The goal of this chapter is to make the following points:
Taxpayers, employers, and employees demand that medicine deliver better value. Physicians need evidence-based medicine to establish which interventions are truly beneficial and modern management techniques to implement those interventions.
Perioperative interventions are investments, each with its costs and, it is hoped, its benefits. The benefits of perioperative interventions are often difficult to quantify in precise dollar amounts (e.g., pain relief).
Three ways of measuring the benefits of health-care interventions are by improved clinical results (i.e., effects), by increased quality-adjusted life-years (QALYs), or by assigned monetary benefit.
Any new intervention being considered must be evaluated in comparison to the best existing alternative. As a result, the fundamental concept is the incremental cost-to-effectiveness ratio.
Because of the multiplicity of health-care stakeholders (i.e., patients, providers, payers, and society as a whole), an economic study in health-care must specify ahead of time—and be consistent in—its point of view.
Costs of perioperative interventions are direct, indirect, and intangible. Direct costs are the easiest to define and quantify, but vary depending on the costing method used. Direct costs decrease over time, because of competition, the learning curve, technological progress, work process redesign, and the bundling of interventions.
Much more attention should be directed toward identifying and addressing barriers to implementation of beneficial interventions. These barriers include lack of awareness (the physician does not know about the new intervention), of familiarity (knows intervention exists but not the details), of agreement (physician does not agree with proposed intervention), of self-efficacy (does not think they can do it), of outcome expectancy (does not think it will work), as well as system factors not allowing successful implementation.
With the aging of the postwar baby boomers and the explosion of technological options at medicine’s disposal, health-care systems around the world, whether fee-for-service or publicly funded, struggle to increase the quality and to decrease the cost of care. As a result, it is more important than ever to know what really works and what does not. The ultimate payers of health-care costs (taxpayers, employers, and employees) demand that physicians provide better value for the money spent. “Pay-for-performance” metrics are one attempt by these payers to reward systems for attaining the biggest bang for its health-care buck.
Physicians must first know which perioperative interventions actually benefit patients and then apply those interventions at the lowest possible cost. This is a moving target, as the economic attractiveness of perioperative interventions changes with time, as new interventions are introduced (e.g., new drugs such as oral anticoagulants), old interventions evolve (e.g., enhanced recovery after surgery), or as new discoveries are made.
In this chapter, we present the conceptual framework used for the economic analysis of interventions so that the reader can critically appraise economic studies of perioperative optimization.
Health-Care Interventions as Investments
Economics describes how people in societies satisfy their needs and wants in an environment of limited resources. During the 19th and 20th centuries, economists, political scientists, and politicians debated the merits of free-market capitalism, with individual freedom of choice and private property versus central planning and state ownership of the means of production. For normal goods, the individualistic free market approach is predominant.
In medicine, however, much central planning occurs because the consumer (the patient) usually does not directly pay the full cost of the care and is often not able to judge the quality of the health-care product. In economic language, physicians are agents for patients, meaning they are asked to make decisions and “invest” on behalf of their patients.
Costs and (We Hope) Benefits
Health-care interventions are investments, very similar to financial investments, in that there is a cost of the intervention, and then there is an outcome —and, one hopes, a benefit. Unfortunately, many medical practices have not been studied adequately to find out whether they are really beneficial, so some of our therapy is of uncertain value. The motivation of evidence-based medicine is to find out whether our costly interventions really do, in fact, produce benefits. Once we know that an intervention actually provides a benefit to the patient, we have to decide whether the increased benefit is worth the increased cost.
No medical intervention can be considered cost effective in isolation but must be compared with an alternative treatment (usually, the standard therapy). This is hugely important. For example, if a new analgesic technique costs $100 and reduces postoperative pain scores by 50%, then its value needs to be compared with the best current treatment to determine whether it is worthy of implementation.
In health-care—as in any type of management—we manage at the margin, constantly analyzing marginal, or incremental, benefits and comparing them with the marginal, or incremental, costs. The fundamental analytical concept of managing at the margin is the incremental cost-effectiveness ratio , in which the incremental cost of the proposed new intervention is divided by the increase in effectiveness achieved by the intervention.
If a new technique achieves a better outcome at lower cost (e.g., polio vaccination compared with polio treatment), the new cost-saving technique is said to dominate the old technique, and it should become the new standard treatment. For example, ultrasound-guided nerve blocks are superior to nerve stimulator-guided blocks. The example is imperfect, however, since ultrasound-guide blocks require initial investments in equipment and training.
Health-care interventions differ from financial investments in one key respect. The primary benefits of health-care for the patient—such as improved health, improved function, and pain relief are difficult to quantify in precise dollar amounts. As an example, if the new postoperative analgesic reduces pain by 50%, how much is that pain reduction worth in monetary terms to both the patient and the payer?
The Agency Problem—Providers as Fiduciaries
Doctors are not the primary beneficiaries of the interventions they choose for their patients. Furthermore, doctors may benefit from the performance of interventions with unclear patient benefit. For example, knee arthroscopy in the treatment of knee pain in patients who are middle-aged or older has “inconsequential” benefit when compared with conservative therapy.
Within provider systems, parochial budgetary concerns may interfere with the simple goal of maximizing benefits to the patient. For example, a hospital pharmacy committee may not want to burden the hospital pharmacy budget by spending extra money on rivaroxaban, despite evidence of better patient experience and less fatal bleeding with rivaroxaban than with warfarin. This silo problem whereby each silo worries about their budget more than the impact on the entire health system is a continuing challenge that must be recognized.
In summary, health-care suffers from the agency problem: Will the CEO of a large corporation and other members of top management (the providers and payers) act in the interest of the “shareholders” (the patients), or will they enrich themselves first, and only later think about what is best for the shareholders?
Health-Care Economic Studies Offer Challenges
The Point of View Must Be Clear and Consistent
Because there are multiple actors in health-care, it is essential that we remain consistent in the point of view of our analysis. Are we looking at an issue from the point of view of the patient, the entity paying the bills, the provider (the physician, hospital, or clinic), or perhaps society as a whole?
Our Goal Is to Know What Works and How Much It Costs
If we were able to surmount the difficulties involved in quantifying the monetary and nonmonetary costs and benefits of our perioperative interventions, and if we were able to specify from whose point of view we were analyzing the issue, we might be able to construct a chart detailing the utility and costs of perioperative interventions.
The US Public Health Service Panel on Cost Effectiveness in Health and Medicine recommends that studies :
Adopt a societal perspective.
Use community- or patient-derived preference weights for utilities (as opposed to expert opinion).
Use net costs (cost of intervention minus savings in future medical costs).
Use appropriate incremental comparisons.
Discount costs and QALYs at the same rate.
Despite the recommendation to approach analysis from a societal perspective, namely from the point of view of all involved parties with all outcomes and costs, the majority of studies do not follow the recommendation. As of 2005, only 29% of studies involving QALY included a complete societal perspective (most were missing key elements such as transportation and caregiver time).
The consensus recommendations were updated in 2016 to include “reference case” and “impact inventory” as a means to standardize cost-effectiveness analysis to improve comparability between studies. A reference case is a set of recommendations covering components of analysis, methods, and reporting techniques. Two reference cases should be done, one from the perspective of the health system and one from that of society. From the health system perspective, the analysis should include an incremental cost-effectiveness ratio and the net benefit (health or monetary). The societal reference case should include a complete “impact inventory,” a structured chart of effects and costs, with components including patient time, caregiver time, transportation, and effect on productivity. All analyses should give clear explanations regarding the perspective being adopted.
With growing evidence and cost pressure, many more studies of the costs and benefits of perioperative interventions have been done to guide physician behavior.
Barriers to Doing the Right Thing
But even if we were able to know exactly what the costs and benefits of various perioperative interventions were, we would still find that implementation of known beneficial interventions lags far behind our awareness of the effectiveness of the interventions.
There are many possible reasons for the gap between awareness of beneficial interventions and their implementation. These include:
Lack of awareness: The physician does not know about the new intervention.
Lack of familiarity: The physician knows the intervention exists but not the details.
Lack of agreement: The physician does not agree with the proposed intervention.
Lack of self-efficacy: The physician does not think they can perform the intervention.
Lack of outcome expectancy: The physician does not think the intervention will work.
Inertia: The physician does not want to change.
External barriers: The physician wants to change but cumbersome and inefficient work processes make it difficult to change.
It is futile to simply exhort physicians to work harder and to perform better. Caregivers need work process redesign as well as concrete positive or negative incentives for changing their behavior.
The goal then is to build compliance with best practices into the system, so that the best care is the default option. For example, best practice for tidal volumes has changed over time toward a “lung protective ventilation strategy” with smaller tidal volumes and positive end-expiratory pressure (PEEP). Changing the default setting on the anesthesia machine is an effective way to change the tidal volumes and PEEP received by patients. The default setting facilitates a change in practice of physicians without adding additional hassle. Another example is that placing hand sanitizer in more convenient and visible locations improves caregiver compliance with hand hygiene, presumably as a result of increased ease of performing hand hygiene.
Costs—What Exactly Do We Mean?
Costs can be analyzed from different points of view—that of the patient, the provider (physician, hospital, clinic), the payer, or society as a whole. For example, the cost of a medical service to the payer (e.g., an insurance company) equals the percentage of charges actually paid by the payer. However, to the patient, the relevant cost is the out-of-pocket expense (that portion not covered by insurance) plus other indirect costs (e.g., inability to work) incurred as a result of the illness. From society’s point of view, the cost of a medical service is the total cost of all the different components of providing the service, plus the costs of any future consequences of that service, such as complications or disability.
To maximize the usefulness of an analysis to different audiences, the perspective of the analysis needs to be clear and hopefully relevant to the intended audience. For example, pharmacists may be more focused on the hospital or provider perspective. The physician should give greatest weight to the patient’s perspective, whereas the health economist is likely to focus on the analysis from society’s perspective.
Aside from the difficulties presented by differing points of view, the term costs has many different meanings.
Direct costs are the value of resources used to prevent, detect, and treat a health impairment. The adjective direct indicates that there is a clear matching of the expenditure with a patient. For example, an antibiotic administered to a patient is a direct medical cost because the antibiotic can easily be attributed to the particular patient who received the medication.
In economic analyses, direct costs should be estimated on a net basis—that is, they are calculated as the cost of the intervention minus any savings in future medical costs. For example, if an antibiotic can be shown to prevent future wound infections, then the cost of the antibiotic for purposes of a cost-utility analysis should be reduced by probable future savings of medical costs. As another example, the full scope of costs associated with intravenous (IV) patient-controlled analgesia include nurse and pharmacy labor, pump and disposables, intangible costs (e.g., adverse events from programming errors), and potential events such as analgesic gaps from malfunctioning pumps or IV line failures.
Perioperative clinics reduce laboratory testing, specialty consultations, and canceled surgeries. For perioperative clinic cost-utility analysis, the cost of operating the clinic should be reduced by the future savings (e.g., reduced laboratory testing). Telephone preoperative clinics have an additional cost benefit to the patient by decreasing transport time, caregiver time, and patient time while maintaining the benefits of an in-person perioperative clinic. The net direct cost of the clinic should be decreased by both the savings of the perioperative clinic discussed previously and the additional savings to the patient and their family. A conceptual, financial, and political problem with this net-cost approach is that it may be difficult within an institution to spend money in one area in order to save money in another area. Proponents of preoperative clinics (and of other efforts to redesign patient care) need to point out to administration that investing in a preoperative clinic may produce large financial returns elsewhere in the hospital, which may more than pay for the expense of running the clinic. The idea of investing in the preoperative clinic is being expanded to include the perioperative setting with a growing trend toward perioperative surgical homes as a method to reduce costs and improve outcomes.
Indirect costs are the value of production lost to society as a result of a patient’s absence from work, disability, or death. Because indirect costs are “opportunity” costs and do not directly influence expenditures for treating disease, they are not easily measurable. According to the human–capital approach, indirect costs are estimated as the income lost while the patient is absent from work.
Intangible costs represent another category of costs that—like indirect costs—are difficult to measure. These are the costs of pain, suffering, grief, and other nonfinancial outcomes of disease and medical care. They are not usually included in economic evaluations but are captured indirectly through quality-of-life scales. For example, epidural analgesia is more expensive than intravenous analgesia but provides better relief of labor pain. The additional expected cost to society of epidural analgesia for labor pain ranges from $259 to $338 per patient (depending on whether nursing costs are assumed to increase as the number of epidurals increases). Patients, physicians, and society need to weigh the intangible value of improved pain relief from epidural analgesia versus the increased cost. For the treatment of pain in colorectal surgery patients, an algorithm has been proposed.
Costs versus Charges
Costs are not the same as charges. For example, a hospital’s cost for giving a medication is usually interpreted to equal the acquisition cost of the medication, plus the true cost of delivering it to the patient. In contrast, charge refers to the amount of money the doctor or the facility bills the insurance company for the medication. Charges often bear little or no relation to acquisition cost, as anyone can attest who has heard about $10 aspirins and $40 plastic bedpans.
To estimate costs, either a top-down or a bottom-up approach can be used. One top-down method of estimating costs uses cost-to-charge ratios, such as those that all US hospitals supply to Medicare. These ratios are used to convert hospital billing (charge) data to estimated costs. The biggest advantage of the cost-to-charge ratio method is that charge data are commonly available and their use is well accepted.
The biggest disadvantage of the cost-to-charge ratio method is that charge data may not reflect the true cost to the facility of providing care, particularly when hospitals markup charges for services in one area to invest in poorly reimbursed departments (e.g., medical records) or to pay for the development of new clinical programs. This cost-shifting is common. For example, areas of the hospital with low cost-to-charge ratios are used to subsidize areas that have high costs in relation to their charges.
Another top-down method of estimating costs uses Medicare diagnosis-related groups (DRGs) to classify episodes of care. Although simple and efficient, this method is limited because it does not account for variations of care within a DRG, or between hospitals of varying efficiency.
Bottom-up costing is an attempt to measure costs more precisely, because, in theory, resources are tracked as care occurs. Bottom-up costing separates costs into fixed and variable components. Fixed costs (e.g., rental of the building that houses the surgery suite) do not change in proportion to the volume of activity, whereas variable costs—such as the price of a disposable anesthesia circuit—are closely tied to the volume of production. The majority of the costs of providing hospital care are related to buildings, equipment, and salaried or full-time hourly labor, all of which are fixed over the short term.
A misleading overstatement of costs can occur when a portion of fixed costs is allocated to what appears to be a variable cost. Sometimes one hears operating room (OR) staff state, “Operating room time costs $40 per minute,” implying—incorrectly—that if a patient left the operating room 10 minutes sooner, there would be a $400 saving. The truth is that the mortgage, the administrators, and the nurses all have to be paid, whether the patient is physically in the OR or not.
Similar discussion frequently occurs regarding whether choice of a certain anesthetic (e.g., regional versus general) can increase patient throughput by decreasing time in either the operating room or the postanesthesia care unit (PACU). This would only decrease cost if a marginally decreased PACU census could be translated into decreased staffing levels.
Skewness in Cost Data
A distinctive feature of cost data in health-care is its asymmetrical distribution (skewed to the right) and large variance. The three measures of central tendency or “average” value of a distribution are mode, median, and mean. High-cost patients (the right-skew) increase the mean to a greater extent than the median. It is estimated that the top 5% of patients utilized 50% of health-care spending, while the lowest 50% only utilized 3% of health-care spending.
If information about the costs of alternative treatments is to be used to guide health-care policy decision making, then the total budget needed to treat patients with the disease is relevant. For example, health-care planners may need information about the total annual budget involved in providing a treatment at a particular hospital. An estimate of this total cost is obtained from data in a trial by multiplying the arithmetic mean (average) cost in a particular treatment group by the total number of patients to be treated. It is therefore the arithmetic mean that is the informative measure for cost data in pragmatic clinical trials.
Future Costs Need to Be Discounted to Present Values
A cost today is not equivalent to a cost in the future. Even when inflation is taken into account, a cost or an outcome today is not equivalent in value to the same cost or outcome in the future. Because people prefer to have something today instead of having it in the future, a future value must be discounted to the present, and this can range from 3% to 7% per year.
Good News about Costs: Marginal Costs and the Learning Curve
Marginal costs (the cost of the next unit of production) normally decline over time. Both individuals and institutions learn and get better at tasks over time. When a technique becomes routine, people develop ways of performing the technique easily. For example, an outside clinical laboratory establishes a blood-drawing station right next to the office of a busy surgeon, thereby saving patient time and travel.
Competition and the freedom to innovate produce process improvement:
Problem identification → Invention of a new treatment → Expensive and difficult implementation → Skill acquisition → Increased volume of treatment → Increased competition on the part of suppliers and caregivers → Drop in costs → A new standard of care is established at a lower marginal cost.
Because of this learning curve, truly beneficial interventions should be performed even if the initial cost seems high, because we can count on costs coming down over time as a result of increased volume, experience, and competition. When curare was proposed as a new intervention to prevent fractures during electroconvulsive therapy, it might have been rejected as too expensive. Fortunately, the innovation was made despite the high initial costs.
Benefits: Clinical Effectiveness and Utility
Three types of benefits are commonly studied in health-care economics:
Desirable clinical results (i.e., effect)
Increased utility (quality and duration of life)
Desirable clinical effects are desirable outcomes or results that can be measured such as:
Reductions in complication rates
Quicker discharge from the hospital
Decreased postoperative nausea and vomiting.
Increased Utility (Quality and Duration of Life)
Health is more than repairing injury, alleviating pain, and eliminating illness. As long ago as 1948, the World Health Organization expanded the boundaries of health to include complete physical, mental, and social well-being. We now consider the impact of a disease and its treatment on patients’ daily lives.
For cost-utility analyses, quality of life has to be quantified—that is, converted into units that can be compared among different conditions. Utilities are numerical ratings, or preference weights, of the desirability of health states that reflect a person’s preferences on a linear scale from 0.00 (death) to 1.00 (perfect health). Preference values for health states are commonly obtained using valuation techniques such as the standard gamble, the time trade-off, or the visual analog scale.
QALYs are now widely used in medical decision-making and health economics as a useful outcome measure that reflects both quality of life and duration of survival. This single-score summary measure is obtained by multiplying the utility value for a given health condition by its duration. For example, the QALY score for an individual in perfect health (with a utility of 1.0) for 1 year (QALY = 1) is considered equivalent to 2 years in a health state with a utility of 0.5 (QALY = 1).
The word benefit is used in a limited sense in the phrase “cost-benefit analysis” to mean an assigned monetary equivalent for a nonmonetary benefit.
Types of Economic Analysis
The three major types of economic analysis in health-care are cost-identification (or cost-minimization) analysis, cost-effectiveness analysis (which includes cost-utility analysis), and cost-“benefit” analysis.
Cost-Identification (or Cost-Minimization) Analysis
Cost-identification analysis simply asks: “What is the cost of a given intervention?” By calculating the cost of delivering a drug or by computing the total cost of the medical services used to treat a condition, the costs of alternative therapies can be compared. Cost-identification analysis is sometimes referred to as cost-minimization analysis, because it is often used to identify which of several therapies has the lowest cost.
Cost-identification analysis assumes that the outcomes of the therapies are equivalent, therefore the goal is to find the least expensive way of achieving a standard outcome. For example, one study found the pharmacy cost of delivering postoperative analgesia to patients undergoing joint replacement surgery represents approximately 3.3% of the total costs of surgery. No statement can be made from that study about how well the analgesia worked.
Cost-Effectiveness Analysis (Which Includes Cost-Utility Analysis)
Cost-effectiveness analyses are the most accepted economic evaluations in health-care because they measure benefits in patient-oriented terms (clinical effects or quality of life). They permit comparison between different interventions by standardizing the denominators (clinical effects or utility).
Cost-effectiveness analysis, in contrast to cost-identification analysis, incorporates both cost and effect. It measures the incremental net cost of performing an intervention (expenditures for the intervention minus savings in future health-care costs) and compares it with the marginal, or incremental, benefit obtained. The incremental cost-effectiveness ratio (ICER) is defined by the equation