How Much Do We Spend and How Can We Spend It Better?
Published Online: March 16, 2012
Kim L. Farina, PhD
Expenditures on healthcare in the United States account for approximately 18 percent of gross domestic product.1 In the absence of costcontrolling changes to the US healthcare system, expenditures are expected to increase even more. It is critically important that all stakeholders, that is, policy makers, payers, providers, and consumers, be represented in the ongoing debate about healthcare reform. A basic understanding of medical economics is necessary for anyone interested in evaluating proposals for reform. During a Medical Economics Clinical Science Forum held December 7 at the 2011 San Antonio Breast Cancer Symposium, Elena Elkin, PhD, Memorial Sloan-Kettering Cancer Center, and Michael Hassett, MD, MPH, Dana-Farber Cancer Institute, attempted to answer 6 basic questions, outlined below, about the cost of cancer care.
How much do we spend?
Between 1990 and 2009, estimated adjusted annual direct medical spending on cancer in the United States doubled.2 During 2010, the United States spent an estimated $125 billion on cancer care.3 Breast cancer spending accounted for 13% ($16.5 billion) of all direct medical spending on cancer in 2010 (Figure 1).3
How do we spend it?
Total average Medicare spending per patient for initial phase care of breast cancer (2 months prediagnosis–365 days postdiagnosis) was $21,000 (2002 US$) in 2002 (Figure 2).4 Surgery and radiation cost little on a per-patient basis: $5700 and $4500 (2002 US$), respectively, and were used in 91% and 51% of patients, respectively. In contrast, chemotherapy and other inpatient services were used in about 25% of patients, but at a higher perpatient cost ($12,800 [2002 US$]). If the data used for this analysis were expanded to include continuing care and end-of-life care, there would be a marked difference in spending patterns. The United States spent an estimated $62,900 to $94,300 per person for end-of-life breast cancer care during 2010 (Figure 1).3
How can we spend our money better?
“In order to make good decisions about investments in healthcare, we need information about value. I hope you would all agree with that statement,” Dr Elkin remarked. A cost-effectiveness analysis, also known as a cost-utility analysis, is one way to obtain information about value. Cost-effectiveness analyses estimate the value for money of different healthcare interventions through a comparison of incremental costs and incremental health benefits of a new intervention. They can inform decisions about individual healthcare choices, clinical policy, and societal resource allocation with respect to preventive, diagnostic, and therapeutic interventions.
Often, a cost-effectiveness analysis compares a new treatment with an older standard treatment. The end point derived from incremental cost-effectiveness analysis is the incremental cost-effectiveness ratio, or ICER, which can be expressed as cost per qualityadjusted life-year (QALY) (sidebar). Dr Elkin cautioned, “Sometimes in these analyses it is not so black and white. If a new treatment is compared to a very poor, older treatment, it will make the new treatment look better. That is why it is very important in an analysis to understand what they are comparing it to.”
The United States does not use costeffectiveness analysis in making decisions on healthcare expenditures or in setting priorities. As part of their decision- making process, several countries, including the United Kingdom, Canada, and Australia, require formal quantitative evaluation in the form of costeffectiveness analysis and consider the recommendations of a designated independent agency.6 Dr Elkin remarked, “I don’t think any of us would advocate that cost-effectiveness should be the sole basis for a decision, but I believe that more information can help us make better decisions.”
What is good value for money?
Willingness to pay for health gain really defines what is considered a good value for money. Available resources influence it and it varies across people and geographical regions. In the United States, there are no strict criteria for good value in healthcare.
Common perceptions suggest upper limits of good value ranging from $50,000 per QALY to $100,000 per QALY. These thresholds, however, are probably too low, are certainly out of date, and might be contributing to resistance to cost-effectiveness analysis in the United States.6 As a result, some researchers and stakeholders have discussed updated thresholds for willingness to pay.6 Adjusting the $50,000 per QALY value (a threshold first proposed in the early 1980s) to 2007 US$ equates to $197,000 per QALY. The World Health Organization’s suggested calculation sets the threshold at $140,100 per QALY in 2008 US$. To put things into context, Dr Elkin presented cost-effectiveness estimates for several breast cancer interventions, a sampling of which is shown in Figure 3.7
Dr Elkin finished with a few comments about cost-effectiveness versus savings. “Cost-effectiveness is not the same as cost saving. Most medical interventions do not save money; most result in net monetary expenditure….Incremental cost-effectiveness should never be the sole criterion for decision making.”
Do we spend too much?
The United States spends more on healthcare than other developed nations. 8 We know from national health expenditure data that annual per person healthcare spending relative to per capita gross domestic product more than doubled between 1990 and 2009; projections through 2019 see a continuation of this trend.9
Despite outspending by the United States compared with other developed nations, outcomes in terms of preventable deaths and years of life lost to malignant disease are not better and, in some cases, are significantly worse.10 Dr Hassett presented the results of an analysis that used Surveillance Epidemiology and End Results (SEER) and Medicare data to look at the relationship between spending in the year after diagnosis, quality of care, and outcomes. In this study, increased expenditures did not correlate with improved quality of care or 5-year overall survival.11
Can we control spending?
Government and employment-related coverage pay for the vast majority of US healthcare (Figure 4).12,13 As a major stakeholder in healthcare spending, the US government has made efforts to control cost; however, the US government does not have a fixed budget or take a unified approach to cost control. The Center for Medicare and Medicaid Services (CMS) sets Medicare payment rates, but there are legal restraints on what measures the agency can take to reduce cost.
The Affordable Care Act and elements of the Stimulus Package were passed by the US government to reduce spending. These pieces of legislation included provisions for:
• Health insurance subsidies
• Health insurance exchanges
• Independent payment advisory board
• “Cadillac plan” tax
• CMS contracts with Accountable Care Organizations
• Incentives for adoption and meaningful use of electronic health records
• Comparative-effectiveness research funding
History has demonstrated that the manner in which government applies its tools can affect not only spending but also utilization. For example, in 2005, changes to Medicare part B reimbursement for chemotherapy effectively reduced payment on many medications. Dr Hassett presented data demonstrating a dramatic drop in paclitaxel usage after the January 2005 payment change, which reduced nominal reimbursement per monthly dosage of paclitaxel by approximately 5-fold.14 Similarly, approximately 65% of US oncologists reduced their use of erythropoietin-stimulating agents (ESAs) after a national coverage determination by CMS regarding the use of ESAs among oncology patients.15
Dr Elkin touched upon this issue during a discussion of value-based insurance, which took place after the presentations. “One of the challenges is let’s increase reimbursement for something that is going to be very effective and low cost so that we don’t see major dropoffs in utilization of good drugs simply because [reimbursement] has dropped. Doctors [behave rationally], like the rest of us, [and] are going to respond to reimbursement incentives.”
Who should decide what we spend and how we spend it?
There are no simple answers to this question and it is really a societal question. Dr Hassett offered perspective by reviewing some of the pros and cons of government, payers, or providers being responsible for decision making. Cost decisions made by society (government) might involve a process that reflects the aggregate priorities of citizens and adopts a single set of standards. In this scenario, however, patients and healthcare providers may end up with less choice. If payers make decisions about cost, a broader array of options regarding the extent of coverage and cost might be available, but critics cite potential inequalities and high administrative costs as drawbacks. Healthcare providers often believe they have a duty to offer a patient any treatment that yields a net benefit regardless of the cost. If they were responsible for considering costs when making treatment decisions, they might be challenged to balance multiple, potentially conflicting responsibilities as taxpayers, business owners, and patient advocates.
PDF is available on the last page.