Prepared By: Amanda Briggs, Laura Jahen, & Audrey Lopez-Valdez
Introduction & Problem Statement
For this project, we utilized state-level mortality data from the Centers for Disease Control Prevention (CDC), as well as Medicare spending data from the Center for Medicare Services (CMS) and Kaiser Family Foundation. The focus of our analysis was on seniors aged 65 and older, and we specifically looked at mortality rates within this age group. The primary health conditions we analyzed were Diabetes, Hypertensive Disease, and Influenza and Pneumonia. Our data sets, which were sourced from multiple government and healthcare organizations, required significant aggregation and cleaning before they could be used for analysis. This cleaning process included re-coding invalid values, handling missing data, as well as standardizing categories to ensure consistency among the data sets. By preparing the data in this way, we were able to conduct statistical analysis and create visualizations that highlight key trends and relationships between mortality rates and healthcare expenditures. These findings are discussed in more details further down in our report.
Based on our analysis, our organization believes that we should increase funding to Medicare programs, particularly for our senior population. With this project, we set out to explore whether a correlation between mortality rates and health care expenditures in the United States existed. Based on the data we found, we recommend changes to public health policy, particularly around senior health and senior preventative health programs. Diabetes, hypertensive disease, and influenza and pneumonia are all diseases that can be minimized with proper preventative health which can be addressed with additional funding. Our organization especially emphasizes programming for diabetes and hypertensive disease because these are chronic conditions that will continue to drain funding if the conditions worsen.
Methods
The final results reported below are the product of three data sets that were cleaned and combined for analysis. The first contains state-level information on enrollment, utilization and spending for Medicare parts A and B. This dataset is from 2021 and provides insights into Medicare spending for individuals aged 65 and older across different states. This dataset provides critical information about Medicare enrollment and spending. The key actions necessary to clean this data are the combination of the spending Medicare A and B, as we are interested in total spending by the State, and ensuring the Total Spending variable is aligned with general knowledge of currency illustrations. The second data set is from the CDC and provides information from all 50 states in the year 2021 regarding the number of deaths by age group and underlying cause of death. This dataset helps inform us of the disease specific mortality rates and the disease specific death rates by state and age. This data was cleaned to focus specifically on individuals 65 and older who’s cause of death was recorded with specific ICD codes related to Diabetes, hypertensive disease, and influenza and pneumonia. The final data set is provisional mortality statistics by multiple cause of death provided by the CDC from all death certificates in the fifty states and DC. The dates included in this dataset are the years of 2018 to the present. This dataset was cleaned to reflect the population of interest and the specific time period of interest, 2021.
We combined all three datasets and created new variables including a mortality rate for each State. Then we created three visualizations to help illustrate potential associations between the reports data elements including, State and Medicare spending.
State | Medicare Spending per Enrollee | Population over 65 | Total Deaths | Mortality Rate |
---|---|---|---|---|
ALABAMA | 10688 | 888817 | 3792 | 0.43 |
ALASKA | 9939 | 97663 | 254 | 0.26 |
ARIZONA | 10092 | 1333046 | 5295 | 0.40 |
ARKANSAS | 9919 | 528101 | 2427 | 0.46 |
CALIFORNIA | 12586 | 5957092 | 23452 | 0.39 |
COLORADO | 9453 | 879653 | 2237 | 0.25 |
CONNECTICUT | 12201 | 649235 | 1680 | 0.26 |
DELAWARE | 11419 | 201646 | 568 | 0.28 |
DISTRICT OF COLUMBIA | 11309 | 85838 | 397 | 0.46 |
FLORIDA | 12170 | 4598386 | 15009 | 0.33 |
GEORGIA | 10687 | 1584071 | 6459 | 0.41 |
HAWAII | 7472 | 282304 | 800 | 0.28 |
IDAHO | 8929 | 315456 | 1003 | 0.32 |
ILLINOIS | 11283 | 2101462 | 6824 | 0.32 |
INDIANA | 10757 | 1114688 | 4258 | 0.38 |
IOWA | 9854 | 565273 | 2110 | 0.37 |
KANSAS | 10533 | 489638 | 1954 | 0.40 |
KENTUCKY | 10121 | 770260 | 2927 | 0.38 |
LOUISIANA | 11650 | 761810 | 3102 | 0.41 |
MAINE | 9159 | 297165 | 1113 | 0.37 |
MARYLAND | 11899 | 1003157 | 3185 | 0.32 |
MASSACHUSETTS | 11661 | 1214693 | 2993 | 0.25 |
MICHIGAN | 10879 | 1822782 | 7179 | 0.39 |
MINNESOTA | 11489 | 955683 | 2873 | 0.30 |
MISSISSIPPI | 11428 | 494244 | 3155 | 0.64 |
MISSOURI | 10477 | 1083767 | 3446 | 0.32 |
MONTANA | 8649 | 216423 | 515 | 0.24 |
NEBRASKA | 10670 | 321890 | 1254 | 0.39 |
NEVADA | 11121 | 518467 | 2216 | 0.43 |
NEW HAMPSHIRE | 9369 | 267521 | 643 | 0.24 |
NEW JERSEY | 12394 | 1565917 | 4235 | 0.27 |
NEW MEXICO | 8665 | 391946 | 1235 | 0.32 |
NEW YORK | 13139 | 3477721 | 13206 | 0.38 |
NORTH CAROLINA | 10106 | 1793314 | 6170 | 0.34 |
NORTH DAKOTA | 11304 | 124641 | 436 | 0.35 |
OHIO | 10630 | 2098999 | 8008 | 0.38 |
OKLAHOMA | 11333 | 644711 | 5433 | 0.84 |
OREGON | 8610 | 788379 | 2481 | 0.31 |
PENNSYLVANIA | 10591 | 2464454 | 7656 | 0.31 |
RHODE ISLAND | 9737 | 198791 | 577 | 0.29 |
SOUTH CAROLINA | 10257 | 966399 | 3444 | 0.36 |
SOUTH DAKOTA | 11399 | 156418 | 565 | 0.36 |
TENNESSEE | 9889 | 1185272 | 5860 | 0.49 |
TEXAS | 11964 | 3875984 | 14633 | 0.38 |
UTAH | 9775 | 389148 | 1281 | 0.33 |
VERMONT | 9206 | 133258 | 406 | 0.30 |
VIRGINIA | 9475 | 1406652 | 4376 | 0.31 |
WASHINGTON | 8737 | 1255178 | 4336 | 0.35 |
WEST VIRGINIA | 10508 | 369420 | 1782 | 0.48 |
WISCONSIN | 10212 | 1057243 | 3158 | 0.30 |
WYOMING | 9975 | 103877 | 321 | 0.31 |
This table organizes our final data by State in the year 2021, and
displays the total amount of funding spent on Medicare per enrollee in
the State, the total population above 65 years, the total number of
deaths, and the calculated mortality rate for that year. The mortality
rate ranges from 0.238 in Montana to 0.843 in Oklahoma. You can also see
in the table that spending per enrolle ranges from $7472 in Hawaii to
$13,138 in New York.
This bar chart highlights states with high mortality rates and low
Medicare spending. North Carolina, Tennessee, and Arizona have
particularly high mortality rates, exceeding 6,000 deaths per 100,000
population, while states like Wisconsin and South Carolina also exhibit
concerning rates.
Discussion
The results from our analysis emphasize the need for increased public healthcare spending. Our analysis revealed variability in mortality rates and healthcare spending across the 50 states. Notably—Arizona, North Carolina, Missouri, Tennessee, South Carolina, Washington, Wisconsin, and Virginia—show a concerning trend of higher-than-average mortality rates along with lower Medicare spending per enrollee.
Diabetes and hypertensive disease were found to be key contributors to preventable deaths in the 65+ age group. These chronic conditions, along with infectious diseases like influenza and pneumonia, are often manageable or preventable with adequate access to preventive healthcare services.
Our analysis indicates that states with higher Medicare spending demonstrate lower mortality rates, highlighting the need for increased investment in healthcare so that it can lead to better health outcomes. For instance, states that allocate more funds to Medicare programs typically have more resources allocated to senior health initiatives. Conversely, states like those mentioned above may struggle to implement such programs effectively due to lack of funding.
These findings strongly support the recommendation to increase Medicare funding, particularly for states with high mortality and low spending. While our analysis provides valuable insights, limitations must be noted. The data is limited to 2021 and may not capture broader trends or the long-term effects of healthcare spending. Additionally, our analysis does not account for other determinants of health, such as socioeconomic factors or access to private healthcare. Further research can address these gaps to provide a more comprehensive understanding of the interplay between healthcare spending and mortality rates.