Man…institutionalized racism is expensive (2)

The World Bank puts India's GDP at $1.22 trillion dollars. U.S. health disparities cost $1.24 trillion between 2003 and 2006.

The Economic Burden of Health Inequalities in the United States (LaVeist T.A., Gaskin D.J., & Richard P., 2009: Joint Center for Political and Economic Studies)

Aims

a) Estimate the direct medical costs and indirect costs to the economy of health inequalities, which relates to b);

b) Estimate the potential financial benefit that would accrue to the economy if every racial/ethnic group in the United States had similar health outcomes.

Methods

Using the same data as the Urban Institute’s study, the Joint Center researchers took a different–and, dare I say, more sophisticated–approach using 2003-2006 MEPS data.

Direct medical costs

To predict adult health expenditures, the researchers also created a regression model including a variety of demographic, socioeconomic, location, and health status variables. On top of the health status and the chronic diseases examined by Waidmann, the researchers included self-reported mental health status and a few other factors you can read about in the report. Two separate models–one for each of the aims described above–were calculated for each year between 2003 and 2006.

To figure out the cost estimates after eliminating disparities, the Joint Center researchers don’t use non-Hispanic whites as the default comparison, and instead determine the best health profile for each health status/condition measure (in most cases, Asians had the best health profile) and simulates health care expenditures when, essentially,  everyone has the best health profile they could possibly have (i.e., a health profile based on that of that of the racial group with the best health profile for a particular health problem).

Indirect medical costs

To determine the costs of health disparities to society, the researchers used demographic, socioeconomic, geographic (census region and urban-rural residence), and health status measures to model days of work lost because of disability . With a little econometric magic to account for self-selection, and after again simulating health expenditures when whites, blacks, Hispanics, and Asians have the best health profile possible for each health status and condition, the researchers are able to compare predictions for the original sample (in which health disparities exist) to the simulated sample (in which we acheive miraculous equity).

The cost of premature death

Yeah, Urban Institute, what about premature death?

The Join Center researchers use National Vital Statistics Reports to cull the number of deaths and the crude death rate by age and race, and then estimate excess deaths related to health disparities. Actual deaths were then compared to the predicted number of deaths if the mortality rate equaled that of Asians–typically the group with the lowest mortality rates by age. In calculating the value of years of life lost, each year was valued at $50,000 (though they note that recent studies have valued “quality-adjusted life years” or QALYs at $95,000 to $264,000).

Findings

Eliminating health disparities for minorities would have reduced direct medical care expenditures by $229.4 billion for the years 2003-2006. More than 59% of these excess expenditures were attributable to African Americans, who have the worst health profile among the racial/ethnic groups.

More than 30 percent of direct medical costs faced by African Americans, Hispanics, and Asian Americans were excess costs due to health inequities – more than $230 billion over a four year period. And when you add the indirect costs of these inequities over the same period, the tab comes to $1.24 trillion.

About 95% of the indirect costs of health inequalities ($957.5 billion) were due to the costs of premature deaths, while the remaining $50.3 billion resulted from illness. African Americans accounted for $782.8 billion, or more than 77% of all indirect costs attributable to health inequalities.

Discussion

Holy crap. U.S. health care disparities between 2003-2006 cost the equivalent of India’s GDP in 2008.  Clearly, bridging health disparities is not just a noble side-effect of health reform, but a sensible driver. As Thomas A. LaVeist, director of the Hopkins Center for Health Disparities Solutions and the report’s author notes,

“What we are arguing in this report is if you want to get a handle on health care costs and quality and ensuring that the U.S. has a healthcare system that is worthy of this nation, you need to account for disparities. And the economic effect of the disparities alone could likely pay for the care for those who don’t have access now.”

Things I found:
Addressing Racial and Ethnic Health CareDisparities (Joint Center for Political and Economic Studies)

A range of policy strategies are available to federal, state and local governments, but it is important to recognize that no single policy – such as expanding access to health insurance – will fully address health care inequality. Health care disparities are complex and are rooted in many causal factors that span across a range of levels – including institutional, governmental and individual levels. It is therefore important to identify, implement and evaluate multi-level strategies addressing health care financing, systems and workforce development.

Fast Fact: Racial health disparities cost the country $50 billion annually (The National Council for Research on Women)

According to a study released this morning by theJoint Center for Political and Economic Studies,racial health disparities are costing the U.S. $50 billion each year. If you’re anything like me, your jaw just hit the floor. At the health reform briefing held this morning at the National Press Club to honor the launch of this study, speakers such as Secretary Sebelius, Majority Whip Clyburn, and researchers from John Hopkins and the University of Maryland discussed the moral and economic imperative of health equity.

Man…institutionalized racism is expensive.

Institutionalized racism is expensive

I’ve never seen the phrase “health disparities” twittered so furiously as it was when connected to “will cost the health care system $23.9 billion dollars.”

Couldn’t type “health disparities” into a search tool for the last two weeks without encountering both the Urban Institute’s and the The Joint Center for Political and Economic Studies’ investigations of how much it’s costing everyone to ignore health disparities.

It seems the Urban Institute and the Joint Center come up with pretty different estimates ($23.9 vs. $229.4 billion) of how much tax revenue is used to systematically undermine my health and wellness.

Estimating the Cost of Racial and Ethnic Health Disparities (Timothy Waidmann, 2009)

Aims

1) To make a “business case” for reducing health disparities by quantifying the economic costs of excess disease burden (to the health care system as a whole and to the Medicare and Medicaid programs, in particular).

2) To capture health disparities in conditions we can prevent and manage (e.g., heart disease, diabetes, hypertension, stroke and renal disease).

Methods

Waidmann analyzed 2003-2005 Medical Expenditure Panel Survey (MEPS) http://www.meps.ahrq.gov/mepsweb/ survey data, using a fairly typical multi-step process to come up with a regression model, simplified below.

Step 1: Spending = Disease prevalence * Age

This basic model was repeated for Medicare, Medicaid, and out-of-pocket spending.

Waidmann controlled for age, sex, insurance coverage, education, and income as alternate explanations for variations in spending. Where race is concerned, we know something ridiculous is going on from the outset. The differentials in age distributions by race. As you can see, things take a turn for the worse between 40 and 50 years-old, which corresponds to theories about key etiologic periods of disease and the cumulative effects of health risks across the life-span.

Given this evidence and since older people tend to have more spending money than younger people, the researchers also

Not. Pleased.

I’d never seen the phrase “health disparities” tweeted so furiously as when connected to the phrase “will cost the health care system $23.9 billion dollars.”

Both the Urban Institute and the Joint Center for Political and Economic Studies recently published investigations of how much it’s costing everyone to ignore health disparities. Working separately, the organizations have come up with different estimates of how much public money is used to systematically undermine my health and wellness.

An Urban Institute 2009 estimate of $23.9 billion is often quoted in the media. Joint Center representatives have quoted $50 billion in annual excess costs.

Estimating the Cost of Racial and Ethnic Health Disparities (Waidmann T., 2009: The Urban Institute)

Study goals

1) To make a “business case” for reducing health disparities by quantifying the economic costs of excess disease burden (to the health care system as a whole and to Medicare and Medicaid programs, in particular).

2) To capture health disparities in conditions we can prevent and manage (i.e., heart disease, diabetes, hypertension, stroke and renal disease).

Methods

Waidmann analyzed data from the 2003-2005 Medical Expenditure Panel Survey (MEPS) (to calculate the share of total costs by program and race/ethnicity), the Current Population Survey (to make state-level analysis possible), and the National Health Expenditure (NHE) (as a baseline for national estimates) to create projections. He used a seemingly straightforward regression analysis procedure, modeling Medicaid, Medicare, private insurance, and out-of-pocket spending by individuals, controlling for age, sex, insurance coverage, education, and income.

Waidmann's calculations from 2002-2005 MEPS survey. Things take a turn for the worse between 40 and 50 years-old, corresponding to theories about key etiologic periods of chronic disease (e.g., heart disease, hypertension) and the cumulative effects of health risk across the life-span.

As you can see in the “Age Distributions by Race” figure, something strange is going on. Given evidence of this strangeness (not to mention that a) older people tend to have more money to spend on health care than younger people; b) chronic disease treatment may get more expensive as people age, especially if these diseases aren’t being managed properly in the first place), Waidmann calculates the total cost effect of disease prevalence disparities by using age specific cost estimates for each disease. When calculating disease prevalence disparities, Waidmann seems to use non-Hispanic whites as the benchmark, since he estimates only the change in expenditures if African American and Hispanic age-specific expenditures are compared that of non-Hispanic whites.

Findings

In 2009, disparities among African Americans, Hispanics, and non-Hispanic whites will cost the health care system $23.9 billion dollars. Medicare alone will spend an extra $15.6 billion while private insurers will incur $5.1 billion in additional costs due to elevated rates of chronic illness among African Americans and Hispanics.

Over 10 years, the combined cost to all four sources of payment analyzed is estimated at $337 billion. The cost to the Medicare program of the two groups’ disparities is approximately $220 billion. The state and federal cost to the Medicaid programs is a more modest $27 billion, while private insurers and individuals are estimated to pay more than $90 billion over the period because of disparities in these conditions.

Thoughts…

…on what it means to me:

Health disparities are too expensive to ignore.

Acknowledging that the moral argument for bridging health disparities is compelling and right, the economic justification seems a mighty motivator for those struggling to identify with the less fortunate. Private insurers, in particular, should be paying attention–particularly since these cost estimates seem a little conservative. Unlike the Joint Center’s analysis, Waidmann looks only at direct medical costs associated with prevalence disparities in a few chronic, somatic diseases among “Hispanics” and “African Americans” compared to non-Hispanic Whites. As we learn from the Join Center’s analysis, there are a variety of other health status disparities, disparities in premature mortality exact significant costs and…Asians are people too.

Still, $337 billion is a pretty large sum–a little over Argentina’s GDP in 2008.

…on the research:

By excluding Asians, the Urban Institute study doesn’t have to deal with the tricky issue of, “which racial group is the healthiest and, therefore, the standard against which we should measure health disparities among other groups” (the Joint Center analysis finds Asians have the best health profiles for some conditions/health statuses, which complicates things for a secondary data analyst). In this study, Whites are healthiest and the evidence supports using Whites as a standard for comparison. From a theoretical perspective, I think this approach somehow incorporates racial discrimination as it is commonly understood and presented in the U.S (I’m not saying this understanding and presentation is right).

What about between-”marginalized group” comparisons?

I wonder about the potential for “between-’marginalized group’” comparisons since reality–and common decency–precludes us from sweeping all Black and Brown people under the same umbrella and stamping “Minority” on their foreheads.

Which brings me to this question:  Does the Urban Institute study conflate race and ethnicity? If so, is that okay, since this is data comes to us as a policy brief and since this is how the lay-public understands race and ethnicity (as suggested by this post)? Waidmann writes “race/ethnicity” several times in the brief, but the use of “Hispanics” and “African Americans” actually implies self-identified ethnicity (assuming only U.S. citizens and residents access Medicare and Medicaid).

Further, the MEPS data collectors seem to use “Black” as the racial category that Waidmann tags as “African American.” From 2002 onwards, MEPS collected data according to the Office of Management and Budget’s 1997 standards for collecting racial and ethnic data, meaning that self-identified Hispanics may belong to more than one race (which might be one reason MEPS uses “Black” and not “African American,” since it carries a different meaning).

That discrepancy suggests “Hispanic” actually measures a cultural affiliation and that the African American and Hispanic categories, as conceptualized in the Urban Institute study, may actually overlap. I guess this highlights the silliness of measuring race, but it also complicates meanings derived from comparisons to non-Hispanic whites and between marginalized groups.

It’s hard to say for sure because the study doesn’t include much information about how race and ethnicity is conceptualized (it’s not a peer-reviewed exercise, afterall).

What might we learn about from rigorous comparisons of non-Hispanic Blacks and Hispanics? What costs are associated with disease prevalence differences between non-Hispanic Blacks, non-Hispanic Whites and Hispanics? What might these costs mean to policy-makers and political strategists?

Next (and I mean it, I’ve got the post almost ready to go): The Economic Burden of Health Inequalities in the United States (LaVeist T.A., Gaskin D.J., & Richard P., 2009: Joint Center for Political and Economic Studies)

Things I found:

Overweight and Obesity: At a Glance (Surgeon General)

In men, Mexican Americans have a higher prevalence of overweight and obesity than non-Hispanic whites or non-Hispanic blacks. The prevalence of overweight and obesity in non-Hispanic white men is greater than in non-Hispanic black men.

Mexican American boys tend to have a higher prevalence of overweight than non-Hispanic black or non-Hispanic white boys.

A(H1N1): Health Disparities Update from Boston Public Health Commission

Well, this shanks my recent "Fall in New England" nostalgia.

Let’s learn about the Boston Public Health Commission’s commitment to addressing health disparities in Boston (dated September 18, 2009):

The video provides a nice–if somewhat general–explanation of the problem and displays Barbara Ferrer’s knowledge. While I have yet to find information describing BPH’s next steps, I’m not saying the information isn’t out there.[1]

In August I posted a bit of a rant about the Boston Public Health Commission’s epidemiological surveillance of A(H1N1) and how that surveillance was being covered in the media.

I knew I’d written a rant when Matt B., friend, Rant Creator, Connoisseur and Curator, left an insightful comment about it. Matt F., former boss and Boston native, followed up with a gentle “holding me accountable” via Google Chat. Capping off the feedback was my mother’s admonition that I…

“…remember there are a lot of people who LOVE Boston. Saying you hate a place in a serious website doesn’t come across as mature. Although I know you hate Boston. By the way, I LOVE Boston.”

Since my mother has been right about so many things–too many things (no, I shouldn’t have stopped going to dance class; yes, I should have taken that full ride to Howard; yes, I should exercise and eat more salad, etc)–I decided to remove the post.

After all, love it (Boston) or leave it alone?

August 26, 2009

First of all, I cringe—CRINGE—at the idea of Barbara Ferrer, executive director of a health commission in a major city (an incredibly and embarrassingly segregated city at that), reportedly being ignorant of racial, ethnic, and socioeconomic disparities in the distribution of flu.

“I thought people were getting sick everywhere at the same level, but it wasn’t true. Some of our neighborhoods were much more impacted by this,’’ said Barbara Ferrer, executive director of the Boston health agency.

Especially because, as this article suggests, Ferrer is/should be pretty knowledgeable!

If the quote of concern is accurate—and not a journalist’s misrepresentation of a venerable public health professional’s carefully chosen language—I’m disappointed and scared. My sense of being invisible in Boston—unless the police wanted to take issue with the nerdy (all black) Public Health parties I attended—isn’t “all in my head,” but is reflected in this report of BF assuming people in America get sick (or get the flu) at the same levels. After all, as executive director of the BPH, how can you be unaware of a persistent, major issue affecting the quality of life of a large section of the population? How can that be if we’re visible and equal constituents?

And so on.

[1] When/if I find it, you can be sure I’ll post it here for you.

Health Communications: What about mental health?

Healthreform.gov offers a very very simple 5-page break down of U.S. health disparities, complete with colorful graphics. At a 5-8th grade reading level, “Health Disparities: A Case for Closing the Gap” is probably helpful for everyone who wants to learn about is forced to learn about what’s going on with racial and ethnic health disparities in the U.S.[1]

One of the most glaring disparities is apparent in the African American community, where 48% of adults suffer from a chronic disease compared to 39% of the general population.

While “Health Disparities” isn’t intended to be some hugely serious treatise—and I’m definitely prone to unnecessary seriousness—I’m still interested in what made the 5-page cut and what didn’t.

Reading, learning from, and being edited by great writers (and designers) has shown me that the structure and organization of a deliverable can be as revealing as its text.

Just entertain this slightly shallow conjecture and speculation for a minute:

What are the perceived priorities? What’s mentioned first? Where are the paragraphs falling on the page?

Well, the usual suspects—obesity, cancer, diabetes, and HIV/AIDS—make an early appearance. The epidemiology of these diseases offers some of the clearest, most straightforward and alarming evidence of differential health outcomes by race (even after controlling for socioeconomic status!). Obesity is, of course, a risk factor for chronic disease. Meanwhile, the pervasiveness of chronic disease among communities of color has some really terrible implications for personal economic prosperity and productivity, as well as wealth accumulation across time.

Public Health folks will be happy to find a section on routine care and prevention that appears later down in the document, but not quite at the end! Colorectal cancer screening data even gets a shout-out with its own histogram!

The conclusion:

Need for Reform

The disparities in health care highlighted in this report demonstrate the need for reform. Meaningful reform must invest in prevention and wellness and ensure that all Americans have access to high-quality, affordable care. We can no longer afford to tolerate disparities in health.

What didn’t make it?

Mental health outcomes are nowhere to be seen, despite the pervasiveness of depression.

Depression (not including bipolar disorder) is the leading cause of disability among men and women of all ages in the U.S. and worldwide. Depressive disorders are common: they affect an estimated 9.5 percent of adult Americans in a given year, or about 20.9 million people (NIMH).

However, data on mental health disparities seem a little harder to conceptualize and communicate. In a pattern generally recognized as a departure from most health disparity trends, evidence suggests the prevalence of mental health disorders is actually lower among people of color than among Whites. One exception is relatively higher levels of schizophrenia diagnosed in the Black community, as revealed by three major community prevalence studies based on data from the National Comorbidity Survey, The National Comorbidity Survey Replication, and the Epidemiologic Catchment Area Study.

This finding could reflect the true distribution…but it might…not. Psychiatric epidemiology related to health disparities may be complicated by the reliability and validity of the measurement instruments administered to study participants. Additionally, mental disorder diagnosis is notoriously fraught (though I’ve known mental health professionals to argue that the diagnosis of somatic illness can generate similar anxieties, errors, difficulties, and angst).

Where mental health disparities are concerned, an important issue is not only missed diagnosis, but also wrong diagnosis and its consequences. Research suggests that Blacks with schizophrenia are overrepresented in the institutionalized population, and that psychiatrists tend to “overdiagnose schizophrenia and underdiagnose mood disorders in African Americans” (McGuire, 2008; Whaley, 2001; Neighbors et al. 1999; Strakowski et al. 1997; Snowden and Cheung 1990)[2].

[1] Look, you’re going to have to accept that the “strikethrough” meme is really never going to get old for me.

[2] If you want to take a look at these, ask me where to find them. In the meantime, I’ve linked to some resources I can quickly access.

Additional Resources (in date order)

Some of these will require subscriptions, unfortunately.

McGuire, T. & Miranda, J. 2008. “New Evidence Regarding Racial And Ethnic Disparities In Mental Health: Policy Implications.” Health Affairs 27(2): 393–403.

Kales, H., Neighbors, HW., et al., 2005. “Race, Gender, and Psychiatrists’ Diagnoses and Treatment of Major Depression Among Elderly Patients.” Psychiatric Services 56:721–
728.

Geiger, HJ. 2003. “Racial and Ethnic Disparities in Diagnosis and Treatment: A Review of the Evidence and a Consideration of Causes.” In Unequal Treatment: Confronting Racial
and Ethnic Disparities in Health Care. Report of Institute of Medicine Committee on Understanding and Eliminating Racial and Ethnic Disparities in Health Care. Brian D. Smedley, Adrienne Y. Stith, and Alan R. Nelson, eds., pp. 417–454. Washington, DC: National Academy Press.

Minsky, S., et al., 2003. “Diagnostic Patterns in Latino, African American, and European American Psychiatric Patients.” Arch Gen Psychiatry.60:637-644.

Neighbors HW, Trierweiler SJ, Ford BC, et al. 2003. “Racial differences in DSM diagnosis using a semi-structured instrument: the Importance of clinical judgment in the diagnosis of African Americans.” Journal of Health and Social Behavior 44:237–256, 2003

Safe Sex and Tupperware Parties

"People need to be given not only reasons to alter risky habits but also the behavioral means, resources, and social supports to do so."

The Pittsburgh AIDS Task Force is busy making moves.

As you’ll read (or may have read), its Girlfriends Project (GP) “targets at-risk [black] women” in the Pittsburgh metro area for HIV/AIDS awareness raising, assertiveness counseling, and domestic violence education. Between January 1 and June 15, 2009, the GP had 135 women to take advantage of free, confidential HIV testing…none of the women have tested positive.

The project’s structure and organization brings the generally uncomfortable topic of HIV/AIDS (associated in the collective consciousness with sex, death, sickness, needles, loss of ability, possible social alienation, etc) into participants’ comfort zones. Seems like a great way to reduce some of the stigma and anxiety attached to HIV testing.

What’s also interesting to me is that the program leverages social networks and social support to address black women’s health issues. 

Here’s how it works (reminds me of a Tupperware Party!):

Women, usually five to eight of them but sometimes as many as 15, get together at the home of a friend, who must be able to provide a private area in which Ms. Dukes can do the HIV/AIDS swab test one woman at a time after her talk. The task force provides refreshments and the hostess receives a $50 gift card for volunteering the use of her home. The attendees get $20 gift cards.

The party starts with some ice-breaking conversation, an HIV/AIDS quiz, raffles and giveaways of things like body lotions, and the opening of gift bags containing condoms and other safe-sex products. Then comes Ms. Dukes’ educational program, in which the women learn how to protect themselves from both HIV/AIDS and domestic violence…

After the presentation, Ms. Dukes gives another quiz on HIV/AIDS to see what the women have learned. Then, she offers the HIV/AIDS swab tests. She sends the tests on to the health department for processing. If there were a positive, the result would come back to Ms. Dukes, who would inform the woman. (Pittsburgh Post-Gazette, 2009)

So far, I guess about five women have been tested per week of the program’s life (though it’s likely the actual rate of testing will increase exponentially as women tell their friends and family about the resource). While 135 women is a tiny number in proportion to the population, it’s certainly better than zero testing, which was apparently the rate in Braddock, Clairton and Duquesne before the project got going.

Verdict on this intervention: 

Backed by reasonable evidence? Check.

Gives me that “Smell of Fresh-Baked Brownie Feeling,” which is good for marketing and other outreach? Check.

Potential for far reach and influence? Debatable.

In all fairness, the program is new and small and hungry. That said, I wonder whether growth (i.e., more money) has been linked explicitly to outcomes (will there be a program evaluation?), or whether the program is more of a symbolic tool to draw necessary attention to HIV/AIDS in Pittsburgh’s black community.

While I personally believe helping even one person justifies social action, professionally, I wonder about this program’s long-term potential to affect hard outcomes–specifically, reductions in rates of new infection in target communities.

Conclusion:

I think the force of GP’s impact will hinge on the breadth of its reach, which in turn relies on the mysterious powers of family, friendship, community, and word of mouth.

Are these enough?

Additional resources:

“Views and Experiences with HIV Testing Among American Americans in the U.S.” (Kaiser “We Don’t Mess Around” Family Foundation, June 2009).

Using data from the Kaiser Family Foundation 2009 Survey of Americans on HIV/AIDS, we find that in many ways, African Americans’ reported views and experiences reflect the disproportionate impact of the epidemic on their community.  Compared with whites and Latinos, African Americans are more likely to say AIDS is a more urgent problem for their local community now than it was a few years ago, more likely to say they know someone who is infected with HIV or has died from AIDS, and more likely to say they are personally very worried about becoming infected themselves.

“Social Cognitive Theory and Exercise of Control over HIV infection” in Preventing AIDS: Theories and Methods of Behavioral Interventions. (Albert Bandura, 1994)

To achieve self-directed change, people need to be given not only reasons to alter risky habits but also the behavioral means, resources, and social supports to do so. 

“From social integration to health: Durkheim in the new millennium.” (Berkman et. al., 2000)

First, social networks via social influence or supportive functions influence health-promoting or health-damaging behaviors such as tobacco and alcohol consumption, physical activity, dietary patterns, sexual practices, illicit drug use. Second, social networks via any number of pathways influence cognitive and emotional states such a self-esteem, social competence, self-efficacy, depression and affect. Third, networks may have direct effects on health outcomes by influencing a series of physiologic pathways largely related to stress responses.