Unintended pregnancy and health disparities

.
In "Health in All" policies to eliminate health disparities are a real answer, August 18, 2011, I discussed the work of Steven Woolf, MD, as it relates to health disparities. The major point of that piece is that the health and mortality differences between groups, particularly racial groups, in the United States accounts for an enormous number of excess deaths. If that gap were closed, and everyone in the US had the same age-adjusted death rate as whites, the number of lives saved would far exceed those saved by all medical care. Indeed, it would far exceed the number of lives saved even by public health interventions, at least as narrowly construed. Many of the social interventions that Woolf and colleagues indicate would be necessary to decrease disparities could be thought of as “public health” in a broader sense, because they would improve the public’s health, but in general eliminating poverty and raising educational levels are not part of the narrower public health construct.

In “Unintended pregnancy in the United States: incidence and disparities, 2006”, published on-line-before-print in Contraception, Lawrence B. Finer and Mia R. Zolna of the Guttmacher Institute report on the disparities in a particular group, women of reproductive age, in relation to unintended pregnancy. They combined data from several sources, “…on women's pregnancy intentions from the 2006–2008 and 2002 National Survey of Family Growth… a 2008 national survey of abortion patients and data on births from the National Center for Health Statistics, induced abortions from a national abortion provider census, miscarriages estimated from the National Survey of Family Growth and population data from the US  Census Bureau,” to assess rates of unintended pregnancy and disparities between groups, and compared  this data to rates in 2001.

 They found that the percent of unintended pregnancies remained high, with a slight increase (from 48% to 49% of all pregnancies) from 2001 to 2006. The actual rate increased from 50 to 52 unintended pregnancies for every 1000 women aged 15-44. There was a significant decrease in the rate of unintended pregnancies in women 15-17 years old, but this group still had the highest rates (79%, down from 89%). While the fact that an increased percentage of pregnancies in such young women were intended is not necessarily a good thing, the overall pregnancy rate per 1000 decreased from 47 to 42 in this group. The rates of unintended pregnancy went down with age, but all other age groups had an increase in their rates from 2001-2006, the largest in women 18-24. To say this again: the rates of unintended pregnancy went up in each age group except 15-17, but that group still had the highest rate, with rates decreased in each older age group.

The most important finding was the disparity in the rate of unintended pregnancy by characteristics other than age: by race/ethnicity, by income, and by educational level. The unintended pregnancy rate for women with less than a HS diploma (80 per 1000) was more than 2.5 times that of college graduates (30); the rates for women who were HS grads and those with “some college” were in between. The rate for Black women (91) and Hispanic women (82) was also 2-3 times that of white non-Hispanic women (36). Income, perhaps, had the greatest disparity: the rate for women at <100% of poverty (132) was more than 5 times the rate for women >200% of poverty (24).

OK. This is a lot of data, and maybe it is hard to follow. But the main point is simple: these are staggering differences, and they are difference based upon the same social factors that Woolf and his colleagues address. The magnitude of these differences overwhelms all the other factors that affect this rate. The women whose resources make them least able to economically provide for unplanned children are most at risk of having them.

The percent of unintended pregnancies ending in abortion also decreased, from 47% to 43%, with the greatest decrease (from 47% to 41%) in women 20-24, but rather than being a positive, this decrease is much more likely to reflect the decreased availability of abortion services than a shift in attitudes toward abortion. That is, a larger number of children are being born as a result of unintended pregnancy to families that will have difficulty caring for them. In addition, these families are getting less and less aid from public sources because the same folks who are against abortion and the protection of the “unborn” are also against social services that will help the families of the born.

This study was also the basis for the excellernt column “Failing Forward” by Charles Blow in the NY Times on August 27, 2011.  He makes these points very strongly, commenting on the policies that restrict access to abortion while effectively punishing the children:

“This is what we’re saying: actions have consequences. If you didn’t want a child, you shouldn’t have had sex. You must be punished by becoming a parent even if you know that you are not willing or able to be one. This is insane.”

As in all of Blow’s columns, he includes a telling graphic, here showing the “States of Child Hunger”, the rate and raw number of children in food-insecure households. There are over 17 million hungry children in the US, or 23.2% of all children. The highest rate is in DC, the lowest in North Dakota. After DC (32.3%), perhaps surprisingly, is Oregon (29.2%). However, after that, unsurprisingly, come the usual suspects , many of the states most commonly associated with poor social supports and frequently conservative Republican leadership: Arizona, Arkansas, Texas, Georgia, Mississippi, Nevada, South Carolina, Florida. Most of the New England states are clustered near the bottom (good) end of the list.

The whole thing is not good. Too many poor and hungry children, too little education, too little opportunity for too many women and their families. Too many people and families caught in the multiple challenges of poverty, poor education, and racial/ethnic minority status, all of which are independently associated with health disparities, and which are synergistic in their effect when found together. This is not a society to be proud of. This is a society that needs great change, and it is the change perhaps we’d hoped for with the election of President Obama.

Frequently, the comic strips (not even the overtly “political cartoons”) capture it best. Here is a link to a “Non Sequitur”, by Wiley Miller. Check out September 4, 2001, with the adventures of super “hero” “CongressMan”. Laugh. And then cry.

.

"Health in All" policies to eliminate health disparities are a real answer

.
At the recent American Academy of Family Physicians (AAFP) sponsored National Conference of Family Medicine Residents and Students (NCFMRS) held in Kansas City (hey, it can be an advantage to live in the Capital of Family Medicine!), I had the pleasure of hearing Dr. Steven Woolf’s presentation “Health Disparities and the Role of the Family Physician”. I have previously cited (eg, Economics and Disease Prevention, February 13, 2009) the work of Dr. Woolf, a professor in the Department of Family Medicine at the Virginia Commonwealth University (VCU) College of Medicine. He is one of the most distinguished health services researchers in the country, and a clear and articulate speaker.

His presentation, appropriately for residents and students, began with a description of the role of the family physician in the care of both individual patients and the communities in which they practice. He noted the major health toll exacted by “personal health behaviors”, headed by tobacco (accounting for 400,000 deaths per year) and including diet/activity (300,000), alcohol (100,000), microbial agents (90,000), toxic agents (60,000), firearms (35,000), sexual behaviors (30,000), motor vehicles (25,000), and illicit use of drugs (20,000, presumably not counting licit use of prescribed drugs). Note that, of these, only microbes are in the usual “traditional” area of medicine. His point was that to truly improve health the practitioner needs to go outside the hospital or office to the community, where these causes of ill health are located. “Health in All” policies include transportation, land use, built environment, taxes, housing, agriculture, environmental justice, etc. As an example he notes that 2.3 million (2.2%) of continental US households are more than a mile from a supermarket and do not have access to a vehicle. While Russell Shorto’s “Opinionator” piece in the NY Times July 31, 2011, “The Dutch way: bicycles and fresh bread” points out that one advantage of using bicycles as much as the Dutch do is that you can’t carry more than a day’s worth of groceries so that the bread can be fresh and preservative-free, this is not the way of things in the US; poor access to healthful food is a big contributor to poor health.

 

Woolf then went on to demonstrate the familiar (but maybe not to residents and students) data on the remarkable health disparities that exist in the US, particularly by race, socioeconomic status, and educational level (although also by geography, language, and other characteristics). For example, the attached table shows a 24-29% disparity between the age-adjusted mortality rates of white and black males over the period from 1991-2000. The other table shows that the health status of college graduates is much better than that of people with some college, which in turn is much better than HS graduates who have been health than those who have not graduated from HS. These disparities are truly remarkable and should be intolerable. Woolf notes that while most of our current research is focused on finding treatments that are effective, we would save far more lives if we were to focus, instead, on equity by eliminating the health disparities gap.

In a “thought experiment” that he published with colleagues in the American Journal of Public Health in 2007, “Giving everyone the health of the educated: an examination of whether social change would save more lives than medical advances”[1], Woolf demonstrates that even if we attribute all reduction in mortality over to medical advances (nowhere near true; most are due to the types of societal change generally characterized as “public health”, such as clean water, sanitation, and cleaner air), eliminating the disparities that exist on the basis of educational level would dwarf that change, as shown in this table. 

But our society spends virtually all of its research dollars on developing new treatments that then are available to only a portion of our population, and have relatively little effect even on those who receive it. Big PhARMA spends $32 Billion/yr. While the National Institutes of Health (NIH), which is part of the Department of Health and Human Services (HHS), spends $28 Billion/yr, primarily on developing new treatments, HHS’ Agency for Health Quality and Research (AHRQ), which looks at systems and utilization, receives only $300 million/yr.

Woolf and colleagues have also developed a great new tool for visually looking at the impact of disparities on health on a state and county level. Called the “County Health Calculator” (http://chc.humanneeds.vcu.edu), it allows you to pick a state or a county and compare its income level and educational level to that of the entire country (for states) or the entire state (for counties), or one state or county to another. It also presents the “best” and “worst” levels (for state or county), and allows (this is really neat!) you to use a “slider” to move the indicator to find out how changing the educational or income levels up or down would affect mortality. That is, you can see how many more lives would be saved – or lost – if your state (or county) had the income or educational levels of the average, best, or worst state (or county).

The results are truly amazing.  In Kansas, for example, the two suburban Kansas City area counties are almost polar opposites. Johnson County has the highest level of education, measured as % of adults with at least some college, at 78%, and income, measured as % of adults in households that have a “basic income” (defined as at least 2x the poverty level), 86%. Wyandotte County is near the bottom, with a “some college” education level of 39% and a “basic income” level of 56%. (Statewide levels in Kansas are 58% for education and 70% for income.) Using the slider, we discover that if Wyandotte County’s basic income level of 56% were raised to Johnson County’s 86%, 201 or 28% of deaths would be averted per year. If the “some college” education level were to go from the actual 39% to Johnson County’s 78%, it would result in a reduction of 272, or 38%, in the annual death rate!

No drug comes close to this. No treatment of any kind comes close to this. If a new drug were shown to reduce mortality from a disease by 5%, or even 1%, it would get incredible advertising – hundreds of millions of dollars – and huge publicity, in both the scientific and lay press. But the simple fact that so many more deaths could be prevented, so many lives could be improved, by addressing the social determinants of health, is scarcely covered, and hardly funded at all.

We are in thrall to, primarily, the pursuit of corporate profit, but also to a “scientific” perspective that sees new discoveries as the true goal of research rather than “moving the bar” and changing the actual outcomes for actual people. Eliminating health disparities, which improve health and decrease the death rate, will also “bend the cost curve” in ways that only improving access will not. In one of his slides, Woolf presents data from Milstein, et al, in a 2011 Health Affairs article [2] demonstrating that over the next 25 years neither providing coverage nor coverage + access to care, but only both plus “protection” – addressing behavioral and environmental risks – will do so.

There is no contest to the value of a dollar spent on changing social conditions as opposed to finding “more effective” treatments.  There is also no contest in where we actually spend our money. There is something rotten in the state, and it isn’t the state of Denmark.

---

[1] Woolf S., et al., “Giving everyone the health of the educated: an examination of whether social change would save more lives than medical advances”, Am J Public Health. 2007;97:679–683

[2] Milstein B et al., “Why behavioral and environmental interventions are needed to improve health at lower cost”, Health Affairs 2011;30:823-832

Disparities in physician income are related to disparities in health

.
A major focus of public health, about which I have written several times, is addressing the disparities in health arising from modifiable conditions (such as inequality of opportunity, income, and racial and ethnic differences). While it is common for those with privilege – wealth, health, opportunity – to believe that they have these privileges because they are “deserving”, i.e., because of their hard work, education, etc., such an outlook minimizes the critically important fact that there are lots of people who work just as hard and have very little. The “illegal immigrant”, working 3 minimum (or sub-minimum) wage jobs to try to just keep his or her family fed and housed, is not only not lazy, but working a lot harder than many of us who are able to enjoy weekends off, play golf, watch the kids’ sporting events.

So it is clearly not hard work alone. It is very much influenced by where you start, and what your opportunities have been. These are the social determinants of health (e.g., Social Determinants, Personal Responsibility, and Health System Outcomes, September 12, 2010) and the capability that people have of acting in healthful ways (Capability: understanding why people may not adopt healthful behaviors, September 24, 2010). It is common for people who have a lot to minimize, rather than to emphasize, the degree to which their position at the start of the race has affected their current position. Attacks on those who would seek to redress some of the most egregious inequities are still couched in terms of “economic class warfare” by those who have already won the war. Those with privilege are very concerned about change that would leave them with less of a leg up; they may give lip service to Horatio Alger heroes, but are more likely to wish to follow the model of George W. Bush. “If you’ve spent your entire life with the wind at your back,” a wise sage once noted (and I do not know to whom to attribute it), “a calm day seems unfair”. In terms of health, the connections are very clear. It is good for your health to be born rich. The Horatio Alger hero, pulled up by their own bootstraps, has worse health outcomes than the child born to privilege.

Health disparities, then, are real and important. But why should we – should anyone – be concerned about the disparities in physician income? After all, even the more “poorly” paid specialists, in primary care, make far more than the average American. Yes, they have worked hard to get into and through medical school and residency training, but, just as noted above, so have a lot of other people who will never make nearly as much. The problem is that, if the presence of a larger number/percent of primary care physicians is associated with improvements in the health of the population, and if the presence of wide disparities in income significantly influence students to choose higher paid specialties instead, then these disparities in health status are likely to continue and the overall health of the American people is likely to suffer. There is good data on both counts. Many of the posts in this blog have addressed the first, the positive influence of primary care on the health of the population (e.g., Lower Costs in Grand Junction: More Primary Care, Less High Tech, October 18, 2010; Primary Care, IMGs, and the Health of the People, August 14, 2010; and many others) and on health disparities[1]. I have also addressed the other point, the decrease in the number of students choosing primary care careers (e.g., Primary care specialty choice: student characteristics, July 12, 2010; Primary Care’s Image: A Problem?, November 17, 2009, and others).

A study published in the Archives of Internal Medicine by Leigh, et. al, “Physician wages across specialties”[2], is the most recent effort to quantify the differences. They utilized the large Community Tracking Study (CTS) of physicians from 2004-2005 to gather information on physician income. They grouped the physician responses into 4 broad categories (surgical, internal medicine and pediatric subspecialties [IMPSS], primary care, and other) and again into 41 specific specialties. They went beyond previous studies to calculate gross personal income on an hourly basis (thus controlling for hours worked per week) and did further adjustments to control for other variables, principally sex and age. They used a statistical manipulation to estimate incomes above the maximum set for the CTS (for some reason set at $400,000, much lower than many subspecialists make).

The outcomes were not surprising in comparison to previously reported data. In the 4 broad-group comparisons, primary care physicians averaged about $60/hr compared to IMPSS at $85, other medical at $88, and surgical at $92. In the single specialty comparisons, General Surgery was taken as a reference being actually near the middle ($86/hr), with the top incomes in neurosurgery ($132), radiation oncology ($126), and medical oncology ($114). At the bottom were family practice, general practice, general internal medicine, geriatric medicine, internal medicine/pediatrics, and “other” pediatric subspecialties (whichever those may have been) with a range of $50-$58.

There are several reasons to think that differences are, in fact, greater than those reported. There was only a 53% response rate to the CTS, and so we do not know if non-respondents made more, less or the same as respondents. “Hospital-based” specialties, specifically anesthesiology and radiology, which are among the highest-paid, were excluded. Other high-end specialties, such as cardiovascular surgery or transplant surgery, do not appear as specific specialties, and may have their incomes hidden when grouped with “thoracic surgery” or “other surgical specialties”. There are many sources of income for many physicians, including a variety of expenses that can be paid by practices and which would presumably be greater for higher income practices. Many highly-paid specialties are paid by hospitals directly (such as anesthesiology and radiology) or through “physician service agreements”. The correction used by the authors of the study for incomes over $400,000 could have been inadequate; certainly anecdotal experience in many locations would suggest that considering $400,000 as a reasonable top end for the highest paid physicians would understate that by at least half.
Nonetheless, the income differences, even in 2004-05, were impressive. Given the debt load that medical students (particularly those, obviously, from the less wealthy families) graduate with, the significant attraction to higher pay is clear.

The Wall St Journal, in two recent articles (“Secrets of the system”) published October 26, 2010, looked at the Relative Value Update Committee (RUC), a group of 29 physicians convened by the AMA from different specialty organizations that make recommendations to Medicare on how to pay physicians for their, well, relative value. One, “Physician panel prescribes the fees paid by Medicare” by Anna Wilde Matthews and Tom McGinty, describes how this group meets to divide up a pie that Medicare seeks to keep constant. In the other, “Dividing the Medicare pie pits doctor against doctor”, Matthews discusses the contentiousness that happened when primary care physicians (greatly outnumbered) challenged their surgical colleagues to get a higher portion of the money (that is, to revalue activities done by primary care physicians relative to surgical specialists).

In the same issue of Archives of Internal Medicine that Leigh’s article appeared in, Federman and colleagues[3] surveyed physicians about whether they thought reimbursements were inequitable or not; 78.4% agreed that they are, with not that much difference between generalists and subspecialists. However, when the idea of shifting payments from subspecialists to generalists was raised, there was a marked difference; 66.5% of generalists supported this, while only 16.6% of surgeons did; overall 41.6% were supportive and 46.4% were opposed. That is, for most specialists, paying generalists more is ok, but paying themselves less is not.

The WSJ‘s Matthews quotes an email from Jonathan Blum, deputy administrator for the Centers for Medicare and Medicaid Services (CMS) saying that the Medicare agency is moving to “improve Medicare's physician systems to correct historical biases against primary-care professionals." That needs to happen. The changes need to be dramatic. And they need to happen soon.

[1] Shi L, Macinko J, Starfield B, Xu J, Regan J, Politzer R and Wulu J, “Primary care, infant mortality, and low birthweight in the states of the USA”,J Epidemiol Community Health 2004;58;374-380
[2] Leigh JP, Tancredi D, Jerant A, Kravitz RL, “Physican wages across specialties: informing the physician reimbursement debate, Arch Int Med 25Oct2010; 170(19):1728-34.
[3] Federman AD, Woodward M, Keyhani S, “Physicians’ opinions about reforming reimbursement: results of a national survey”, Arch Int Med, 25Oct2010;170(19):1735-42.
.