The supplies represent a variable cost. For each transplant done, the hospital will need additional clinical supplies. However, the equipment represents a fixed cost. After the equipment is purchased, it can be used for many transplants. Both the supplies and the equipment are needed for the first transplants, which means that both the supplies and equipment costs are marginal costs of adding that service. If one considers the total costs incurred by an organization before it makes a change and the total costs after it makes the change, the difference in costs represents the marginal costs of that change. Mixed costs Costs that contain both fixed and variable cost elements. An example of a mixed cost is electricity. The OR department uses some electricity every day for lighting hallways and for other purposes not related to the number of patients. That portion of electric usage is a fixed cost. Much of the electricity used by the department is used in the operating suites. The more surgeries, the more electricity used. Therefore it has a variable component as well. Similarly, a home health agency would have mixed costs for transportation if it used agency-owned cars. Some costs would be fixed, such as annual maintenance, and some costs, such as gas, would vary with the number of home visits. Mixed costs create some problems for managers. If the number of service units for a responsibility center rose by 10%, one would expect fixed costs to stay unchanged. One would expect variable costs to rise by 10%. Mixed costs would be expected to rise but by some amount greater than 0% and less than 10%. Methods are provided later in this chapter for estimating the change in mixed costs resulting from a change in patient volume. Step-fixed or step-variable cost Costs that are fixed over small ranges of activity that are less than the relevant range. Nursing units often require a fixed number of nurses on duty for a range of patients. Within that range of patients, the nursing personnel cost remains fixed. However, if the number of patients increases by a large enough number, additional personnel will be needed. That personnel level is then fixed over a new, higher range of activity. The key to step costs is that they are fixed over volume intervals but vary within the relevant range. These concepts are discussed in greater detail throughout this chapter. We will use a variety of examples and graphs to explain cost concepts. Fixed versus variable costs The total costs of running a department are generally divided into costs that are fixed and costs that are variable. The concepts of fixed and variable costs are often conceptualized with the use of graphs. In the following example, the service unit measure is assumed to be patient days. Fig. 8.1 provides an example of fixed costs. Specifically, the graph shows the annual salary for a unit nurse manager for the coming year. The salary is $100,000.1 That salary is a fixed cost for the organization. The salary paid to a nurse manager is not dependent on any patient-volume statistic. In Fig. 8.1, the vertical axis shows the cost to the institution. As one moves up this axis, costs increase. The horizontal axis shows the number of patient days. The farther to the right one moves, the more patient days the institution has. FIG. 8.1Fixed costs: cost for a nurse manager. Note that the fixed costs appear as a horizontal line. This is because regardless of the volume, the salary for the nurse manager will remain the same. Thus the cost for 8000, 10,000, or 12,000 patient days is the same. Variable costs fluctuate with the volume of service units. Most variable costs are supply costs (e.g., general patient care supplies, linens, dietary stock, and pharmacy supplies). To illustrate how variable costs are determined, we can suppose that each patient on a 32-bed general medicine unit has his or her temperature taken twice a day with either an oral or a tympanic thermometer. However, the unit’s practice council recommends that the unit switch from taking patients’ temperatures with a tympanic device to taking them with an oral device with a disposable cover. Based on this policy change, the nurse manager expects that each patient on the unit would use a minimum of two oral thermometer covers per day, but use of disposable covers will vary directly with patient volume. The nurse manager determines that, when purchased from an approved hospital vendor, each thermometer cover costs about $0.05; if two a day are used for each patient, the cost of those items is $0.10 for each patient each day. The more patient days, the more the cost for that disposable item in the total nursing unit budget. Consider Fig. 8.2. This graph plots the cost for disposable thermometer covers as their number varies with patient volume. As in Fig. 8.1, the vertical axis represents cost, and the horizontal axis represents patient volume. Unlike Fig. 8.1, which shows some amount of cost even at a volume of zero, Fig. 8.2 shows zero cost at a volume of zero because with zero patient days, none of this particular supply is used. The total variable cost increases by $0.10 for each extra patient day. FIG. 8.2Variable costs: costs of a disposable supply item. Rather than assume that the unit is filled to capacity every day, the nurse manager could use the average number of patient days to calculate the number of oral thermometer covers needed and consequent costs. The general principle is that the greater the number of patient days, the greater the cost for a disposable or any variable cost item in the total nursing unit budget. For instance, in Fig. 8.3, dashed lines have been inserted to show the cost when there are 10,000 patient days and the cost when there are 50,000 patient days. As you can see, $1000 is spent on the disposable item if there are 10,000 patient days, and $5000 is spent if there are 50,000 patient days. The total of the fixed and variable costs is shown in Fig. 8.4. This figure combines the fixed costs from Fig. 8.1 with the variable costs from Fig. 8.3. Note that in the graph, the total costs start at $100,000, even if volume is zero, because the nurse manager’s salary is a fixed cost. Although this example is used to illustrate a point, the same principle applies for small equipment that is “charged” per patient use or for any other purchase that must be determined based on patient volume. Another important issue is that some variable supply costs, such as linens and dietary, are purchased as a contracted service (e.g., the organization “contracts” with a linen service to provide all of the linens used), whereas others are obtained through negotiated rates with vendors (e.g., the organization negotiates with a vendor to provide certain supplies at an agreed-upon rate). On some contracts, the costs of supplies per unit may actually decrease as volume increases. FIG. 8.3Variable costs: costs of a disposable supply item at volumes of 10,000 and 50,000 patient days. FIG. 8.4Total costs for a unit. Cost graphs and the relevant range One potential problem exists with this type of graphic analysis of fixed and variable costs. That concerns the relevant range. The relevant range represents the likely range of activity covered by a budget. Variable costs increase proportionately over the relevant range. However, it is unlikely that the hospital will pay $0.05 for each disposable thermometer cover at any given volume level. If purchases increase substantially, the hospital will probably get a price reduction per unit. On the other hand, if purchases decrease substantially, the hospital would possibly have to pay more per unit. However, the variable costs may reasonably be considered to increase proportionately over the relevant range. As noted earlier, fixed costs are not fixed over any range of activity. If a nursing unit has zero patient days, the hospital will close the unit and not have any fixed cost for a nurse manager. If patient volume rises substantially and exceeds the capacity of the unit, the hospital might need to open a second unit and incur the additional cost of a second nurse manager. The costs, however, are fixed over the relevant range. Essentially, variable costs do not increase by exactly the same amount per unit over any range of volume, and fixed costs do not remain fixed over any range of volume. However, for most budgets, volume expectations for the coming year do not assume drastic changes. When fixed and variable costs are graphed, the relevant range issue is often ignored; costs appear fixed over all ranges of activity in the graph. However, the user of the graph should bear in mind that the graph’s information is accurate only within the relevant range. Many costs are step-fixed and vary within the relevant range, but not smoothly. They are fixed over intervals shorter than the relevant range. See Fig. 8.5 for a graphic representation of step-fixed costs. FIG. 8.5Example of step-fixed costs. For example, staffing patterns may be such that a nursing unit will use several nurses to cover a range of workload (or demand). If that range is exceeded, the unit will have to hire new permanent nursing staff, use or hire additional per diem staff, or contract with an outside agency for additional nurses on a short-term basis to staff the unit. Clearly, more patient days or greater average acuity requires more nursing care hours. However, if the staffing pattern is based on the delivery of about 10.6 hours of nursing time per patient day, the unit would not expect to hire a nurse for an additional 10.6 hours every time the patient-day census increased by one. If the hours were consistently at this level, the nurse manager could use this information to help justify an additional permanent position. It should be noted that the hours per patient day also vary by type of unit. The nursing care hours per patient day on a special care unit are often higher than on a general medical unit. As long as a staffing plan exists indicating the number of nurses needed for any volume of patient days, step costs should not present a major budgeting problem.2 Generally, because of the use of overtime and agency nurses, a step-fixed pattern of cost is estimated by treating the staffing costs as if they were variable. Although this will not give a precise result, it is usually a reasonable approximation. The impact of volume on cost per patient If a nurse manager were to ask what it costs to treat patients on the unit, accountants would probably answer, “It depends.” Costs are not unique numbers that are always the same. The cost to treat a patient depends on several critical factors. One of these is the volume of patients for whom care is being provided. Although other factors may explain the costs to treat patients on a particular unit—namely, acuity, type of patients being cared for on the unit, and the experience or expertise level of staff—our focus here is primarily on volume. Suppose that a unit has fixed costs of $400,000 and variable costs per patient day of $400. With these hypothetical data, what is the average cost per patient day? If there are 3000 patient days for the year, the total costs are the fixed cost of $400,000 plus $400 per patient day for each of the 3000 patient days. The variable costs are $1,200,000 (i.e., $400 per patient day × 3000 patient days). The total cost is $1,600,000 (i.e., $400,000 fixed cost + $1,200,000 variable cost). The cost per patient is $533 (i.e., $1,600,000 total cost ÷ 3000 patient days) per patient day. However, what if there are only 2500 patient days? Then the variable costs, at $400 per patient day, are $1,000,000, and the total cost is $1,400,000. In this case the cost per patient day is $560. The cost is higher because there are fewer patients sharing the fixed costs. Each patient day causes the hospital to spend another $400 of variable costs. The $400,000 fixed cost remains the same regardless of the number of patient days. If there are more patient days, each one shares less of the $400,000 fixed cost. If there are fewer patient days, the fixed cost assigned to each rises. Table 8.1 calculates the fixed, variable, total, and average costs per patient at a variety of patient volumes for this scenario. ❃ TABLE 8.1 Fixed, Variable, Total, and Average Costs at Various Patient Volumes Volume (A) Fixed Cost (B) Variable Cost (C = $400 × A) Total Cost (D = B + C) Average Cost (E = D ÷ A) 1 $400,000 $ 400 $ 400,400 $400,400 50 400,000 20,000 420,000 8,400 100 400,000 40,000 440,000 4,400 500 400,000 200,000 600,000 1,200 1,500 400,000 600,000 1,000,000 667 2,500 400,000 1,000,000 1,400,000 560 3,000 400,000 1,200,000 1,600,000 533 Fig. 8.6 shows the average cost at different patient volumes. The cost declines as the volume of patient days increases because more patients are sharing the fixed costs. In Chapter 4 this result was referred to as economies of scale. FIG. 8.6Average cost per patient day. Suppose there are only 500 patient days. The total variable costs of $400 per patient day are $200,000, and the total cost is $600,000. The cost per patient day is $1200, more than double the previous cost per patient results at 2500 and 3000 patient days. In trying to understand costs, it is critical to grasp the concept that because fixed costs do not change in total, the cost per patient or per patient day does change as volume changes. The greater the volume, the greater the number of patients available to share the fixed costs. Because costs depend on volume, there is no unique answer to the question “What is the cost per patient day?” That question can only be answered by giving the cost per patient day assuming a specific volume of patients. The volume of patients is critical. One implication of this result is that HCOs almost always find higher volume preferable to lower volume. As volume increases, the average cost per patient declines. If prices can be maintained at the original level, the declining cost will result in lower losses or higher profits. Marginal cost analysis Another costing concern is the issue accountants and economists refer to as marginal cost analysis. Decisions to change the volume of a service or to change the specific types of services offered should be based on marginal rather than average costs. If someone were to ask the nurse manager the cost of treating a particular type of patient, the answer should be, “It depends.” The previous section pointed out that cost depends on the number of patients. It also depends on why we want to know. If the question is just one of historical curiosity, the average cost is an adequate response. However, if the information will be used for decision making, that response may well be incorrect. Suppose the hospital was trying to decide whether to negotiate with an insurer to accept additional patients of the same average acuity and mix as the 2500 patient days the hospital currently has. The insurer has offered $500 per patient day for 500 patient days. From the earlier calculations, the cost per patient for 500 patient days is $1200! However, the hospital would not be providing only 500 patient days of care. It already has 2500 patient days. From the earlier chart, the average cost is $560 per patient day at 2500 patient days. At 3000 patient days, the average cost would be $533. Given that information, would it pay to accept the additional patients at a price of only $500? It definitely would. Why should the hospital accept $500 if the additional patient days will cost at least $533? Actually the additional patient days will not cost at least $533. All of the patients, on average, will cost that amount. The $533 includes a share of both fixed and variable costs. If the unit is going to have at least 2500 patient days regardless of the insurer negotiation outcome, the fixed costs of $400,000 will be incurred no matter what. The fixed costs will not change if the hospital has the extra 500 patient days. Decisions such as this one require marginal analysis. The “margin” refers to a change from current conditions. “A patient on the margin” refers to adding one more patient or reducing volume by one patient. Marginal costs are the costs for treating one more patient. On the margin in this case, if the hospital were to take the additional patients, it would have more variable costs but would not have any additional fixed costs (assuming that 3000 patient days is within the relevant range). Each extra patient causes the hospital to spend only the additional variable costs of $400 per patient day. That is less than the $500 the insurer has offered to pay. The hospital will be better off by $100 for each additional patient day. The additional costs incurred for additional patients are often referred to as the marginal, out-of-pocket, or incremental costs. If fixed costs were to rise because the relevant range was exceeded, those costs would be included appropriately in the incremental costs along with the variable costs. The key element in marginal costing is that the only costs relevant to a decision are those that change as a result of the decision. The decision may be to add a new service or to close down an existing one. It may be to expand volume (as in the previous example) or to contract volume. In any case, when a decision is being made that involves changing the patient volume, the revenues and changing costs are the essential information. Effective managerial decisions require that the manager know the amount by which total costs will increase and the amount by which the total revenues will increase or, alternatively, the amount by which both will decrease. Costs that do not change in total for the organization are not relevant to the decision. Fixed costs generally do not increase when additional patients are added (within the relevant range), and therefore they do not affect marginal costs. Marginal cost analysis is sometimes referred to as relevant costing. The concept of so-called relevant costing is that all decisions should be based only on costs that are relevant to the decision. The simplest way to think about this concept is to consider costs before and after a change. The only relevant costs are those that change as a result of the decision. The approach applies equally to revenues. In the hospital-insurer example, suppose that before the insurer negotiation, the hospital was receiving $550 for each of its 2500 patient days. Total revenue ($550 × 2500) was $1,375,000. Total costs were $1,400,000 (calculated earlier). The hospital was losing $25,000, the amount that the $1,400,000 total costs exceeded the $1,375,000 of revenues. If the insurer business were accepted, the additional revenue would be $500 times 500 patient days, or $250,000. The total costs for 3000 patient days (calculated earlier) are $1,600,000. The cost increase of going from 2500 patient days to 3000 patient days is only $200,000 (i.e., $1,600,000 total cost for 3000 patient days versus $1,400,000 total cost for 2500 patient days). The total costs with the insurer’s patients are $1,600,000, and the total revenues are $1,625,000 (i.e., the original $1,375,000 + $250,000 revenue from the insurer). The unit has gone from a loss of $25,000 to a profit of $25,000. The costs have increased by $200,000, and the revenues have increased by $250,000. The amount by which the extra revenues exceed the extra costs for the 500 insurer patients accounts for the turnaround from a loss to a profit. This should not be surprising. The extra revenue per patient day is $500. The cost per additional patient day is $400. The difference between incremental revenue of $500 per patient day and incremental cost of $400 per patient day is a profit of $100 per patient day. This extra profit of $100 for each of the 500 insured patient days accounts for exactly the $50,000 profit from the insurer patients, which turned a $25,000 loss into a $25,000 gain. If the hospital had used average cost information for its decision, it would have turned away the extra business and lost the chance to turn a $25,000 loss into a $25,000 profit. The $500 per patient day revenue is less than the $533 average per patient day cost. However, the average cost is not relevant for such decisions because it incorrectly assumes that each extra patient day will cause the hospital to have additional variable and fixed costs. The incremental cost is relevant because it considers only the additional revenues and additional costs that the hospital will have as a result of the proposed change. Relevant cost case study Assume that a hospital is trying to decide whether to perform liver transplant surgery.3 The finance department has prepared a financial projection of the expenses for the service, which appears in Table 8.2. Which elements of the table are not relevant costs for the decision? ❃ TABLE 8.2 ABC Hospital Liver Transplant Program Financial Projections Operating Expenses Nurse program supervisor $ 100,000 Operating room nurses (RN: circulating, scrub) * 190,000 Technician * 55,000 Orderlies * 40,000 Operating room receptionist/secretary * 35,000 Benefits * 126,000 Medical supplies * 300,000 Allocation of building depreciation 20,000 Allocation of operating room equipment depreciation 40,000 Telephone allocation * 7,500 Office supplies * 5,000 Allocation of malpractice insurance and overhead not included above * 300,000 Medical director fees 150,000 Nursing staff orientation and training costs 15,000 Computers and software 10,000 Fees (Center of Excellence, others) 20,000 Total $1,413,500 *Expense items assumed to vary directly with the number of operations. Note: These numbers are hypothetical and do not reflect the true costs related to liver transplantation. The first questionable item in the financial projection is the cost of the program supervisor. It is quite possible that one individual will be given the responsibility for the liver transplant program. However, one must question whether a new manager will be hired just for that position or whether the responsibility will be assigned to a manager who would be managing other units or working on various other things, in any case. Would the hospital actually spend more for supervisors if it had this program than it would if it did not? If the answer is that no more money is spent on supervision if the program exists than if it does not, the cost is not relevant. The costs of the staff in the OR during the transplantation are probably relevant. By adding more patients, more nursing staff will be needed. The same is true of the technician. On the other hand, it is not clear that the secretary and the orderly are relevant costs. The manager must assess whether the addition of the program will necessitate having an additional full- or part-time secretary or orderly. If existing staff carry a heavier burden with no staff additions, the costs are not relevant unless additional overtime is incurred. In that case, it would be the overtime cost that was relevant. Allocation of building overhead is clearly a nonrelevant cost. The building will depreciate in any case. Unless remodeling is done to accommodate the program or unless the organization must acquire additional space, building depreciation should not be considered in making a decision regarding whether to offer the program. The same is true for equipment depreciation. Assuming that no additional equipment is acquired for the program, depreciation is not a relevant cost. On the other hand, some new equipment will likely be required, and depreciation of any equipment specifically purchased for the program is a relevant cost. Malpractice insurance is a peculiar item. In some hospitals, it is based on past experience. In other hospitals, it is based on the sum of the riskiness of the patients and services of the hospital. Liver transplantation can be a risky operation. However, patients are aware of the associated risk. An assessment must be made concerning whether this program increases the risk of lawsuits and malpractice findings against the hospital. If so, the malpractice insurance costs will probably increase as a result of the program, and the cost is considered relevant. Accountants may still argue that a portion of the nonrelevant cost items must be assigned to the program so that the full costs of the liver transplantation program will be known. That is, “All costs must be allocated to all activities to be fair, or else why should any costs be allocated to any activities?” Managers would argue that not allocating costs would be preferable to allocating costs in such a way as to result in poor managerial decisions. If it is decided not to add the liver program because of costs allocated to it that are not relevant, then the accounting system is a hindrance to the organization’s success. The analysis of the new program must allow the manager to know how much more cost the organization will have with the program than without the program. Whether or not preexisting costs are allocated to the program, the important thing to keep in mind is that when decisions are made, relevant costs are the only costs that should be considered. ❃ Cost estimation techniques One of the most difficult parts of financial management is the prediction of costs. This is an essential component of financial management. Budgets contain cost estimates. Trying to predict how much will be spent on each type of expenditure in the coming year presents great problems for both inexperienced and experienced managers. Managers must have ways to estimate such costs. One approach is simply to look at what happened during the current year and to use the same information to predict next year’s cost, plus an additional amount, or increment, for inflation. At the other extreme is an approach that says it is desirable to do better next year than this year, so it is appropriate to budget a certain percentage less than was spent this year. In each case, the approach is far too simplistic. A priori, there is no reason to believe that next year will be just like this year, and simply wishing to spend less than in the current year will not make it happen. Some sort of clear method is needed that will help predict what will happen next year based on the past. Some way to formally consider why that prediction may not come true is also needed. Finally, if costs are to be reduced below the predicted outcome, a specific plan must be in place that the nurse manager believes will accomplish the cost cutback. This section considers several methods of cost estimation. Not all elements of the budget are simply costs. Items such as the number of patient days must be predicted as well. A discussion of general forecasting is presented in Chapter 20. Here the focus is solely on prediction of costs. Often historical information about costs incurred can be a great aid in predicting what costs will be in the future. This is especially true in the case of mixed costs, which have both fixed and variable cost elements. Cost estimation techniques look at historical information and compare the change in cost over time with the change in volume over time to isolate fixed and variable costs. If costs rise as volume rises, what could account for the increase in costs? Fixed costs, by definition, do not change as volume changes. Therefore any change in cost as volume changes must be attributable to the variable cost. By determining how much costs change for a change in volume, it is possible to calculate the variable cost per unit. After the variable cost per unit is known, the total variable cost for any volume can be determined by multiplying the variable cost per unit by the volume. Then fixed cost can be determined as well. The difference between the total cost for any volume and the total variable cost for that volume is the fixed cost. The fixed and variable cost information can then be used to estimate costs for the coming year based on a forecast of the volume in the coming year. There is one critical problem in the flow of logic that allows cost estimation. Changes in cost over time are assumed to be the result only of changes in volume. However, to some extent, changes in cost over time are the result of other changes. One possible type of change is evolving clinical practice. For example, a new technology may require more intravenous solutions than were needed before patients had access to that technology. Other reasons that costs may change include the implementation of evidence-based practices or the treatment of certain patient populations with new types of medications. Managers should always adjust any estimates they make according to the impact of any changes that they anticipate. Another problem is inflation. If inflation were a constant percent each year, one could argue that past inflation could be ignored and that inflation would be built automatically into predictions for the future. However, inflation rates tend to fluctuate from year to year. Over a period of years the fluctuations can be substantial. Therefore to predict fixed and variable costs, the data should be adjusted for the effect of inflation. Adjusting costs for inflation Suppose that a nurse manager is interested in determining the total increase in registered nurse (RN) staff costs for the unit in fiscal year 2019.4 This hypothetical unit is staffed with a minimum of 10 full-time equivalent (FTE) RNs for any volume up to 9000 patient days at a certain acuity level. The cost of those 10 FTEs is fixed because the unit will always have at least that cost. As volume increases above 9000 patient days, additional nursing time will be needed. In 2017 the patient days numbered 9800, and the cost, including fringe benefits, was $666,400. In 2018 the patient days totaled 11,000, and the cost was $792,000. The cost increase of $125,600 was attributable to both the increased volume and inflation. Most readers are probably familiar with the Consumer Price Index (CPI), the most widely used measure of inflation. The CPI and many other indexes of inflation, such as the hospital market basket index, were developed by or for the federal government. The CPI measures the relative cost of a typical basket of consumer goods. Whatever the basket of goods costs in the base year is considered to be 100% of the cost in that year, or simply 100. The index is revised, and a new base year is established from time to time. If it costs twice as much to buy the same goods in a year subsequent to the base year, the index would be 200% of the base year costs, or simply 200. The U.S. Department of Commerce’s Bureau of the Census annually publishes the Statistical Abstract of the United States. Included in that book are “Indexes of Medical Care Prices.” Several quite useful indexes are listed under that heading, including the index of medical care services and the hospital daily room rate index. The CPI is also published online and available for reviewing historical information. In this example, the nurse manager wants to forecast the variable cost per patient day of nursing labor for 2019 using current dollars as of the end of fiscal year 2018. If information from 2014 through 2018 is used, the nurse manager will have to find the value of an appropriate index in each of those years. The financial managers in most health care institutions can provide nurse managers with appropriate indexes adjusted for labor costs in the specific geographic area. Failing that, most library reference sections can assist with current index information. Assume that an appropriate index has values as follows: 2014 258 2015 287 2016 318 2017 357 2018 395 Suppose also that the following cost and volume information is available: Year Patient Days Cost 2014 8,000 $720,000 2015 8,700 809,100 2016 8,850 862,875 2017 9,800 996,600 2018 11,000 1,188,000 It appears that costs have risen from 2014 to 2016 even though volume is below 9000 patient days in each of those years. Because the staffing is fixed at 10 FTEs for any volume below 9000 patient days, the cost is expected to be about the same in each of those 3 years and to increase only as volume increases above 9000 patient days, thus requiring more nursing staff. The cost information, however, is not comparable because of the impact of inflation. Even if the number of FTEs did not change, the total cost would rise because of annual pay raises. To make the numbers reasonable for comparison purposes, they must be restated in constant dollars, that is, in amounts that have been adjusted for the impact of inflation. That adjustment can be made by multiplying the cost in any given year by a fraction that represents the current value of the index divided by the value of the index in the year the cost was incurred. This is not a complicated procedure. For example, in 2014 the cost was $720,000. The hypothetical index value is 395 for fiscal year 2018. In 2014 it was 258. Multiplying $720,000 by the fraction 395/258 results in a cost of $1,102,326, which is the 2014 cost adjusted to year 2018 dollars. Now the $1,188,000 spent when there were 11,000 patient days in 2018 can be compared with $1,102,326, the constant-dollar cost of 8000 patient days in 2014. In a similar fashion, all of the data can be restated in year 2018 dollars, as in Table 8.3. ❃ TABLE 8.3 Adjusting Costs for Inflation Year Patient Days Original Cost Index Fraction Adjusted Cost 2014 8,000 $ 720,000 × 395/258 = $1,102,326 2015 8,700 809,100 × 395/287 = 1,113,570 2016 8,850 862,875 × 395/318 = 1,071,810 2017 9,800 996,600 × 395/357 = 1,102,681 2018 11,000 1,188,000 × 395/395 = 1,188,000 Inflation accounts for changes in prices. One example of a change in prices is a change in wages. Wage rates are the price an organization pays for labor. Note that adjusted for inflation, there was very little change in costs from 2014 to 2016, the period during which the staffing was fixed, because patient days were less than 9000. However, the actual dollars spent in those years, before adjusting for inflation, rose from $720,000 to $862,875. Such an increase might at least partly reflect the impact of rising salaries. These index values and costs are hypothetical. Managers should consult their organization’s financial managers for an appropriate inflation index and its actual values for their specific geographic region. High-low cost estimation The high-low approach is a relatively simple, quick-and-dirty approach to cost estimation. It is unsophisticated and therefore not terribly accurate, but in many cases, it may be “good enough.” It certainly is better than simply taking a guess. The key to the high-low method is the fact that fixed costs do not change at all in response to changes in volume. The method examines the organization’s cost for a specific item over a period of approximately 5 years. Costs adjusted for inflation should be used, as described previously. The use of 5 years is arbitrary. It might be more appropriate to use a longer period, but one would not want to use data from much less than 5 years. Less than 5 years should be used only if there have been substantial changes in the unit that make earlier data no longer relevant. For the period chosen, find the highest volume and the lowest volume and compare the costs at these two volumes. The amount by which the costs changed from the lowest to highest volume should be compared with the amount by which the volume changed. In the example from the previous section, the highest volume in the last 5 years was 11,000 patient days, and the cost for nursing labor that year for the unit was $1,188,000 in constant dollars. The lowest volume in the last 5 years was 8000 patient days, and the constant dollar inflation-adjusted cost in that year was $1,102,326. In this case, whereas inflation-adjusted costs increased by $85,674, volume increased by 3000 patient days. If $85,674 is divided by 3000 patient days, the result is $28.558 per patient day. Although it is certainly likely that nursing labor is a step-fixed cost and therefore will not go up by $28.558 for each additional patient day, that volume provides a reasonable measure of the amount of additional nursing services needed per patient day when changes in volume are significant. If the variable cost per patient day is $28.558, what is the fixed cost?5 The yearly total variable cost is found first by multiplying the variable cost per patient day by the number of patient days ($28.558 × 8000 = $228,464). The total nursing labor cost for 8000 patient days was $1,102,326 in 2008; if $228,464 is the variable cost, then the remainder, $873,862, represents the fixed cost. Similarly, for 11,000 patient days at $28.558 per patient day, the variable cost is $314,138; given a total cost of $1,188,000 in 2018, the fixed cost would be $873,862. The fixed cost is expected to be the same at either volume level because by definition it is fixed. This fixed and variable cost information can be used in preparing next year’s budget. If 12,000 patient days are expected, costs will be expected to rise by $28.558 × 1000 patient days, or $28,558. The fixed cost portion will not change. Because this information was calculated using 2018 constant dollars, both the fixed and variable costs will have to be adjusted upward for the expected 2019 salary increases or, more generally, for the expected impact of inflation during the next year. The high-low method is not accurate because it considers only the experience of 2 years. One or both of the 2 years chosen may have had some unusual circumstance that would skew the costs in that year. A superior prediction is possible if some method is used that takes more experience into account. Regression analysis can provide such a prediction. Regression analysis The volume of patient days and the total cost for those days for a number of years can be plotted on a graph. The horizontal axis represents volume, and the vertical axis represents cost. The result is a scatter diagram. Each point on the graph represents a volume and the cost at that volume. If a line is drawn approximating the points, it can be used for future predictions. By selecting any expected volume on the horizontal axis, it is possible to go vertically up to the line and then from the line move horizontally across to a point on a vertical cost axis. That point represents the prediction of cost. For example, Fig. 8.7 shows a scatter diagram with a line drawn approximating the points. FIG. 8.7Predicting costs from a scatter diagram. The difficulty in drawing the diagonal line connecting those points is properly placing it so that it will give accurate predictions. Regression analysis is a technique that applies mathematical precision to a scatter diagram. Regression selects the one line that is effectively closest to all the individual points on the scatter diagram and that will therefore best predict cost for the future. This method can predict fairly accurately a breakdown of costs into their fixed and variable components. Simple linear regression analysis can take all available past information into account in estimating the portion of any cost that is fixed and the portion that is variable. The phrase simple linear regression refers to several issues. First, it is simple in the sense that there is only one dependent variable and one independent variable. Cost is the dependent variable being estimated. Cost depends on the value of the independent variable. An independent variable is a causal factor. For example, the most significant causal factor for nursing costs might be patient days. The more patient days, the greater the costs for nursing. Patient days cause costs to be incurred. For the admissions department of the hospital, it is not patient days but rather the number of patients that is important because admission time is the same for each patient regardless of the ultimate length of stay. The second part of the phrase simple linear regression refers to the presumption that cost behavior can be shown in a linear fashion—that is, using a straight line. For example, what if a slightly lower price per disposable supply unit is paid for every increase in volume (e.g., $0.75 for one unit, $0.7499 per unit for two units, $0.7498 per unit for three units, and so on)? In that case, Fig. 8.2 is not an accurate reflection of how variable costs change. It is necessary to draw a curved line on the graph, but the mathematics behind determining curved lines instead of straight lines is far more complicated. Variable costs generally are treated as if they are linear, even if that is only an approximation of their true behavior. Finally, the term regression refers to trying to regress, or bring all the points from the scatter diagram as close as possible to the estimated line. For example, suppose that the CNE desires to make a rough starting prediction for the total cost of all patient care units in a hospital for the coming year. If last year there were 50,000 patient days and this coming year patient days are expected to be 52,000, there is a 4% expected increase in the number of patient days. However, it cannot be assumed that all costs of running the unit will go up by 4%, because some costs, such as the salary of the CNE, are fixed and will not rise in proportion to the number of patient days. The high-low method is one way to make the prediction, but this method relies on only two data points. Far greater accuracy in breaking out fixed and variable costs is possible if the costs from past years are examined on a very detailed basis, cost item by cost item, to determine which were fixed and which were variable. That is a very time-consuming procedure. Gathering information costs money, and even if the information is gathered, there are always some costs that cannot be separated into fixed and variable components without using some estimating method, because they are mixed costs. For example, nonmedical supplies, such as paper, pens, and forms, are needed to some extent regardless of patient volume. On the other hand, the more patients, the more nonmedical supplies used. How can costs be divided into their fixed and variable components and less money be spent on gathering information than if each line item from past years is examined? Regression analysis can help separate these mixed costs. Mixed costs and regression analysis Suppose it is known that last year the combined cost of the salary of the nurse manager and the disposable supplies was $105,000 and that the number of patient days was 50,000. This uses the information represented in the graphs presented in Figs. 8.1 through 8.3. The volume of patient days is expected to rise to 52,000 next year. Should the $105,000 cost be increased by 4% because volume is increasing by 4%? No. Some costs are fixed. Only variable costs increase as volume increases. In this scenario, one would expect costs to increase by $200, or $0.10 for each extra patient day, because it is known in this simple example that variable costs for the disposable thermometer covers were $0.10 per patient day. However, in dealing with a more realistic example with many different fixed, variable, and mixed costs, the variable costs per patient day are not necessarily known. Suppose that the historical information in Table 8.4 were available (already adjusted for inflation using the indexing technique). If the high-low technique were used to evaluate the fixed and variable costs, there would be a very strange result. The highest cost is $105,000, and the lowest cost is $104,100; thus, costs have risen by $900. At the same time, volume has increased from 40,000 patient days to 50,000, or an increase of 10,000. When $900 is divided by 10,000 patient days, a variable cost of $0.09 per patient day results. Is that an accurate estimate? No, because it is known that the variable cost is $0.10 per patient day. What might be the cause of the discrepancy? ❃ TABLE 8.4 Historical Data for Thermometer Cover Costs Year Patient Days Cost 2009 40,000 $104,100 2010 42,000 104,200 2011 43,000 104,300 2012 44,000 104,400 2013 45,000 104,500 2014 46,000 104,600 2015 47,000 104,700 2016 48,000 104,800 2017 49,000 104,900 2018 50,000 105,000 It is possible that 2009 was the first year that a new disposable oral thermometer cover was used. Perhaps many of them were defective and were thrown away, or perhaps some were wasted because of lack of familiarity with using them. In any case, if more than $0.10 per patient day was spent on disposable thermometer covers in the low-volume year, the costs were unduly high in that year. Therefore the change in cost from 2009 to 2018 looks unrealistically low, and the variable cost measure is unrealistically low. At the other extreme, had there been unusual waste (perhaps because of nursing practice or use by other clinicians, but possibly because of quality problems with a large batch of the disposable item) in the most recent, high-cost year, the change in cost would look especially high, and the variable cost per unit would have come out to more than $0.10. As has been stated before, if one relies on just two data points, as the high-low method does, results are subject to the whims of unusual events in either of those years. In reality, one would not expect to use exactly $0.10 per patient day on disposable supplies in any year. For one reason or another, some patients will have their temperature taken only once on a given day. This might be caused by admission to the hospital late in the day, for instance. On the other hand, patients with a fever will no doubt have their temperature taken more often. A more likely pattern of costs is shown in Table 8.5. ❃ TABLE 8.5 More Likely Historical Cost Pattern Year Patient Days Cost 2009 40,000 $104,100 2010 42,000 104,180 2011 43,000 104,360 2012 44,000 104,380 2013 45,000 104,530 2014 46,000 104,670 2015 47,000 104,690 2016 48,000 104,880 2017 49,000 104,890 2018 50,000 105,000 Simply looking at this list does not provide a lot of insight about fixed and variable costs. Fig. 8.8 shows a scatter diagram for these data. One can roughly see how costs increase as volume increases. A straight line cannot be drawn through all of the points on this scatter diagram. However, the regression technique uses all of the available information to select a line that will provide the best estimate in the absence of any other information. Regression analysis uses information about the dependent and independent variables in the past to develop an equation for a straight line. As part of that process, it calculates a constant value and a coefficient for the independent variable. If the dependent variable is the cost, the constant represents the fixed cost, and the coefficient of the independent variable represents the variable cost. Regression analysis is a statistical technique. A detailed discussion of statistics is beyond the scope of this book. However, many mechanical approaches to regression have made it a workable tool used in HCOs. Regression can be performed easily using a wide variety of statistical and spreadsheet programs for personal computers and many handheld devices (including “old fashioned” calculators). Turning back to the scatter diagram in Fig. 8.8, one can see that it is possible to use regression analysis to predict what the costs will be for the nursing unit next year if there are 52,000 patient days. Regression analysis will determine a specific line to plot through this scatter diagram that will give the best possible estimate of fixed and variable costs and therefore allow prediction of the cost next year. FIG. 8.8Scatter diagram of total costs for nursing. Basically, the process requires several simple steps. First, determine the cost associated with each volume of patient days. For instance, when there were 40,000 patient days, the cost was $104,100. The independent variable, patient days, is often referred to as the X variable because it is plotted on the horizontal axis. The dependent variable, cost, is often referred to as the Y variable because it is plotted on the vertical axis. When using such data in a computer program or spreadsheet, the X and Y values for each year must be indicated. With that information, it generally is necessary only to give a command to compute the regression to complete the process. Although regression is a quantitative tool often learned in a statistics course, in practice it is a tool that supports managers in doing their jobs and making better decisions. The major difficulty in using regression is having a conceptual understanding of what the tool does, as well as learning the software program. In the example, regression analysis using the data in Table 8.5 predicts that the fixed cost is $100,226 and that the variable cost is $0.0956 per unit. Fig. 8.9 shows the resulting line. If extended to the left, it would have its intercept at $100,226, increasing with a slope of 0.0956. These figures are not exactly the expected variable cost of $0.10 per patient day and fixed cost of $100,000 for the salary of the nurse manager. They are, however, better estimates than those the high-low method would give. The high-low approach predicts a fixed cost of $100,000 and variable cost of $0.09 per patient day. Regression analysis is an inexpensive, potentially very useful, and relatively simple way of estimating fixed and variable costs and helping predict future costs. FIG. 8.9Simple linear regression of total costs for nursing. For any number of patient days predicted, it is now possible to multiply by 0.0956 and then add $100,226 to get a forecast of future costs. With many computer software programs and packages, the process is made even simpler by requiring only that the forecast volume be entered into the computer along with the historical data. The software will then generate the estimated cost for the coming year automatically based on the forecast volume. Remember, however, that it is necessary to adjust upward the resulting cost for expected increases caused by inflation for the coming year. Although you will soon find this approach quite simple, it is useful only as a tool to aid in managing. It should not be allowed to take over the role of judgment. The mathematical model is quite accurate in predicting the future if nothing has changed. It is the nurse manager’s or executive’s role to know whether there are reasons that costs are likely to change from their past patterns. For instance, if it is known that in 2009 there was excessive waste of disposable thermometer covers, you might want to eliminate that year from the analysis. If you do, your regression results will show fixed costs of $99,997 and variable costs of $0.1005. Recall that the fixed costs were actually $100,000 and the variable costs $0.10. As you can see, the input of judgment into the process can substantially improve the resulting estimates. When regression analysis is performed, one statistic that is generally provided by the calculator or the computer is R-squared (denoted as R2). That value can range from a low of zero to a high of 1.0. A value close to zero means that the independent variable does not do a very good job of explaining the changes in the dependent variable. An R2 value of 0.20, for example, might indicate that patient days are not a good predictor of nursing cost. On the other hand, an R2 of 0.80 would indicate a very good predictor. However, it is possible to become even more exact in estimating costs. Multiple regression analysis There is a type of regression analysis that is more sophisticated than simple linear regression. It is called multiple regression because it allows for the use of multiple independent, or causal, variables. The use of simple linear regression can be a substantial aid in estimating future costs because it is so efficient at predicting the fixed cost and the variable cost per unit when there is one major independent variable. Sometimes, however, there are several key variables. For instance, suppose that nursing costs vary not only with the number of patient days but also with the number of patients. That is most probably the case. Certainly, the costs vary with the number of patient days. The more patient days, the more temperatures to be taken, pulses to be checked, medications to be administered, and so on. However, certain costs vary with the number of patients, not with the number of patient days (e.g., a health history must be recorded, an electronic chart must be set up, a patient care plan must be established, valuables must be stored, orientation must be given, discharge planning and education must be done, and so on). Thus there may be instances when having fewer patients with a long length of stay will cost less than having more patients with a short length of stay, even if the total patient days are the same. For example, consider 10 patients each staying on a unit 5 days versus 5 patients each staying 10 days. There are 50 patient days in both cases, and the costs of taking vital signs and giving medications are the same. In the 5 patient case, the number (and cost) of admissions, histories, charts, plans, and discharges will be lower. Thus it is likely that the cost of a nursing unit varies with both the number of patient days and the number of admissions. Most handheld business calculators cannot perform multiple regression. However, most statistical programs for personal computers, tablet computers, and other handheld computerized devices can handle this easily. Instead of simply entering the X and Y values for each year into the calculator or computer, the user enters an X value for the historical information for each of the independent variables as well as the Y value. To predict a future cost, the user provides the computer with, for example, the expected number of patient days and the expected number of patients to predict the expected costs. Sometimes the multiple regression level of sophistication adds extra work and complexity without substantially changing the results. Recall that when all is said and done, the result is just an estimate; all types of events can happen in the future that will throw off the estimate, no matter how finely tuned it is. It is not necessary to add complexity for its own sake. At times, however, multiple regression can produce information that otherwise would not be available. For example, more and more attention has been placed on measures of patient acuity, or the level of intensity of required nursing services. It certainly is clear that the amount of nursing services varies not only with the number of patient days but also with the severity of the patients’ illnesses. If data about the number of patient days and the average acuity level are used as independent variables, the accuracy of estimated costs might improve substantially. Another use for multiple regression is in investigatory work with respect to costs. Suppose that there is a strong feeling by the nursing staff that the way a particular physician practices medicine is extremely costly. This is common in the OR, where particular surgeons often exhibit out-of-the-ordinary behavior. The number of operations by a specific physician each year can be used as an independent variable. Costs increasing as a result of more cases by that physician will show up as a positive coefficient for that independent variable. The nurse manager will then have evidence to support the more general feelings of the staff that the physician is an unusually high consumer of resources. Readers of this book are encouraged to pursue the topic of regression analysis further. This should be done on both a conceptual basis, reviewing the underlying principles and theories of regression analysis, and on a practical basis, using a computer software package to perform some regression analyses for financial decision making. ❃ Break-even analysis Up to this point, the general behavior of costs (fixed versus variable) has been discussed, as well as the techniques for cost estimation (high-low and regression). Attention will now be focused on using cost information for understanding whether a particular unit or service will lose money, make money, or just break even. This technique is useful for the evaluation of both new and continuing projects and services. It is often used in developing a business plan. Business plans are discussed in Chapter 21. Nurse managers and executives in many instances find it necessary to be able to determine whether a program or service will be profitable. One key to profitability is volume. Prices are often fixed. Average cost, however, is not fixed. As the number of patients rises, the cost per patient falls because of the sharing of fixed costs. One cannot simply compare price and average cost and determine that a program or unit will make a profit or a loss. To determine whether something will be profitable, it is critical to know the volume of patients. As noted earlier, BEA is a technique to find the specific volume at which a program or service neither makes nor loses money. Forecast information about the likely volume of the service can be compared with break-even volume to predict whether there will be profits or losses. BEA is based on the following formula: or where Q is the number of patients needed just to break even, FC is the total fixed cost, P is the price for each patient, and VC is the variable cost per patient. At a quantity lower than Q, there will be a loss; at a quantity higher than Q, there will be a profit. The basis for the formula is the underlying relationship between revenues and expenses.6 If total revenues are greater than expenses, there is a profit. If total revenues are less than expenses, there is a loss. If revenues are just equal to expenses, there is neither profit nor loss, and the service is said to just break even. Expenses are the sum of total fixed costs and total variable costs. Example of break-even analysis Suppose that a new home health agency opens in a rural area.7 It charges, on average, $50 per visit. The agency has fixed costs of $10,000 and variable costs of $30 per patient visit. If there are no patients at all, there is no revenue, but there are fixed costs of $10,000, and there is a $10,000 loss. If there were 100 patients, there would be $5000 of revenue ($50 × 100 patients), $10,000 of fixed costs, and $3000 of variable costs ($30 × 100 patients). Total costs would be $13,000 ($10,000 of fixed costs + $3000 of variable costs), revenues would be $5000, and there would be a loss of $8000. Each additional patient brings in $50 of revenue but causes the agency to spend only $30 more. The difference between the $50 price and the $30 variable cost—$20—is called the contribution margin (CM). A positive CM means that each extra unit of activity makes the organization better off by that amount. The CM from each patient can be used to cover fixed costs; if all fixed costs have been covered, the CM represents a profit. In this example, when there are 100 patients, there is $20 of CM for each of the 100 patients, or a total CM of $2000. Note that the loss with zero patients was $10,000, but it was only $8000 when there were 100 patients. The loss decreased by $2000, exactly the amount of the total CM for those 100 patients. How many visits would the agency need to break even? The answer is 500. If each additional patient generates $20 of CM, then 500 patients would generate $10,000 of CM (500 patients × $20 = $10,000), exactly enough to cover the fixed costs of $10,000. If the agency has 500 patients, it will just break even. This could have been calculated using the formula or BEA can also be viewed from a graphic perspective, as shown in Fig. 8.10. The total cost line starts at $10,000 because of the fixed costs. The total revenue line starts at zero because there is no revenue if there are zero patients. Where the revenue line and the cost line intersect, they are equal, and the agency just breaks even. Note that with fewer patients than at the break-even point, the cost line is higher than the revenue line, and there will be a loss; with more patients than at the break-even point, the revenue line is higher than the costs, and a profit is made. FIG. 8.10Break-even analysis. Before the introduction of prospective payment systems, Diagnosis Related Groups (DRGs), and now value-based purchasing, most break-even analyses in hospitals focused on the number of patient days needed to break even. Hospitals no longer get paid for extra patient days. Therefore attention is now focused on the total number of patients needed to break even, rather than on patient days. BEA can also be performed based on the number of surgeries or clinic visits or other appropriate service unit volume measures. When there are different types of patients, BEA becomes somewhat more complicated. The formula presented at the beginning of this section assumes that there is only one price and one variable cost and therefore one CM. If there are different types of patients with different prices and different variable costs, it is necessary to find a weighted-average CM. That weighted average can be divided into the fixed costs to find the break-even volume for all patients. For example, suppose that there are three classes of home care visits, referred to here as complex, moderate, and simple. The price for the visits is $80, $50, and $30, and the variable costs for the visits are $55, $30, and $20, respectively. The CM for each type of visit can be calculated by subtracting the variable cost from the price, as shown in Table 8.6. ❃ TABLE 8.6 Contribution Margin by Type of Patient Home Visit Price (A) Variable Cost (B) Contribution Margin (C = A − B) Complex $80 $55 $25 Moderate 50 30 20 Simple 30 20 10 The crucial piece of information for calculating the break-even point is the relative proportion of each type of visit. Management of the home health agency expects that 20% of all visits are complex, 30% are moderate, and 50% are simple. This information can be used to determine a weighted-average CM. This requires multiplying the individual CM for each type of visit by the percentage of patients receiving that type of visit. The results are added together to get an overall weighted-average CM, as shown in Table 8.7. ❃ TABLE 8.7 Calculation of Weighted Average Contribution Margin Visit Type Percentage of Visits Contribution Margin Weighted Average Contribution Margin Complex 20 × $25 = $ 5 Moderate 30 × 20 = 6 Simple 50 × 10 = 5 Total 100 $16 This $16 weighted CM represents the average CM for all types of visits. It can be used to calculate the break-even quantity. Assume that fixed costs are $10,000. The break-even quantity of visits is as follows: or Of the total of 625 visits needed to break even, 20%, or 125, are complex; 30%, or 188, are moderate; and 50%, or 312, are simple. This method works for three different kinds of patients. What if there are more than three kinds? The same weighted-average approach that can be used to find the break-even volume when there are three different types of patients can be used even if there are hundreds of different types of patients. What if there is more than one price for each type of visit? Medicaid pays one price, Medicare another, private insurers another, and self-pay patients yet another. This still can work within the same framework as has been presented. It will be necessary to calculate a weighted-average CM, treating each payer for each type of visit as a separate group. For example, if Medicaid pays $40 per visit regardless of the type of visit and the other payment rates are the same as indicated earlier, the CM by type of visit by payer will be as shown in Table 8.8. ❃ TABLE 8.8 Contribution Margin by Type of Patient Home Visit and by Payer Visit Type Price (A) Variable Cost (B) Contribution Margin (C = A − B) Complex Medicaid $40 $55 ($15) Complex other 80 55 25 Moderate Medicaid 40 30 10 Moderate other 50 30 20 Simple Medicaid 40 20 20 Simple other 30 20 10 If it is possible to anticipate the percentage of each type of visit, a weighted-average CM can be estimated. Assume that 10% of all visits are Medicaid complex visits and 10% are other complex visits. Assume that 20% of all visits are Medicaid moderate visits and 10% are other moderate visits. Assume that 30% of all visits are Medicaid simple visits and that 20% are other simple visits. The weighted-average CM will be as shown in Table 8.9. ❃ TABLE 8.9 Calculation of Weighted-Average Contribution Margin With Multiple Payers Visit Type Percentage of Visits Contribution Margin Weighted-Average Contribution Margin Complex Medicaid 10 × ($15) = ($1.50) Complex other 10 × 25 = 2.50 Moderate Medicaid 20 × 10 = 2.00 Moderate other 10 × 20 = 2.00 Simple Medicaid 30 × 20 = 6.00 Simple other 20 × 10 = 2.00 Total 100 $13.00 The break-even volume can then be calculated as follows: or The number of visits of any type can be determined by multiplying the 769 break-even volume times the percentage of visits in any given class. For example, because 30% of the visits are Medicaid simple, 30% of 769, or 231, Medicaid simple visits can be expected at the break-even level. Some managed care organizations (MCOs) pay home care agencies on a case basis rather than a visit basis. With some patients paid on a case basis and others on a visit basis, these calculations become complex. Using break-even analysis for decision making If a particular service is expected to have a volume of activity well in excess of the break-even point, managers have a clear-cut decision to start or continue the service. If the volume is too low to break even, several options exist. One approach is to lower the volume needed to break even. There are three ways to reduce the required break-even level. One approach is to lower the fixed costs. In some cases, it might be possible to do that. Another alternative is to increase prices. Price increases will increase the CM per patient. This strategy will also have the effect of lowering the break-even point. However, price increases might reduce the expected volume. In that case, the price increases will defeat their purpose. Also, prices are sometimes regulated and beyond the control of the organization. Finally, one can try to reduce the variable cost per unit. This might be accomplished by increased efforts toward improved efficiency. If it is not feasible to change fixed costs, price, or variable costs, an organization can try to attract more patients so that volume will rise above the break-even point. In the example presented, what type of visits would be desirable? The most desirable type of visit is a non-Medicaid complex visit. That visit yields a CM of $25. The least desirable is a Medicaid complex visit. The CM is negative. For each additional Medicaid complex visit, the agency loses money. In this particular example, the most attractive visit brings the highest revenue. However, the focus should not be on revenue. If the highest-revenue visit also has extremely high variable costs, that visit might not be as attractive as one with lower revenue and much lower variable costs. The attractiveness of additional visits is determined by how much CM they provide to the organization. Break even and capitation As managed care has become more and more prevalent, many negotiated contracts call for capitated payments. In such an arrangement, the MCO pays the health care provider a set amount for each member for each month. This is called the per member per month (PMPM) payment. Under capitation, an increase in the amount of services provided to patients will not cause revenues to increase at all. On the other hand, an increase in the number of members will increase revenues. Suppose that an MCO offered to pay a home health agency $1 PMPM to provide all home care services for all of its members. Over the course of a year, revenue would be $12 per member ($1 PMPM × 12 months = $12). Assume that the agency’s variable costs are $30 per visit. Furthermore, the agency will have increased fixed costs of $10,000 if it takes the MCO members. How many members will the MCO need to have for the home health agency to break even on the contract? We can use BEA to calculate the break-even number of members. The fixed costs are $10,000. The variable costs are $30 per visit. The price is $12 per member per year. However, we do not have enough information to calculate the break-even point because we know only the variable cost per visit. We need to know the variable cost per member per year. To find the break-even volume of members, it is necessary to predict the utilization levels—that is, how many home visits each member will have. Most individuals will not need any visits in a typical year. Suppose that the average person consumes 0.3 visits in a given year. The variable cost for 0.3 visits per year is $9 (the $30 variable cost per visit multiplied by 0.3 visits per year). That $9 represents the variable cost per member per year. We can now calculate the break-even point as follows: Both the price and the variable costs in the calculation are per member per year. The quantity calculated represents the number of members needed to break even. If the MCO guarantees 5000 members, the agency will likely make a profit. If the MCO has only 2000 members, the agency will lose money. What if the agency has other fixed costs as well? Are they needed for the calculation? No. The determination of whether the MCO contract is profitable depends only on the marginal costs of the contract. Fixed costs that exist whether the agency contracts with the MCO or not are not relevant to the decision or the calculation. What if the MCO contract did not cause fixed costs to rise at all? Then the contract would be profitable for the agency as long as the revenue per member per year exceeded the variable cost per member per year. Break-even analysis cautions A few words of caution are advisable when working with BEA. First, after a break-even point has been calculated, one must decide whether it is likely that volume will actually be achieved. That requires a volume forecast (see Chapter 20). To the extent that the forecast of volume is incorrect, the decision to go ahead with a new service may turn out to be a bad one, even if the BEA is perfect. Another potential problem is that BEA assumes that prices and costs are constant. If it can reasonably be expected that prices will fall over time, a higher volume will be needed to keep a service viable unless variable or fixed costs fall as well. On the other hand, if prices are expected to rise faster than costs, a marginal service today may become profitable over time even without an increase in volume. Another consideration is the assumption that the mix of patients will stay constant. Suppose that in the earlier example, over time, there are more and more Medicaid complex visits. The CM for such visits is negative. If the demographics of the population are such that a shift in mix in that direction is likely, the results of the BEA require close scrutiny. Will there be enough of those visits to shift a profitable service over to a loss? As with all budgeting tools, judgment is essential. The nurse manager or executive must examine the assumptions of any modeling technique through experience, insight, and thought and consider the reasonableness of the results. If a result does not seem to make sense, often that is because it does not! However, BEA is a tool that can help give a manager a firm starting point in understanding whether a project or service is likely to be financially viable. ❃ Implications for nurse managers and executives Assessing costs is complex. In general, costs do not increase in direct proportion with volume. The implications of this fact are that if money is lost on a particular program, the solution may be to increase patient volume for that program. More patients do not necessarily mean greater losses. It is possible that volume increases can turn a loss into a profit. Understanding how that can happen requires an understanding of cost behavior. Some costs are fixed; others are variable. The result of that basic nature of costs is that the cost per patient will decline with increasing volume. The greater the number of patients who share the fixed costs, the lower the average cost per patient. Costing is further complicated by the fact that additional patients do not cause costs to increase by the average cost per patient. Decisions that are in the organization’s best interests often require marginal cost analysis. An important part of the budgeting process is the prediction of costs. Estimated costs can be based on historical cost information. Some estimation relies on using the historical information to isolate variable costs from fixed costs. To make such calculations, it is first necessary to convert historical cost information into common or constant dollars. This requires indexation of costs to account for the impact of inflation. Indexation is a process that adjusts a dollar value for the impact of inflation over a period of time by using a price index, such as the CPI. A price index is a tool that indicates year-to-year changes in prices. Through the use of indexed historical costs, the results of the cost estimation process will be in constant dollars. In preparing next year’s budget, the cost estimate has to be adjusted upward by the anticipated inflation rate over the next year. After constant-dollar information is available, costs can be estimated using the high-low method, simple linear regression, or multiple regression analysis. The ability to estimate fixed and variable costs is a potentially valuable tool. To apply the results, however, projections of the estimated number of patients, patient days, acuity level, and so forth are needed. Chapter 20 focuses on the process of forecasting such data. Break-even analysis, or BEA, is a tool that allows one to focus specifically on the quantity of patients needed for a program, project, or service to be financially viable. Its foundations are in fixed and variable costs. At low volumes of patients, the average cost may surpass the revenue per patient. As the number of patients increases, the cost per patient falls because fixed costs are shared by more patients. Eventually, the cost per patient falls below the revenue. BEA allows the manager to determine what the break-even quantity is so that a reasonable decision can be made about the likely financial viability of a program, project, or service. From the perspectives of nurse managers and executives, the topics of this chapter have critical implications. At the most basic level, falling volume will mean rising cost per patient. In such cases, it is likely that a revenue crisis will exist, and actions to restrain costs should be contemplated immediately. On the other hand, rising volumes represent an opportunity. They not only bring in more revenue but also decrease average cost per patient. Therefore there is the opportunity for profit from more patients and for more profit from each patient. Profits ultimately allow the organization to replace buildings and equipment, add services, add staff, improve quality, and raise salaries. In addition, in preparing budgets, nurse managers and executives should take into account the behavior of costs. The fact that certain costs vary in proportion while others are fixed may reorient the manager’s thinking from the notion that a 10% increase in volume requires 10% more resources. This in turn can allow the manager to prepare budgets in a more sophisticated and exact manner. Finally, managers and executives should remember that decisions to begin, continue, or stop a program are not made solely on the basis of profit-and-loss projections. Some programs are continued as a community service, and the institution makes a conscious decision to subsidize them from other sources of revenue. However, such decisions require as much information about anticipated profits or losses as possible. ❃ Key concepts Service unit Basic measure of the product or service being produced by the organization, such as discharged patients, patient days, home care visits, ambulatory care visits, emergency department treatments, or hours of operations. Direct costs Costs that are incurred within the organizational unit for which the manager has responsibility, or costs that are directly related to patient care. Indirect costs Costs that are assigned to an organizational unit from elsewhere in the organization, or unit costs that do not directly relate to patient care. Full cost Total of all costs associated with an organizational unit or activity. This includes direct and indirect costs. Average cost Full cost divided by the volume of service units. Fixed costs Costs that do not change in total as the volume of service units changes within the relevant range. Variable costs Costs that vary directly with changes in the volume of service units. Relevant range Normal range of expected activity for the responsibility center or organization. Marginal costs Extra costs incurred as a result of providing care to one more service unit, such as for one extra patient day. If one considers the full or total costs incurred by an organization before it makes a change and the total costs after it makes the change, the difference in costs represents the marginal costs of that change. Mixed costs Costs that contain both fixed and variable cost elements. Impact of volume on cost per patient Average cost declines as the volume of patients increases, because more patients are sharing the fixed costs. Therefore HCOs almost always find higher volume preferable to lower volume. Marginal cost analysis Decisions about changes should be based on the marginal costs of the change, not on full or average costs. Cost estimation Prediction of costs. This process is complicated by the necessity to divide historical mixed costs into their fixed and variable cost components and by the necessity to adjust historical costs for the impact of inflation to use them for predicting future costs. Adjusting costs for inflation Part of the change in costs over time is a result of volume changes, but part is attributable to inflation. To adjust for the impact of inflation, a historical cost must be multiplied by the current value of an appropriate price index divided by the value of that index when the cost was incurred. Regression analysis After historical costs have been adjusted for the impact of inflation, regression analysis can be used to estimate the fixed and variable costs. Costs are the dependent variable, and service units are the independent variable. The constant term of the regression represents fixed costs, and the coefficient of the independent variable represents the variable costs. Multiple regression analysis and cost estimation Technique that allows superior cost estimates by incorporating information from several independent variables rather than just one. Break-even analysis (BEA) Technique that allows the user to determine the volume of patients required for a program or service to be financially self-sufficient. At volumes above the break-even point, a profit is made; below that point, a loss occurs. BEA is based on the following formula: Contribution margin (CM) Price minus the variable cost per service unit. This represents the additional financial benefit to the organization from each additional service unit. This benefit can be used to cover fixed costs or provide a profit. A weighted-average CM is used for break-even analysis when there is more than one type of service unit or more than one price for each type service unit.
AGFOX
Chapter8.docx
0
