The increasing gap in life expectancy between the rich and poor

An article was recently published online by the Journal of the American Medical Association (JAMA) describing the association between income and life expectancy in the United States between 2001 and 2014. The article highlights major differences in life expectancy by income group and that these differences increased from 2001 to 2014. The figure below, an adaptation of Figure 3 from the article, shows that in 2001 the difference in life expectancy among men between those in the lowest income group (mean income =  $17,000 per year) versus those in the highest income group (median income = $256,000 per year) was 9 years and by 2014 this difference had increased to 10.6 years. Among women, the difference in life expectancy in 2001 between those in the lowest income group (mean income =  $16,000 per year) versus those in the highest income group (median income = $243,000 per year) was 4.3 years and by 2014 this difference had increased to 6 years. From 2001 to 2014 the gap between those in the highest income group and the lowest income group widened by about the same amount for men (1.6 years) and women (1.7 years).

Change in race- and ethnicity-adjusted life expectancy from 2001 to 2014 among men and women by income group

income inequality

Notes: The data points for this figure were extrapolated from Figure 3 of the Chetty et al article. The quartiles of median income were based on the median household earnings among working individuals. Only the top and bottom quartiles are shown to highlight the gap between these two extreme income groups. 
Data Source: Figure 3 from Chetty et al. The Association between Income and Life Expectancy in the United States, 2001 to 2014. Published online by JAMA on April 10, 2016.

How much does a colonoscopy cost?

A colonoscopy is a procedure that uses a colonoscope, a thin, flexible tube, to inspect the inner lining of your rectum and colon for ulcers, polyps, tumors, inflammation or bleeding. While the procedure itself is fairly standard, the cost for the procedure is not. The figure below highlights the differences in the cost of a colonoscopy both within and between the 20 most populous cities in the U.S.

The average cost of a colonoscopy is lowest in Memphis, TN ($1,666) and highest in San Jose, CA ($3,718). That’s a $2,052 difference in the cost of the procedure just based on whether you live in San Jose or Memphis. A city many may assume would be on the more expensive end, New York, is actually on the lower end, with an average cost for a colonoscopy of $2,062.

Differences in the cost of a colonoscopy within and between 20 most populous U.S. cities

What is perhaps more interesting than the difference between cities is the difference found within cities. San Jose not only has the highest average cost for a colonoscopy but also has the greatest difference in cost. The lowest available cost for a colonoscopy in San Jose is $1,924, while the highest cost is $6,618. That’s a $4,694 difference in cost depending on where in San Jose you have the procedure performed.

The figure is interactive. Hover over the colored bars to get the exact costs in each city as of March 11, 2015.

Source: http://www.guroo.com/#!care-bundles/CS001

How do countries of the 2015 women’s world cup stack up when it comes to social progress?

The 2015 women’s world cup just ended and the U.S. came out on top. However, when we look at the social progress of the same countries that competed in this year’s world cup, Norway wins out as the most socially progressive nation. The two teams that competed in the final this year, Japan and the U.S., rank 10th and 11th, respectively.

Social Progress, as measured by the Social Progress Index, measures how well nations meet the needs of its citizens. The Social Progress Index is made up of three domains: basic human needs, foundations of well-being and opportunity. The basic human needs domain consists of nutrition and basic medical care, water and sanitation, shelter and personal safety. The foundations of well-being domain consists of access to basic knowledge, access to information and communications, health and wellness and ecosystem sustainability. The opportunity domain consists of personal rights, personal freedom and choice, tolerance and inclusion and access to advanced education.

In the figure below, the countries competing in this year’s women’s World Cup are plotted along the Y-axis, ranked from top to bottom. The most socially progressive countries (e.g. Norway) appear at the top of the y-axis and the least socially progressive countries (e.g. Nigeria) fall at the bottom. Along the x-axis is the Social Progress Index, which can range from 0 to 100. The minimum for this sample of countries is 43.3 and the maximum is 88.4.

Countries ranked according to the 2015 Social Progress Index

Countries ranked according to the 2015 Social Progress Index

Note: There is not enough publicly available data to calculate the social progress index for Côte d’Ivoire.

Source: 2015 Social Progress Index http://www.socialprogressimperative.org/

Drugs oncologists prescribe cost a lot more than other specialty drugs

Yesterday, for the first time, the Center for Medicare and Medicaid Services (CMS) released data on prescription claims under Medicare part D at the provider-level. This means you can now see the number and cost of drugs prescribed by each individual health care provider (e.g. physician, physician assistant or nurse practitioner). Over the course of 2013, over one million unique providers prescribed $103 billion in prescription drugs.

There are a lot of interesting ways to look at this data, but one quick view is to look at the average cost of a prescription by specialty of the health care provider. The figure below shows that when a hematologist/oncologist prescribes a drug it costs hundreds of dollars more than the average drug prescribed by providers in different specialties. Despite the very high average cost per drug prescribed by hematologists/oncologists, the total cost of prescription drugs prescribed by hematologists/oncologists was $5 billion in 2013 compared to the $23 and $27 billion prescribed by family and internal medicine physicians, respectively, due to the large volume of patients seen by family and internal medicine physicians every year.

Prescription drug claims made in 2013 under Medicare Part D

Graphic 31Notes: For a detailed description of the limitations of this data please see the original press release referenced below.
Data Source: Chart 2 from the Center for Medicare and Medicaid Services press release on April 30, 2014.

The lifetime risk of being diagnosed with cancer in blacks vs. whites

The first data viz below shows the lifetime risk of being diagnosed with any cancer among blacks and whites in the U.S. by sex. White males have the highest lifetime risk (44.3%), while black women have the lowest risk (36.4%). Black men have the same risk as white women (42.0%).

It is important to note that these estimated lifetime risks take into account competing causes of death. This means that because black men are much more likely to die of homicide than white men, black men may have a lower lifetime risk of cancer because they have a higher risk of dying from other things, like homicide.

Lifetime risk of being diagnosed with cancer

Males and females

Now let’s look at the black vs. white difference in the lifetime risk of a cancer diagnosis by cancer type among men. Along the y-axis in the figure below is cancer type, ordered from top to bottom according to the difference in risk. Melanoma is at the top because the black vs. white difference is -4.66%. This means that black men have a 4.66% less chance of being diagnosed with melanoma in their lifetime compared to white men. Prostate cancer is at the bottom of the y-axis because the difference is +4.92%, meaning that black men have a 4.92% greater chance of being diagnosed with prostate cancer in their lifetime compared to white men. For the majority of cancers, black men and white men have a very similar lifetime risk of diagnosis.

Lifetime risk of being diagnosed with specific cancers

Quantitative difference
Black males vs. white males

If we were to classify a difference in lifetime risk between black and white men of + or - 1% as being roughly the same risk then there would only be four types of cancer, not including all sites, for which the lifetime risk in black vs. white men was different. These four cancers are melanoma, non-Hodgkin lymphoma, prostate and bladder cancer, as shown in the data viz below. The cancers in green are those for which the lifetime risk is lower in black men vs. white men (i.e. <1% different). The cancers in red are those for which the lifetime risk is higher in black men vs. white men (i.e. >1% different). The cancers in grey, which are the majority of the cancer types, are those for which the lifetime risk is roughly the same (i.e. + or - 1%).

Lifetime risk of being diagnosed with specific cancers

Qualitative difference
Black males vs. white males

Now let's take a look at the difference in the lifetime risk of a cancer diagnosis, by cancer type, in women. In the data viz below, the y-axis is sorted from top to bottom according to the difference in risk in black vs. whites, with the smallest value at the top and the largest value at the bottom. Black women are less likely, similar to what we saw above with black men, to be diagnosed with melanoma than white women. Black women are also less likely to be diagnosed with breast and lung cancer than white women.

Lifetime risk of being diagnosed with specific cancers

Quantitative difference
Black females vs. white females

If we were to classify a difference in lifetime risk between black and white women of + or - 1% as being roughly the same risk then there would only be three types of cancer, not including all sites, for which the lifetime risk in black vs. white women was different: breast, lung and melanoma (shown below). There are no red bars in this data viz for women, as there were for men. This is because there are no cancers for which black women are more likely to be diagnosed during their lifetime than white women.

Lifetime risk of being diagnosed with specific cancers

Qualitative difference
Black females vs. white females

Overall, what the five data viz in this post highlight is that the overall risk of a lifetime diagnosis of cancer is lower in blacks vs. whites. However, the overall lower risk is really driven by a few cancers, since the lifetime risk is roughly the same for the majority of cancers. Given this overall conclusion, one might assume this translates to a lower, if not similar, overall lifetime risk of death from cancer in blacks vs. whites. To find out if this is true stay tuned for the next post.

Notes: A cancer diagnosis includes both invasive and In situ cancers. 

Data Source Males: SEER Cancer Statistics Review 1975-2011. Table 1.16. Lifetime Risk (Percent) of Being Diagnosed with Cancer by Site and Race/Ethnicity Males, 18 SEER Areas, 2009-2011.
Accessed on February 24, 2015.

Data Source Females: SEER Cancer Statistics Review 1975-2011. Table 1.17. Lifetime Risk (Percent) of Being Diagnosed with Cancer by Site and Race/Ethnicity Females, 18 SEER Areas, 2009-2011.
Accessed on February 24, 2015.

Is thyroid cancer being overdiagnosed?

An article published in the New England Journal of Medicine and discussed on Vox, showed that the incidence of thyroid cancer in South Korea has skyrocketed in recent years while mortality from the cancer has remained relatively stable. The figure below shows the relationship between incidence and mortality from thyroid cancer for high-income countries in 2012. While the difference between incidence and mortality from thyroid cancer is extremely large in South Korea, indicating a potential problem of overdiagnosis, many of the other high-income countries, like the United States, are likely to suffer from the same problem.

Incidence and mortality from thyroid cancer in high income countries in 2012

Notes: The y-axis contains a list of 46 high-income countries based on the World Bank classification as of July 1, 2014. Countries are rank ordered based on incidence from high to low.

The next figure shows the relationship over time in the United States between incidence and mortality of thyroid cancer. The incidence of thyroid cancer has been rapidly increasing since 1975 while mortality has remained relatively stable, indicating that like South Korea, the United States likely has an overdiagnosis problem when it comes to thyroid cancer.

Incidence and mortality from thyroid cancer in the U.S., 1975-2010

Notes: Rates are age-adjusted to the 2000 US Standard Population.

Data Source Figure 1: Ferlay J, Soerjomataram I, Ervik M, et al. GLOBOCAN 2012 v1.0, Cancer Incidence and Mortality Worldwide: IARC CancerBase No.11 [Internet]. Lyon, France: International Agency for Research on Cancer; 2013. Available from: http://globocan.iarc.fr. Accessed on February 11, 2015.
Data Source Figure 2: Incidence data comes from SEER 9 areas (San Francisco, Connecticut, Detroit, Hawaii, Iowa, New Mexico, Seattle, Utah and Atlanta) and Mortality data comes from US Mortality Files, NCHS, CDC. All data was downloaded from http://seer.cancer.gov/faststats/ on February 11, 2015.

Nursing home staff have very low flu vaccination rates

A recent article published in the American Journal of Infection Control found that only 53.9% of nursing home staff in three states (Florida, Georgia and Wisconsin) had been vaccinated against the flu during the 2011-2012 flu season. Low vaccination rates among nursing home staff are of public health concern as they may put many nursing home residents at risk.

The figure below is a re-design of a figure published in the original article and shows the wide variation in vaccination rates by facility. In one facility in Florida only 15% of nursing home staff were vaccinated against the flu.

Nursing home staff vaccination rates by facility in three states

Vaccine FigureNotes: Each square represents the estimated proportion of nursing home staff members that received a flu vaccination at each facility. The line represents the 95% confidence interval of the proportion vaccinated. Facilities are ranked based on the proportion of staff vaccinated at each facility and ordered from low to high along the x-axis   

Data Source: Re-design of Figure 1. JD Daugherty et al. Influenza vaccination rates and beliefs about vaccination among nursing home employees. American Journal of Infection Control, Volume 43, Issue 2, 2015, 100 – 106.

 

Is the amount of disease-specific NIH funding related to burden of disease?

Update: This Data viz was featured in an article on Vox

A recent article published in JAMA looked at the relationship between NIH funding and the U.S. burden of disease. The figure below is based off of data published in a supplementary appendix. For most conditions the relationship between the % burden in the population and the % of NIH funding is fairly similar. However, some conditions receive substantially more in funding relative to its burden (HIV/AIDS and cancer), while some receive much less (injuries).

Relationship between amount of NIH funding and burden of disease

Burden of disease and funding

Notes: % of total NIH funding was calculated by dividing the condition-specific funding amount by the total NIH spending amount for the 27 listed conditions. % of total burden of disease was calculated by dividing the condition-specific disability-adjusted life years (DALYs) amount by the total DALYS for all 27 listed conditions.

Source: eTable 3.  NIH Disease-Specific Research Funding and Burden of Disease for Selected Conditions. Moses HM III, Matheson DHM, Cairns-Smith S, George BP, Palisch C, Dorsey ER. The Anatomy of Medical Research: US and International Comparisons. JAMA. doi:10.1001/jama.2014.15939

Are scientists guilty of newsjacking Ebola?

Given that Ebola is such a hot topic, newsjacking of Ebola to make stories relevant is commonplace. However, journalists aren’t the only ones who benefit from staying relevant, scientists benefit as well. This dataviz was created to capture any uptick in scientific articles about Ebola since the most recent epidemic.

Trend in the % of scientific articles about Ebola from 1990-2014

Graphic 10

Notes: To obtain the percent of publications published each year the tile/abstract field was searched for ‘Ebola’ every year from 1990 to 2014 in PubMed. 2014 data is up to 10/27/2014.

What the dataviz above shows is that there was a 360% increase in the number of scientific articles published about Ebola in 2014 compared to 2013. This significant uptick shows that the scientific community clearly responded to the mass media attention about the epidemic. Of course we want science to be relevant, but are there really that many Ebola experts out there and that much new Ebola science to discuss to account for the uptick? Or is the uptick just a result of newsjacking Ebola in order to seem relevant?

Hospitals account for majority of health care-related debt

According to a recent 2013 report, health care-related debt is the leading type of consumer debt collected. Student loan debt is the next most common consumer debt collected, followed by credit card debt. The visualization below takes a look at the major types of debt contributing to the high levels of health care-related debt. Debt owed to hospitals accounts for 70.2% of the health care-related consumer debt, followed by debt owed to physician groups (18.7%) and clinics (11.1%).

Types of health care debt collected in 2013

Horizontal stacked bar chart showing the percent of debt due to hospitals, physicians and clinics

Data Source: Table 7.  The Impact of Third-Party Debt Collection on the U.S. National and State Economies in 2013. Prepared for ACA International July 2014.