Going Analytical for 2020: My First Data Project
2020 always sounded like a really cool futuristic year to me while growing up. Like “grab your hoverboard, let’s fly to the virtual 3d movie theater on the skyway in our shiny futuristic garb.”
Instead, 2020 started with me losing the sister that I grew up with, Karma. We were inseparable as children, but lost touch for many adult years. She was the most creative and vibrant individual that I have ever known. Sadly, I spent very little time with her during her last dozen years.
I was a retail manager arriving at work when I got the news that Karma was gone. Losing Karma changed the way I viewed life.
A few weeks later, I left retail for good and started a Data Analytics course online.
My sister, Karma, certified as the “Coolest Chick ya ever met!” by many people lucky enough to have met her.I had spent most of my life breaking down everything into charts and spreadsheets. Scrawling sports statistics in tiny chicken scratch handwriting as a kid and then invented weekly music charts as a teenager.
As a young adult, I analyzed my emotional world writing and editing poetry…recording it all into an endless Table of Contents like an ancient Egyptian scribe.
Analyzing data to garner insights was my favorite part of being a retail manager, so I decided that I wanted to pursue a Data Analytics career.
My first Data Analytics project for CareerFoundry involved delving into statistics from the CDC regarding influenza deaths in the United States from 2009 until 2017.
Over 50% of all
Influenza Deaths
are from 8 States!
1. California~~~~~~110,710
2. New York~~~~~~83,985
3. Texas~~~~~~~~~~56,514
4. Pennsylvania~~~47,178
5. Florida~~~~~~~~~46,764
6. Illinois~~~~~~~~~42,448
7. Ohio~~~~~~~~~~~40,386
8. North Carolina~33,724
Fig. 1: Total number of deaths from influenza from 2009-2017 in the United States.I started this project just as the world was shutting down due to the Covid-19 pandemic back in 2020.
I was to act as a medical staffing agency that sends temporary workers to clinics and hospitals strategically.
The first statistic that jumped out at me was finding that over half of all influenza deaths for the United States were in only 8 states total.
Fig. 2: The total number of influenza deaths from 2009-2017 by month of the year in the United States.The next step was looking at influenza deaths by month. Clearly, the months of December, January, February, and March are the months to focus extra staffing on.
One extremely important factor to consider is each state’s vulnerable population. The vulnerable population is defined as over-65, under-5, pregnant women, individuals with HIV/AIDs, cancer, heart disease, stroke, diabetes, asthma, and children with neurological disorders. The CDC estimates that adults over 65 account for 90% of all flu-related deaths (cdc.org).
Fig. 3: Percentage of each state in terms of population considered “vulnerable” which mainly consists of adults over 65 and children under 5.
Fig. 4: Looking at high need, medium need and low need states based on the amount of vulnerable population each state has.I divided all states into three sections: high need, medium need and low need. The high need states all have higher vulnerable populations while the low need states have far less vulnerable populations.
It is no surprise to see states like Florida, West Virginia and Maine near the top. Clearly, the biggest determinate of vulnerable population is senior citizen population.
According to U.S. Census Bureau, senior citizens (65 and over) make up 21% of Florida, 20% of Maine, and 19.5% of West Virginia. The national average is 16.5%.
Utah and Alaska are amongst the lowest with Utah at 10.8% and Alaska at 11.1% of senior citizens as part of total population.
Fig. 5: Average number of patients per provider by state from 2009-2017. The pink bar in the middle represents the rates from 90% to 110%. States #1-#13 are considered "understaffed", States #14-20 are "properly staffed" and States #21-51 are "overstaffed."It is important to study what each state already has in place in terms of staffing for the influenza season. How many patients does each provider have to take care of by average?
Something interesting to note is that many of the “understaffed” states actually have some of the lowest rates of “vulnerable population”. Washington D.C and Colorado may have the lowest levels of staffing, but they also have some of the smallest percentage of vulnerable population.
The only two states that have both a high vulnerable population (high need) and a high patients per provider number (understaffed) are Iowa and Arizona.
Fig. 6: A chart showing the number of influenza deaths by state against the population of every state. Please note that this chart does not include the eight states with the highest number of influenza deaths. Moving on to the rate of influenza deaths by population, the line in Fig. 4 represents the average rate of influenza deaths by population. The states located to the right of the line all have higher than average influenza deaths by population.
States such as Tennessee, Missouri, Alabama, and Kentucky have a higher influenza rate while Georgia, New Jersey, Washington
and Arizona all have a lower influenza rate.
So in conclusion…
There are a multitude of factors to consider when it comes to extra staffing in preparation for influenza season…
The focus should be on the months of December, January, February and March.
The Top 8 states for total influenza deaths along with the states that have a higher influenza death rate like Tennessee and Missouri.
States with the highest vulnerable population such as Florida, West Virginia and Maine.
States that have both a high vulnerable population and lower staffing such as Iowa and Arizona.
This project showed to me the importance of looking at every factor in making informed and efficient decisions….especially when it comes to the loss of human life.
References
Centers for Disease Control and Prevention (CDC): https://wonder.cdc.gov/ucd-icd10.html
Washington State Department of Health: https://doh.wa.gov/you-and-your-family/illness-and-disease-z/flu/are-you-high-risk-flu
US Census Bureau Population data: https://images.careerfoundry.com/public/courses/data-immersion/A1-A2_Influenza_Project/Census_Population.xlsx
