Author + information
- Edward P. Havranek, MD⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Edward P. Havranek, Denver Health Medical Center, 777 Bannock Street #0960, Denver, Colorado 80204-4507
“The other America, the America of poverty, is hidden today in a way that it never was before. Its millions are socially invisible to the rest of us … .”
The Other America: Poverty in the United States (1)
It has been a full 50 years since the publication of Michael Harrington's The Other America in 1962. Written near the height of America's post-war prosperity, the book exposed high rates of endemic poverty, caught the attention of the Kennedy and Johnson administrations, and spurred the development of “safety-net” programs such as Medicare and Medicaid. While these programs were successful in reducing endemic poverty, their success was incomplete and temporary. The nation's poverty rate stood at nearly 20% when the book was published and was nearly halved to 11% by 1974. Since 2000, however, it has risen steadily and now stands at over 15% (2). When we define the poverty threshold as less than one-half of median income—to account for the shape of the income distribution—the proportion of people in our country with incomes below the poverty threshold is now higher than in Turkey, South Korea, or Poland (3).
This rise in poverty makes our jobs as cardiovascular healthcare providers more difficult in 2 ways. First, poverty increases the risk for cardiovascular morbidity and mortality (4,5) and thus increases the number of patients requiring treatment. Second, it makes the appropriate treatment of cardiovascular disease more difficult because rising poverty increases the likelihood that patients will be unable to pay for care. The strong link between poverty and a lack of health insurance in the United States is well known. In 2011, 25.4% of those with a household income <$25,000/year were uninsured; for those with a household income >$75,000/year the rate was 7.8%.
In this issue of the Journal, Smolderen et al. (5) present a rather remarkable study of the impact that an inability to pay for care has on both physicians and patients. They examined the relationship between insurance status and quality of care for chronic coronary artery disease using data from a voluntary quality-improvement registry sponsored by the American College of Cardiology. They studied over 60,000 patients in 30 practices who were under 65 years of age and who had prior myocardial infarction or coronary revascularization. The practices were distributed across the United States from east to west and north to south, were in rural and in urban settings, and were in counties of varying median income levels. Compared with patients covered by private or government insurance, the uninsured were less likely to receive prescriptions for beta-blockers (81% vs. 73%), angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers (ACE-I/ARB) (76% vs. 67%), and lipid-lowering medications (95% vs. 89%) in situations in which prescription would have been guideline-indicated. Thankfully, rates of prescribing antiplatelet therapy were not affected. Practices seeing a greater proportion of uninsured patients had slightly poorer overall quality performance, and adjustment for the proportion of patients in the practice who were uninsured attenuated the differences in beta-blocker, ACE-I/ARB, and lipid-lowering drug prescriptions.
The first of several important findings was that providing care for patients with a restricted ability to pay for that care is a common feature of contemporary cardiology practice. Overall, 9.4% of patients in the study sample were uninsured, and in several of the practices, uninsured patients made up over 20% of the sample. The financial impact on these practices is not known but cannot be negligible. In addition, the impact may go beyond the financial. Even when visit fees are waived or paid out of pocket, physicians and uninsured patients are faced with difficult decisions that insured patients are not: Should diagnostic tests be skipped and treatments be empiric?; Should less expensive medical options be pursued when interventional options might be superior? The toll that providing less-than-optimal care takes is also not known but also cannot be negligible. Offering care that we know is less than optimal is corrosive to our core professional values.
The second important finding in this report is that practices with higher proportions of uninsured patients were slightly less likely to prescribe guideline-based medications. The reasons for this finding are not clear, but 1 possible explanation is financial. Practices with higher proportions of uninsured patients may be put under greater financial strain, which diminishes their ability to put in place quality-monitoring and -improvement programs. As the authors suggest, directing assistance toward practices disproportionately serving the uninsured may help. This assistance will likely need to be both specific, focusing on adherence to quality measures, and general, providing for the development of infrastructure such as information technology. One example of ways this might be done is in the Medicare Shared Savings Program's Advanced Payment Model (7). This model allows low-volume rural hospitals to receive payments in order to set up accountable care organization infrastructure, and then to repay the advances out of savings realized for Medicare in the future. Many readers will notice that this way of thinking is nearly opposite of the “pay-for-performance” model that has been touted in recent years.
The third important finding is perhaps the most significant: It is prescription rates, not medication use rates, that are low. Had the investigators found low rates of medications use, this could have been explained by patients' inability to afford the medication. Instead, physicians did not prescribe the medication. With the widespread availability of inexpensive generic alternatives for beta-blockers, ACE-I/ARB agents, and lipid-lowering drugs, it should be uncommon for physicians not to offer prescriptions for these medications. It is possible that, over time, physicians learn which patients skip medication and therefore no longer write prescriptions. However, more subtle factors of which patients and providers may be unaware might also be involved.
One possible reason for physicians is that we are influenced by a flawed conception of medication adherence. As Steiner (8) pointed out, despite the fact that sociodemographic and clinical characteristics are quite weak predictors of adherence, clinicians might be using these characteristics to make decisions about withholding therapy from patients who they believe will be not be adherent.
Closely related is the possibility that providers' implicit biases are playing a role. Implicit biases are subconscious and can influence behavior without an individual being aware of the influence. Implicit bias can, and frequently does, exist in the absence of overt explicit bias. Several studies have suggested that implicit racial bias might influence medical decisions (9,10). Whether such biases can be activated by perceived socioeconomic status has not been studied, so this mechanism should be viewed as speculative.
The implication of these findings is that improved access alone might not improve care for those currently uninsured. Strained practices disproportionately serving uninsured patients may be put under greater strain by an influx of new patients and might have difficulty improving infrastructure for quality monitoring and improvement. Merely providing insurance might not change providers' assumptions about and behavior toward those recently uninsured.
In the end, I agree with the authors' conclusion that efforts to expand insurance access are a necessary part of the journey toward higher-quality care. The past decade's rise in poverty and decline in access to health care both seem likely to begin a reversal in the next few years. Indeed, in concert with the onset of some provisions of the Affordable Care Act, the proportion of people without health insurance decreased from 16.3% in 2010 to 15.7% in 2011 despite the fact there was no change in the proportion of those with employment-based insurance (2). We will need to consider that greater access alone might not improve the quality of care; we will need to strengthen infrastructure and change practice patterns in ways that improve the care of those currently on the edges of the system. Trying to treat insured and uninsured patients in one healthcare system has had unseen consequences for physicians and patients. Fixing this flawed system, however, will benefit both.
This paper was supported in part by grant no. NHLBIR01 HL88198.
↵⁎ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
- American College of Cardiology Foundation
- Harrington M.
- DeNavas-Walt C.,
- Proctor B.D.,
- Smith J.C.
- Organisation for Economic Cooperation and Development
- Woodward M.,
- Brindle P.,
- Tunstall-Pedoe H.,
- ASSIGN Group on Risk Estimation
- Smolderen K.G.,
- Spertus J.A.,
- Tang F.,
- et al.
- Centers for Medicare and Medicaid Services Center for Medicare and Medicaid Innovation