A medical agency preparing for flu season. Using historical influenza and population data to determine when and where to deploy staff
Assumptions
-
​States with a higher vulnerable population aged 65+ will see a higher number of influenza related deaths
Tools
-
Excel
-
Tableau
Data
-
Influenza deaths
-
Population census
-
Source: US Census Bureau
-
Data: Data (Excel)
-
Skills
-
Data cleaning, integration & transformation
-
Statistical hypothesis testing
-
Visual analysis
-
Forecasting
-
Storytelling in Tableau
Analysis & Insights
The CDC & Census data sets were cleaned, merged, profiled, & transformed.
Conducted statistical analysis to examine data spread and test hypothesis.
Tableau visualisations were created to derive insights and form the basis of recommendations.
-
Strong positive correlation between vulnerable population and deaths
-
More deaths in states with higher vulnerable populations (aged 65 and over)
-
Run up in autumn and peak in winter months (Jan - March & Dec)
Conclusions
-
Confirmed hypothesis that states with a higher vulnerable population aged 65+ will see a higher number of influenza related deaths
Recommendations
-
Prioritise staffing across the 5 states with most influenza related deaths
-
California, NY, Texas, Pennsylvania & Florida
-
-
Timing of staffing to follow seasonality i.e., winter months and autumn run up
-
Further analysis into normalised deaths findings and Florida and NY as outliers in the correlation between the vulnerable population and number of deaths
-