In this issue of Arthritis Care & Research, Wolfe and Michaud present an extensive analysis of the out-of-pocket expenses incurred by patients with rheumatoid arthritis (RA) and discuss the financial burden imposed by those expenses (1). Since the amount of out-of-pocket expenses faced by patients is the direct consequence of their health insurance status and coverage type, one needs to first understand the role of health insurance, how it works, and how a patient's ability to choose their desired type of insurance coverage is based on their income and risk preferences, as well as their subjective perceptions of financial burden when faced with illness-related treatment costs.
On a day-to-day basis, healthy individuals do not require any health services. However, illness can strike at any time and depending on severity, the associated health expenditures can be high or low. The resulting uncertainty due to the stochastic nature of health care needs, coupled with the fact that individuals with average earnings cannot cover the treatment costs of catastrophic illness, provides the rationale for health insurance. By providing a hedge against uncertain future costs, insurance also reduces the level of individual savings that would be needed, thereby increasing expenditures in other areas of the economy. Health insurance can take many forms. It can be mandatory or voluntary, privately or publicly provided, and its coverage may involve copayments, deductibles, or be first-dollar coverage. Two concerns related to the provision of health insurance are known as adverse selection and moral hazard.
Adverse selection occurs when groups of people choose not to buy health insurance, or they buy insurance coverage that is inappropriate for their risk class. At its worst, adverse selection makes health insurance unsustainable because of escalating costs, but most often it leads to incomplete coverage of the population. For example, in most current developed economies, even people without health insurance can still receive care since hospitals and health care providers are required to provide medical care regardless of whether a patient has health insurance or the ability to pay. In the US, therefore, there are many uninsured people who benefit from free or subsidized treatment. In other resource-poor settings, the consequences may be that the health care system cannot help those with low income get treatment. A mandatory insurance system is often seen as the solution. In addition, a mandatory insurance system requiring open enrollment and insurance premiums independent of health status, universal coverage, and taxation based redistributes resources in favor of those with a higher risk of illness by nature and therefore ensures equality of opportunities with regard to health (2).
Moral hazard refers to the phenomenon whereby individuals who are insured for their illness risks behave differently than they would without the insurance, i.e., if they were fully exposed to the risk. For example, an insured individual will do less by way of preventative health behaviors (ex ante moral hazard). In the context of health care, this problem is exacerbated due to the principal-agent problem, whereby the type of health care service utilized because of illness is determined by a health professional instead of the patient. Depending on how it affects their earnings, the provider may choose a treatment that is more expensive than warranted. Therefore, full insurance coverage for almost all types of medical care is not considered optimal because of the possibility of moral hazard. Full insurance coverage might undermine the individual's incentive to prevent illness, and therefore it fails to combat ex ante moral hazard. Furthermore, people under full insurance coverage are more likely to increase their demand for (utilization of) medical care, which gives rise to ex post moral hazard.
A solution to combat moral hazard is cost sharing between patients and insurers, and even between providers and insurers. From a patient perspective, there are 3 common types of cost sharing: premium cost sharing, deductibles, and copayments (3, 4). Under premium cost sharing, the insured is required to contribute to part of the insurance premium cost. A deductible is the amount that the insured is required to pay before any covered medical expenses are compensated by the insurance plan. A copayment is the amount that the insured has to pay out-of-pocket for their medical care after paying the deductible amount. For example, under a plan requiring a $250 deductible with an 80:20 coinsurance, once the cumulative medical expenses exceed the deductible, the insurer starts paying 80% of all covered medical care expenses, and the insured has to pay the remaining 20%. In addition to shifting expenditures from the insurers (often governments) to the insured, cost sharing is appealing because it supposedly mitigates moral hazard by making patients pay for a portion of all expenditures and this makes them more cost-conscious. There is limited empirical evidence on the impact of cost sharing on ex ante moral hazard. Zweifel and Manning (5) reviewed the evidence and summarized that as the out-of-pocket payment for insurance increases, the demand for preventive care services declines. Also, Zweifel and Manning reviewed empirical evidence on the impact of cost sharing on ex post moral hazard, which comes from natural experiments, observational comparisons, and a randomized trial of insurance (i.e., the RAND health insurance experiment). There is substantial and supporting evidence in the literature that cost sharing reduces ex post moral hazard, i.e., when the level of cost sharing by the insured increases, the utilization and expenditures of health care decline.
It is also well known that cost sharing, when applied to a specific type of health service, not only reduces expenditures on this specific service but also impacts expenditures on other types of health care services. In economic terminology, goods or services are considered to be “complementary” when the increase or decrease in the price of one of the goods or services leads to a corresponding decrease or increase in the consumption of the other. In contrast, goods and services are defined as economic “substitutes” when the quantity of demand for one rises or falls if the price of the other rises or falls. In a managed-care setting, studies have shown that cost sharing was associated with reduced expenditures for prescription drugs, physician services, and outpatient hospital services (6–9). The results of these studies suggested that prescription drugs and other health services were “complementary” goods.
However, contrary to most of the literature, we found that in the Canadian context the demand by RA patients for health services such as physician visits is positively correlated with drug price, and therefore prescription drugs and other health care services are economic “substitutes” (10, 11). We also found that the increases in the prices of prescription drugs inherent in copayments indeed decreased the demand for drugs, which is consistent with the literature. The possible explanations for why our findings contradicted other studies with respect to complementarity/substitutability between prescription drugs and other health care services might be as follows. First, the increase in cost sharing of prescription drugs may worsen senior adults' symptoms due to fewer medications taken, and therefore may lead them to see physicians or be admitted to the hospital more frequently. Second, in Canada, all treatments and health services (physician visits, emergency department visits, hospital admissions, and drug therapies provided in the hospital) except prescription drugs are covered by the government and are completely free of charge. Therefore, when prescription drugs become more expensive due to increasing copayments, patients are more likely to seek substitute treatment modalities.
In general, because of the complexity of insurance markets due to the presence of uncertainty, asymmetric information, and the principal-agent problem leading to adverse selection and moral hazard, empirical studies are very difficult, primarily because it is almost impossible to establish causality. Nonetheless, the study by Wolfe and Michaud provides a wealth of information that can now be considered within the context of health insurance and its known consequences (1). While the authors found that average out-of-pocket expenses for RA patients were high ($1,798), they had not increased substantially over the past decade when compared with the US average. However, their finding that 3.4% of RA patients <65 years of age have no health insurance suggests that adverse selection is a problem, and the solution lies in examining how the existing menu of insurance options, voluntary or mandatory, causes this group to prefer the no insurance option. Wolfe and Michaud's finding regarding the uptake rates for biologic therapy hint at moral hazard because they do not lead to an increased out-of-pocket burden.
Although cost sharing reduces moral hazard, a negative consequence is that it may exacerbate inequity in health care finance, which can be interpreted in terms of “vertical equity (persons or family of unequal ability to pay making appropriately dissimilar payments for health care) and horizontal equity (persons or families of the same ability to pay making the same contribution)” (12). Wolfe and Michaud focus specifically on vertical equity related to out-of-pocket expenses by asking the subjects to rate their burden of out-of-pocket expenses according to the following categories: no problem or limited problem (I am able to pay the bills without much problem); a moderate problem (paying the bills takes away some money I need for other activities); or a great problem (I can't purchase all of the medications or medical care that I need). The study found that the stated burden of out-of-pocket expenses varied with household income, RA severity, and the type of health insurance.
A limitation of this approach is due to the circular nature of the insurance puzzle. While the real burden of out-of-pocket expenses should be according to the actual level of expenses incurred, Wolfe and Michaud introduce the concept of “affordability” based on a patient's subjective evaluation of these expenses, which in periods of illness will depend on the decrease in their disposable income (income after out-of-pocket expenses), which in turn depends on their insurance coverage. Furthermore, the available data allow Wolfe and Michaud to control for insurance type only, which is an imperfect proxy for the actual costs of treatment covered by the insurer in the event of illness. However, when a patient ex ante chooses a specific type of insurance coverage, their subjective forecast of the ex post burden of potential out-of-pocket expenses will influence their choice of insurance type, more specifically the level of out-of-pocket expense they are willing to accept and not vice versa. In other words, patients with the same household income and the same amount of out-of-pocket expenses may rate “affordability” differently according to their attitudes toward risk.
The impact of the circularity limitation is somewhat mitigated because Wolfe and Michaud tested the consistency between the ratings and actual expenses by showing that the proportion of household income that was consumed by out-of-pocket spending was 2.4%, 7.2%, and 19.2% for the no burden, moderate burden, and great burden groups, respectively. In addition, as the authors pointed out, the high missing rate for out-of-pocket data (>40%) and the difficulty in testing the validity and reliability of patients' reports might discourage the use of the actual out-of-pocket expenses as the indicator of burden. The study results showed that the self-rated burden of out-of-pocket expenses falls continuously as income rises, implying that the American financing system is regressive in terms of vertical equity. Despite the limitations of out-of-pocket data, it is still worthwhile in further studies to investigate whether the finance of health care is progressive, regressive, or proportional by plotting the proportion of household income that was consumed by out-of-pocket spending and the different household income categories.