First I looked at HIV Diagnoses per 100,000 individuals for each state. I coded the region of the state based off of the CDC’s categorization. I then color coded the state’s individual data point, plotting the number of HIV Diagnoses per 100,000 against the alphabetical index of the state.
Because of the District of Columbia’s high rate, I removed it from the dataset.
Now I could see that it appeared there was a difference in the HIV incidence in various regions, so I calculated the mean diagnoses rates. The South, even when removing the District of Columbia, still had substantially higher mean diagnoses rates per 100,000 people. Just from looking at the graph, I could see clustering of red points near the top of the graph and predominance of blue and black at the bottom of the graph. Visualizing it in this way showed me that there was difference between the regions. Because of this marked difference between the various regions, I grouped the states by region and showed the overall diagnoses and the diagnoses per 100,000 with a line representing the mean diagnoses rate.
Here we can clearly see that HIV is most prevalent in the South and least prevalent in the Midwest and West. This trend is true for both relative diagnoses rates and total diagnoses rates. After discovering that there are higher rates of HIV infection in the South and Northeast even when adjusted for population size, I wanted to see if these trends held in the past. Examining CDC data from 1981 to 2002, I created graphs that showed the number of HIV cases in the South and the Northeast have consistently been the highest in the nation, even when controlled for the population size.
From these graphs I saw that there were differences in the slopes before 1993 for the different regions and differences in the slopes after 1993. I then separated out each of the regions into individual graphs to see what the various slopes are. I subsetted the data from each region into pre- and post-1993 and ran OLS regressions for each region’s pre- and post-1993 data. I plotted line segments over the data to show the difference in how effective the post-1993 policies were in decreasing the incidence of HIV in each of the regions.
The South had the steepest HIV incidence rate between the years 1981 and 1993, followed by the Northeast, the Midwest, then the West. In the post-1993 period, it had the second-steepest decrease in annual HIV diagnoses relative to its population. Its rate of -0.223 people per 100000 per year was substantially lower than the Northeast’s -0.341 decrease, which suggests that the policies were more effective at reducing the transmission of HIV in the Northeast. The Midwest and West had substantially less steep slopes in the pre- and post-1993 period, which may suggest that the policy changes in 1993 were more effective for high-HIV-incidence areas than they were in the Midwest and West that had lower diagnoses rates before.
I then wanted to look at HIV diagnoses based on race. I looked at the overall number of people diagnosed with HIV by their racial background and then looked at the number of people per 100,000 per race.
These trends held over time with the total HIV Cases between 1981 and 2002 showing that, since 1995, Black individuals have consisted of the plurality of HIV diagnoses. Looking at the relative number of cases per 100,000 people, we see that the Black community has faced the highest number of diagnoses relative to its population size, followed by Hispanics then American Indians and White individuals.
If we look at the pre- and post-1993 periods, we see that there was substantial decline in the total number of White individuals diagnosed with HIV but a smaller dip for both the number of Black and Latino individuals. Looking at the total number of diagnoses, we see that White individuals had the steepest rise and the steepest decline in the pre- and post-1993 period, respectively. The steep decline post-1993 compared to the half as steep decline in the Black population in that same period explains why in 1995 Black people became the plurality of HIV diagnoses. These numbers look at total diagnoses since racial data from the intercensal period was unavailable from 1981-1990 (which would have allowed us to determine relative diagnoses rates) and these numbers reveal the overall public policy change in terms of societal health (reducing the numbers of individuals which translates directly into decreased health spending) instead of looking at each racial group’s relative change (which would say nothing on the net effect on society).
I then looked at the average diagnoses rate for the total diagnoses and the three most affected racial groups (Black/African American, Hispanic/Latino, and White individuals). The top four graphs look at the number of HIV Diagnoses per 100,000 people, while the bottom four look at the total number of cases of HIV in 2011. The averages for each region are plotted as horizontal lines which reveals that either the South or the Northeast has the highest rates of HIV Diagnoses per 100000 and in total across all racial groups.
If we examine this by region, we find that in the South, Black individuals make up the plurality of HIV diagnoses since 1993 and in the Northeast they are close to the number of White individuals, with the first time Black diagnoses being more common occurring around 1996.
Black people are the least prevalent in the West, but the Midwest’s high concentration of White individuals (who have the lowest rates of HIV diagnoses per 100,000 people of the three most affected racial or ethnic groups by total number of cases—Hispanics, Whites, and Latinos) contributes to the Midwest being the lowest HIV-prevalent area, though it is close to that of the West.
If we examine this by region, we find that in the South, Black individuals make up the plurality of HIV diagnoses since 1993 and in the Northeast they are close to the number of White individuals, with the first time Black diagnoses being more common occurring around 1996.
Since in no region besides the South are Black individuals the plurality of HIV diagnoses, then it seems that the South is driving the national trend in which Black individuals are the most affected by HIV diagnoses. I then looked at the racial composition of each of these regions to see how their racial composition could explain the overall trends of the region. The high proportion of Black people in the South also explains why the South has such high rates of HIV diagnoses. Since Black people are now the plurality of national annual HIV diagnoses, then the high concentration of Black people in the South would drive up the South’s overall rates of HIV diagnoses.
Black people are the least prevalent in the West, but the Midwest’s high concentration of White individuals (who have the lowest rates of HIV diagnoses per 100,000 people of the three most affected racial or ethnic groups by total number of cases—Hispanics, Whites, and Latinos) contributes to the Midwest being the lowest HIV-prevalent area, though it is close to that of the West.
Datasets Accessed
HIV/AIDS Data 1981-2002: Sexual Orientation, Region, Race
US Department of Health and Human Services (US DHHS), Centers for Disease Control and Prevention (CDC), National Center for HIV, STD and TB Prevention (NCHSTP), AIDS Public Information Data Set (APIDS) US Surveillance Data for 1981-2002, CDC WONDER On-line Database, December 2005.
Estimated HIV Rates 2011
Centers for Disease Control and Prevention, National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention (NCHHSTP) Atlas, accessed July 2013. http://kff.org/hivaids/state-indicator/estimated-rates-per-100000-of-hiv-diagnoses-adults-and-adolescents-by-raceethnicity/
Population 1990-2012
United States Department of Health and Human Services (US DHHS), Centers for Disease Control and Prevention (CDC), National Center for Health Statistics (NCHS), Bridged-Race Population Estimates, United States July 1st resident population by state, county, age, sex, bridged-race, and Hispanic origin, on CDC WONDER On-line Database. http://wonder.cdc.gov/bridged-race-population.html
No comments:
Post a Comment