Economics of Health Care in the Intensive Care Unit

One of the catalysts driving the growth of pharmacoeconomics is the staggering cost of health care. In 2009, healthcare spending in the United States increased to a total of $2.6 trillion, and it is projected to be 4.7 trillion in 2019, which would represent 19.3% of the U.S. Gross Domestic Product (GDP).⁷ Hospital sector spending in the United States increased 86% in 2009 to $761 billion.

U.S. prescription drug expenditures were $300 billion in 2009, representing a 5% increase from 2008.⁸ Data in nonfederal hospitals for 2008 reveal that drug expenditures increased by 2.1% to $27 billion, and injectable drugs accounted for 71% of these expenditures.⁸ In 2008, the top 10 therapeutic classes accounted for 73% of hospital drug expenditures, with antineoplastic agents being the highest expenditure at $3.3 billion. The top two individual drugs in 2008 were enoxaparin at $1.1 billion and immune globulin at $868 million.

The cost of drug therapy is complex and comprises multiple components. Table 190-1 summarizes the cost of the drug product versus the cost of complications.⁹ It is easy to obtain data on acquisition costs of drugs and materials to prepare and administer drugs. Determining the cost of drug failures or adverse drug events is far more challenging, however, and often is not considered. The estimated annual costs of drug-related problems in the United States increased from $77.6 billion in 1995 to $155 billion in 2000.¹⁰

TABLE 190-1 Cost of a Drug Versus Cost of a Complication

Intensive care units (ICUs) consume significant hospital resources. Quantifying costs attributable to ICU care is complex, since both patients and their costs are constantly being shifted to and from the ICU, and different hospitals use different cost accounting systems. However, the large economic burden of the ICU is out of proportion to the number of ICU beds in the institution. In 2005, 95,000 ICU beds accounted for only 7% of all inpatient beds in the United States yet consumed about 13% of inpatient costs, or $82 billion annually.¹¹ Furthermore, this figure represents 4% of national health expenditures. The cost of an ICU day is estimated to be three to four times the cost of a ward day and has increased 30% from 2000 to 2005.¹² One study reported that daily ICU costs averaged $3518 during 2000-2005,¹² but a large database of more than 50,000 patients from 252 ICUs revealed a mean ICU cost of $19,725.¹³ Daily ICU costs were greatest on day 1 (average $7728), decreased on day 2 (average $3872), and stabilized on day 3 and beyond at approximately $4200 in 2009 dollars. Mechanically ventilated patients had the highest ICU costs; use of a ventilator increased average daily cost by $1800 compared to ICU patients who were not receiving mechanical ventilation. Therapeutic interventions that can reduce ICU length of stay by even 1 day can have a significant impact on total hospital costs, particularly in patients requiring mechanical ventilation. Cost saving initiatives are particularly relevant in this setting, since only 83% of hospital costs are covered for Medicare patients with an ICU admission.¹⁴ This discrepancy can result in a $5.8 billion loss to hospitals when ICU care is required.

Drug costs in the ICU are difficult to quantify because most hospitals are not sufficiently computerized to track these data. In one academic medical center, 15 drugs accounted for more than 50% of drug costs in the ICU.¹⁵ Drug costs in the ICU averaged 38% of the hospital’s total drug costs and increased at a higher rate than non-ICU drug costs over the 4-year period studied (12.4% versus 5.9%). Fiscal year 2002 data revealed an ICU drug cost of $312 per day compared with $112 per day outside of the ICU.

Economic Evaluations in Critical Care Medicine

Although any economic analysis could be performed in a critical care environment, some of the more appropriate are cost-effectiveness analysis, cost-benefit analysis, cost minimization analysis, cost utility analysis, and cost of illness.¹⁶ Cost-effectiveness evaluation is discussed in greater depth because it is the most commonly performed and the approach recommended by expert bodies.¹⁷

Cost-effectiveness analysis is a full economic evaluation because both costs and outcomes are considered. A drug is evaluated on the basis of cost and outcome in reference to a comparator, which is usually the current standard of care. In a cost-effectiveness evaluation, the most preferred therapy has increased effectiveness at decreased cost.

In 1996, the Panel on Cost-Effectiveness in Health and Medicine (PCEHM) published seminal work consisting of guidelines for the conduct and reporting of economic analyses.¹⁸ This work resulted in some key points to consider when employing a cost-effectiveness evaluation, including use of a reference case for comparison, the importance of transparent methods and logic, and consideration of the perspective being evaluated. The PCEHM recommends that the societal perspective is the most comprehensive and considers workforce and familial aspects of illness. The PCEHM recommends using the following steps when designing a cost-effectiveness analysis:

Figure 190-1 Example of a decision tree.

The American Thoracic Society convened a workshop to address the application of the PCEHM guidelines to a critical care environment.¹⁹ A group of experts compiled key considerations for a cost-effectiveness evaluation in the ICU, and details can be reviewed in their report.

Cost utility analysis is a form of cost-effectiveness evaluation that examines the utility or value of an outcome.¹⁶ Patient, family member, provider, or societal preferences can value health outcomes. Cost per quality-adjusted life-year can be measured for alternative therapies by assessing the length of time a patient is in a state of health rated on a scale of 0 to 1, where 0 equals death, and 1 equals perfect health.

Cost minimization assumes equal effectiveness for each alternative and evaluates the impact on an identical outcome. For example, if two ICU sedatives produce the same quality of sedation, but one requires a more labor-intensive administration protocol, a decision maker could apply a cost minimization analysis to determine the preferred, less costly therapy.

Cost-benefit analysis compares the costs and benefits of alternatives. This approach is rarely used in medicine.

Cost-of-illness studies describe the economic burden of a specific condition or disease state and are frequently part of epidemiology studies. This type of analysis may take into consideration the workforce and societal impact of illness, in addition to the financial implications for payers and providers.

The American Recovery and Reinvestment Act of 2009 provided considerable funding for comparative effectiveness research and mandated that the Institute of Medicine (IOM) of the National Academy of Sciences recommend initial national priorities for this research.²⁰ Its purpose is to compare the ratio of cost to effectiveness of two interventions used for the same condition.¹⁶ These comparisons should be evaluated in the real-world setting and could provide data about which interventions are most effective for patients under specific circumstances.⁸ It has been estimated that these data could reduce spending by Medicare and Medicaid by $0.1 billion from 2008 to 2012 and $1.3 billion between 2008 and 2017.

Determining Costs in the Intensive Care Unit

Three approaches frequently are used to assess the economic burden of a disease state: prospective study design, retrospective analysis, and decision modeling.¹⁷ A prospective study gives the investigator an opportunity to measure important variables completely and accurately. Retrospective database analysis reviews data that already have been assembled and has the advantage of being much less costly and time consuming than prospective studies with the ability to review many patients easily.

In retrospective studies, the subjects already are assembled and have been de-identified, baseline measurements have been made, and the follow-up period has occurred. The total direct and indirect costs of a condition can be readily assessed. Total direct costs include the value of all goods, services, and other resources consumed in the provision of an intervention or in dealing with the side effects of the intervention or other current or future consequences linked to the intervention.²¹ Indirect costs are the costs that result from a certain therapy or illness, such as lost wages, workforce replacement, or child care that may be necessary.

Patient billing information and summary estimates of department-level expenditures can be used to estimate costs when hospital administrative data are used. At one extreme, hospital charges can be used as a proxy of costs. This approach may be reasonable in a comparative analysis of interventions, assuming that charges per admission are roughly proportional to economic costs per admission. Another approach is to use the department’s cost-to-charge ratio, which has been shown to perform accurately when evaluating average costs per diagnosis-related group.²²

Cost of Intensive Care Unit–Related Conditions

Evaluation of the economic impact of medical conditions is one of several areas of focus since the 1990s in ongoing efforts to decrease overall healthcare spending and identify high-cost diseases to target therapies. More recently, hospital-acquired conditions have received increased attention. Several of these are so-called never events that are not reimbursed by government payers.²³ Representative conditions of the 10 identified areas for 2009 and 2010 include deep venous thrombosis (DVT) or pulmonary embolism (PE) developing after total knee or hip replacements, extreme manifestations of poor glycemic control, and surgical site infections. Recently, a study analyzed the performance of several hundred hospitals for hospital-acquired conditions and their associated additional costs.²⁴ Top conditions and their associated annual costs were decubitus ulcers ($536,900), DVT and PE ($564,000), and infections ($252,600). Preventing these conditions from developing could result, for example, in a 200-bed hospital saving $2 million per year.

In this section, we will summarize the data on selected acute care conditions for which information on costs is known. Although the data presented may not be directly from patients in the ICU, the results represent the best data available.