Economics of Cancer Care, 2019

Hearing the words “you have cancer” is extremely distressing and immediately raises a number of concerns about life expectancy, functional status, income and employment, family relationships, and the impact of treatments on each of these. Multiple surveys have documented the classic response mechanisms of denial, anger, bargaining, and sadness, but immediate practical considerations, such as impact on job, income, bills, and quality of life, dominate early conversations. Increasingly, another serious consideration is coming to the forefront—that of the costs of care, especially the self-borne ones. This tends to arise once the treatment plan is formulated, as the direct line of association crystallizes between the diagnosis and the plan of action. It is also highly dependent on the presence or absence of insurance benefits, the details of those benefits, and how well they are understood by the patient. A continuous and rapid rise over the last few years of these patient-borne costs, which are projected to persist and increase over an indeterminate period of time, not only threatens one’s ability to pay for treatment but also the access to such treatment. As the article by Paul points out, once the words “you have cancer” are spoken, everything that contributes to a delay in getting to a treatment plan, or at least information about the diagnosis, becomes high priority for most individuals. In this article, at least one-third to one-half of patients in Sydney Australia reported treatment-phase dependent concerns about access to treatment, especially those of lower socioeconomic status, those who were younger, and those born outside of Australia. The same concerns are true in the United States, especially in rural areas, for which insurance coverage is a contributing factor. In rural areas, Medicare is the predominant insurance for eligible individuals, and having no insurance has decreased slightly to approximately 15% to 25%. Initiation of the treatment itself, when disease is evaluable, provides a temporary emotional buffer to the costs of treatment, as it at least contributes subjective data about the imputed benefit-to-cost ratio. This is much more nebulous when the treatments are purposed to prevent recurrence, as the benefit/cost ratio is much more statistical in nature, rather than directly evaluable. Once treatment is concluded, as the proportion of survivors continues to increase (a positive outcome), long-term side effects of care will continue to escalate, leading to an increase in chronic noncancer but treatment-related financial and medical issues. ^, The Kline article proposed a risk-based model to optimize post-treatment care to the site most clinically appropriate for that care, and potentially delivered at a lower cost, such as the active incorporation of a primary care provider into the model. Future employment and insurability issues for those who have had cancer continue and are subject to local, regional, and federal regulatory changes. The looming threat of recurrence likely never disappears but certainly remains an emotional burden for a long time and is very disease dependent. Moreover, some treatments may actually predispose individuals to other cancers, further increasing long-term concerns. Adolescents who have survived cancer carry this burden for the rest of their lives. While this is shared with diseases such as diabetes, heart disease, chronic lung disease, and arthritis, the rapid increase in the costs associated with cancer has recently received much attention.

Magnitude of the Issue

In any discussion about costs, one must decide whether the focus is on total societal costs of care (usually difficult to measure), payer costs (usually easiest to measure as the payments to providers and applicable to those patients with some sort of coverage), provider costs (extremely variable, unique to the provider, and usually estimated from charges to reimbursement), patient-borne costs (relevant to insured and uninsured patients), and those costs not captured by classical reimbursement systems, such as travel expenses and out-of-pocket expenditures (very difficult to capture). Nonetheless, when reported, and regardless of the above variable definitions, all have shown a significant increase in cancer costs, especially over the last decade, from $124.57 billion in 2010 to between $186.59 billion and $206.59 billion estimated for 2020. These estimates are based on both incidence and survival trends. There is no evidence that the increase in costs will change in the near future, and most projections are that the rate of increase will also increase with the explosion of new technology and drugs. The cost associated with the initial year of treatment has grown by 19.4% in that time frame, while the cost in the last year of life has grown by 28.3%. The increasing role of continuing care (between the two endpoints) is shown by an increase of 31.6% over the same 10-year period. Not only is cancer being treated more effectively initially, it is increasingly a chronic disease with treatment extended over a longer period of time (similar to the diseases above). Furthermore, relapses that occur between the initial and last year of life have increasing options available to bring the disease back under control, contributing to longer survival after relapse. The average initial year of cancer treatment varies by disease from $5000 for most confined melanomas to $115,000 for brain tumors, ranging from a few hundred dollars a year to $11,000 for pancreatic cancer but these are averages and obviously very stage dependent. However, this group of costs is most likely to continue to increase as the periods of “maintenance” therapy expand for most cancers. The last year of life also varies by disease but ranges from $56,000 for melanoma to $135,000 for brain tumors. Much of this care is palliative in nature, some in intensive care units, and a majority in the hospital. Efforts to reduce these latter figures have thus far been unsuccessful, despite the increase in palliative care units and hospice locations. End-of-life care remains a considerable cultural issue and is relatively unique to the United States.

Given the lag in tabulation, most of the cited data are either 5–10 years old or rely on projections based on newer treatments. In a recent study released by the American Association of Cancer Research (AACR), projections are for cancer costs of approximately $450 billion by 2030, worldwide. Long-term data and even projections, by their very nature, are becoming subject to extreme variation and based on the utilization of current and newly released treatments, research developments and expectations of results, and population demographics. It is likely that even this projection will be refined long before that date to a much larger number. Medicare, Medicaid, and private payers provide some of the most recent expenditure data available and are generally searchable on the web pages of The National Cancer Institute and the American Cancer Society.

Medically Induced Bankruptcy

At an alarming rate, due to the cost of treatments and the distress caused by the rising cost of treatments, the number of people who file for a subsequent bankruptcy is increasing. ^, A recent article published in the American Journal of Medicine found that 42.5% of patients lost their entire life savings within 2 years of diagnosis. Total financial insolvency extended to 38.2% over 4 years. Fifty-seven percent of the patients in the study were Medicare and 34% had private insurance. Medicaid accounted for only 2% of the population, for whom the results would be even more dramatic. Cancer patients are 2.65 times more likely to file for bankruptcy than those without cancer. Estimating expected costs before the initiation of treatment remains an imprecise art, based on the experience of others or on the averages of a population, and subject to unexpected events during treatment. Historically, physicians have not felt comfortable about having financial discussions with patients, mostly in fear that the perception would be that doctors were basing recommendations on costs. In most locations, this role has been delegated to financial advisors after a specific treatment recommendation has been offered. This makes it difficult for financial advisors to discuss other options on a comparative basis. The American Society of Clinical Oncology has been wrestling with this issue for over a decade and, at an annual meeting in 2018, noted the continuing barriers to a difficult situation. Frequently, costs become an issue long after anything can be done to reduce them or inform decision-making.

To compound the issue, financial insolvency has been shown to be a risk predictor for early mortality among patients with cancer according to Ramsey, with a hazard ratio of 1.79. The endless circle of high costs, insolvency, inability to continue care, and worse outcomes is an unfortunate reality and story repeated over and over. It has become an all-to-frequent newspaper, magazine, or television article.

Besides the sheer cost of treatments, this is also a reflection of changing structures in the way cancer care is paid for. Cancer is not specifically treated any differently than other major medical illnesses; however, due to its escalating costs, it has drawn the attention of preexisting condition legislation, targeted preauthorization requirements, and coverage policies for newer treatments. Changing insurance plans with a history of cancer treatment or current treatment is difficult. With the very conscious effort to shift costs to the beneficiaries, regardless of plan, patients with cancer tend to be hit by coinsurance and out-of-pocket requirements very rapidly. As alluring as high-deductible plans can be with a lower premium, it is not always clear to purchasing patients how this affects cancer treatment access and costs. Delays in access and treatment have been reported in patients with breast cancer. Cancer-specific insurance policies do exist but frequently have significant limitations and exclusions. For those with Medicare, it is critical to have a supplemental plan, as out-of-pocket exposure can be unlimited.

Financial Distress and Toxicity

Because of the above and across the board regardless of insured status or resources, the emotional and psychologic aspects of financial distress are increasing. In that article they cite the effect on all age groups: young adults with generally fewer assets and savings are dramatically affected for life. Blending into the middle adult age group, there is a need for loans, debt, and an inability to pay for care. Minorities were especially affected, depending on their socioeconomic status. The intensity of treatment, which also increases in duration, is a major driver of certain diseases. Newer molecular therapies are increasingly expensive and a major contributor. There are also data that it is either delaying care or discouraging care. It has also been shown to correlate with increased pain, greater symptom burden, and poorer quality of life in a study by Lathan. Methods of measurement for financial distress are evolving and will hopefully be applied as a routine part of assessment at more institutions, just as patient experience of care and patient-reported outcomes are currently. ^,

The Ultimate Solution to Reduce Costs

Of course, to the extent it applies, which is not known to be true for all cancers, prevention and early detection are the optimal solutions. Risk factors for the most common diseases have been known for a long time, and while some risk reduction efforts have been successful, there is still much room for improvement. Many revolve around lifestyle changes, which have been adopted, would reduce cancer incidence. They are (1) a reduction in tobacco use, (2) adoption of a healthy diet, (3) maintenance of healthy weight, (4) physical activity, (5) protection from excessive sun exposure, (6) limitations of alcohol exposure, (7) vaccinations for relevant viruses, (8) regular medical care, and (9) avoiding risky behaviors. Estimates by the Harvard School of Public Health are that adoption of these could reduce preventable cancer deaths by up to 75%. In the case of smoking, a 50% reduction in smoking (for smokers of more than 15 cigarettes per day) reduced the hazard ratio for future lung cancer to 0.73. In addition, regular medical care with appropriate screening for selected cancers has been effective in detecting disease in earlier stages and has been shown to reduce the costs associated with more advanced disease. Depending on the criteria and population chosen for analysis, the results vary when applied to many cancers. Many of the comorbidities that are present in people who are cured of their cancer would also be reduced by such activities, such as smoking reduction. It is recognized that we do not know how to prevent all cancers from developing, but the effects on the most preventable ones, colorectal, lung, cervix, pancreas, bladder, breast, ovarian, melanoma, stomach, kidney, hepatocellular, uterine, and prostate, as well as the genetically linked inheritable ones, happen to overlap greatly with the most common cancers and therefore have the biggest effect.

Health Plan Efforts in Wellness

Unfortunately, benefits programs in virtually all plans have spent more dollars in treating illness than preventing illness. This is beginning to change for the first time since the implementation of health maintenance organizations (HMOs), with some benefits now being provided for free or at markedly reduced costs, such as screenings and vaccinations. Health coaches and wellness programs are becoming a common feature of plan offerings. A recent survey found that 86% of employers use their wellness programs.

Strategies in Benefit Design to Lower Costs

Cost-sharing strategies take different forms, but the most common are increases in copays and coinsurance and tiering of pharmaceutical and imaging benefits. Structural strategies, including out-of-network obligations, narrow networks, annual or lifetime maximum benefit caps, and utilization management programs with financial implications for “nonpathway” treatments have been implemented by insurance plans and self-insured employers.

Changes in copay, coinsurance, and deductible requirements play a key role in the design of most plans. In fact, the increases in cost-sharing payments exceed wage growth over a 10-year period, thus becoming less affordable for the average patient with cancer. In the study by Claxton et al. the average deductibles increased from $303 to more than $1200 between 2006 and 2016. Unfortunately, this contributes to prescription abandonment and delaying refills in order to postpone costs. In the study by Doshi et al. the abandonment rate was 18% overall and highly correlated with the out-of-pocket cost, with a rate of 49.4% for out-of-pocket costs greater than $2000. Delayed initiation was prolonged in this group.

Pharmaceutical benefit tiering is another tactic for increasing sharing costs, particularly relevant to cancer. The article by Dubanski discusses this issue with regard to oral anticancer drugs and Medicare Part D, noting that a 25%–33% coinsurance requirement is in place for drugs in the highest tier, which includes many cancer drugs. Almost all private insurance plans have similar structures. This trend has not abated and is not likely to do so in the current benefit structures.

As cancer costs increase, it is critical to know whether the provider is in-network or out-of-network. The difference in plan responsibility can vary from 0% (100% for you) to 100% (0% for you) of a very large number. As the 1-year costs of treatment (extremely disease- and stage-dependent) can vary from approximately $10,000 to well over $500,000, the implications are obvious. Unfortunately, not all geographic areas of the country have equal access to “high-quality” oncology providers, although most cancer care is available at a reasonable distance. Legislative efforts have been directed at ensuring access to high-quality care for all, but this is still not reality for most people with cancer. Geographic location has been shown to be one of the most important factors in cancer mortality, partly due to access and partly due to incidence. ^, Although health plans are required to have specialty coverage in the network as required by state laws, there continue to be opportunities for improvement.

Alternate Payment Models

In an attempt to deal with the rising costs, alternative payment models may be tested with various levels of one-side or two-side risk. Such programs include the oncology care model, centered on chemotherapy, a proposed radiation oncology model, and many examples of various bundled payment programs in both the private and public sectors. In most of these, payment is also tied to quality outcomes and the achievement of system efficiency. It is still too early to assess whether these types of programs will evolve into a primary payment model for cancer.

The Uninsured

Those without insurance, especially if not well resourced, avoid these issues but are faced with the cost of care directly and immediately, unless public funding is available. Frequently, it forces one immediately to ask questions that nobody should have to ask: Is it worth it? Or, should I or my family go bankrupt? Financial distress has risen to the top of the chief concerns of people with few resources who are diagnosed with cancer. Unfortunately, there appears to be no immediate solution on the horizon to alleviate these concerns or questions.