Access to Contraceptive Services in Florida

Transcription

1 Access to Contraceptive Services in Florida Introduction This project aims to determine which Florida county has the least access to family planning services through Title X facilities. With data gathered from the Census and the State of Florida s Health system, I compared each county s access to Title X facilities based upon a combination of economic, social, situational, and health criteria. First, I created a sociodemographic profile for each county. Then, I created a profile for women s health by county, determining which criteria would define a county as in need, or less accessible. The results of this project concluded that Alachua County has the worst access, with over 16,000 women per facility. These facilities provide healthcare to women, ranging from birth control to abortions. Women s health is a topic that has become controversial in the past several decades. On the basis of politics, many American states and counties are losing funding for family planning services, which provide contraceptives, as well as screenings and testing for millions of women. Many of these facilities also perform abortions, and on political grounds, this has led to the decline of services. Title X facilities are those which provide discounted women s health services, and can range from the local health department to Planned Parenthoods.

2 Methodology Manual Analysis Filters The first phase of this project, titled Filter 1, narrowed down the counties most in need of facilities, based upon sociodemographic data. A five-mile buffer was created from urban areas throughout Florida. Urban populations likely have a greater need for these facilities, as they have a greater range within their sociodemographic context, as well as a larger population of women. Then I created a selection, starting with poverty rate. The average rate of poverty in Florida is estimated at 14%, so I selected the counties in Florida which intersected the urban buffer and had a poverty rate equal to or greater than 14%. Education data was available at a zip code level, rather than county, so I had to convert the data for this. Due to the nature of the data, which had clear natural breaks, I selected 37% as my criteria, because it narrowed my results significantly. I selected the counties which contained zip codes where over 37% of the population did not have a college education, and proceeded to select these counties from the previous selection. This concluded my results for Filter 1, which left 13 of 67 counties.

3 Top: poverty, bottom: college education & urban buffer, right: intersection The next phase, Filter 2, incorporated women and their access to healthcare. This filter began by looking at every county in Florida, without interaction from Filter 1 data. The counties in Filter 2 are intended to be those which have a high population of women between the age of 15-44, and a high percentage of uninsured women. I selected the counties which had a population of at least 10,000 women in the previously mentioned age range, because I believe this is a significant value. The second variable in this filter was an estimate of uninsured, childbearing women. For this measure, I used the percentage of women who selected self-pay as payment method for child delivery, which indicates that they do not have health insurance. In addition to population, I selected counties which also a percentage value for uninsured as over

4 8.3%. Again, this value was selected in context of the data and its natural breaks. The counties which met both of these criteria formed Filter 2, which left 9 of 67 counties. Top: % uninsured, bottom: population of women, right: intersection The final stage of my analysis was overlaying the counties from Filter 1 and Filter 2. Only three of these counties met the criteria set within both of these filters. These three counties, Alachua, Osceola, and Gadsden composed the final filter, which would be the final determinant of the county which has least access to contraceptive facilities. Finally, these counties were joined with the contraceptive facility data, which gave the quantity of Title X facilities in each Florida county. The final county was chosen based upon an accessibility ratio: number of facilities to the population of women at child-bearing age. The accessibility ratio left Alachua county as that with the least access to government-funded contraceptive services, with only one facility per 16,223 women. With 8 facilities total, in the county, this places a stress on these facilities as they may have to treat thousands of clients each, on limited government

5 funding. Alachua County is a primarily Democratic county, having voted for the Democratic candidate in the past 7 elections. This county also has a large population of college students, which may also contribute to the Democratic nature of the politics within Gainesville, although the more rural areas of Alachua county identify otherwise. Automated Analysis The methodology of my original project was extremely straightforward and linear, which allowed for a simple translation into Model Builder, so most of the steps were identical to those from the first version. As seen in the flow chart from the original project, there were three

6 distinct phases. One filter consisted of selections based on demographics, while another filter was based on women s health data. These filters produced a final selection through intersection, which was compared to a third filter, analyzing the number of facilities in each of these selected counties. Sub-Models To conduct this within Model Builder, I created 4 sub-models. The first sub-model consisted of Filter 1. The first step was to select zip codes where the poverty rate was at or above 12%. I intersected this with the shapefile of counties which contained zip codes where over 30% of the population did not have a college education. From here, I created a 2-mile buffer from urban areas, and selected the counties remaining which intersected this buffer. Unlike the first method, this left 18 remaining counties. I believe that the difference lies in my order of operation, since I did not conduct the selections simultaneously (in either version). The second sub-model analyzed women s health. A selection was made for counties where the female child-bearing population was over 10,000, and another selection for over 8.3% of uninsured births. These were intersected, which left 21 remaining counties. This particular section did not differ from my original analysis.

7 The third sub-model was the intersection of the first two models, which left 4 remaining counties, named Counties in Need. These three counties varied from the first version s, consisting of Broward, Orange, and Miami-Dade. Manually, I joined women population data with Title X location data, and returned to the attribute table to add a new field. This field, center_pop, calculated the female population in each county PER title X facility in that county. These were classified by three quantiles, with the third being over women per facility. A fourth sub-model was created, automating the selection of only counties within the high-value classification of facilities, which was then

8 intersected with Counties in Need. This model was redundant, due to the fact that the three counties in need were all in this classification. From this point, my results were manually calculated. Broward County, with only 9 Title X facilities, was limited to one facility per ever 80,444 women. This is over double the ratio for Orange County and Miami-Dade, and over 4 times the ratio for my previous result, Alachua County. Alternative Analysis Change in Parameters The first change alternative that I considered was the change in parameters. The focus of this project is Title X facilities, so I thought it would be a great opportunity to take a look at the different options which that particular options data provides. The field of center_pop, which calculated the population per TOTAL number of Title X facilities in each county, was manually created, but the data provides the quantity of each type of facility as well (Planned Parenthood, health departments, hospital-based, etc.). For this alternative, I chose to change the calculation from TOTAL to the population per HEALTH DEPARTMENT of each county. This did change the results, leaving Broward County with over 360,000 women per health department. This alternative was straight-forward, and could easily be changed to any of the other types of facilities.

9 Change in Methodology Due to the linear nature of my project, I had some difficulty creating an alternative for the change in methodology. Originally, I thought to remove the buffer from the Sociodemographic filter/sub-model, in order to change the analysis to focus on rural areas. While this did work in theory, I believe the accuracy was skewed due to the fact that the other parameters were tailored to focus on urban areas. However, once some of the parameters were changes, for example, the women s population (lowered to 5,000), I continued with the alternative analysis. I added the buffer to Sub-Model #3, but this time it was set to exclude counties that intersected a ten-mile buffer. There was a single county remaining, Hardee

10 County, which met all of the criteria. Hardee County has 3,700 women per facility (total of 3). Results This research has several limitations, specifically regarding the population of the counties. Primarily, the lack of population data for several counties may provide inaccurate results. Several counties were excluded because I could not obtain information for the childbearing population within that county. Another limitation is the location of residents within the county. For example, groups of people living near the border of another county may use the facilities of neighboring counties because they may be closer. This project also excluded rural areas, which may have more difficulty accessing these facilities to begin with, even if they are available. An additional limitation is that there was limited data regarding the client base of the facilities, so it is unknown whether or not there is a large population of women in this county who need contraceptives, but cannot access them. This project is meant to determine which county has the least access to contraceptive facilities, but if women are not using these facilities, it is difficult to determine whether or not they are needed.