Flu Season Staffing Plan


OVERVIEW

A medical staffing agency requested a plan to help prepare for flu season.

OBJECTIVE

Determine how many temporary staff to send to each state and when to send them.

ASSUMPTIONS

States with higher numbers of vulnerable people will have a higher need for temporary staff.

DATA

  • US Census Bureau
  • US Centers for Disease Control and Prevention

SKILLS

  • Translating Business Requirements
  • Data Cleaning, Transformation, and Integration
  • Statistical Hypothesis Testing
  • Forecasting
  • Visual Storytelling

TOOLS

  • Excel
  • Tableau

CHALLENGES

From 2010 to 2016, between 23 and 39 million people suffered from the flu, and over 40,000 people died every year. People in the vulnerable population (children under 5 and adults over 60 years old) develop serious complications and end up in the hospital. Hospitals and clinics need additional staff to treat these patients. A medical staffing agency that provides temporary staff needs a plan to determine when and where to send staff. They are not able to hire more staff, so they need to use their existing resources.

ANALYSIS

One-tailed hypothesis testing determined vulnerable people are more likely to die of flu: 91% of all flu deaths in the US can be attributed to adults over 65. States with higher numbers of vulnerable populations have higher rates of the flu.


Linear regression illustrates the strong correlation (0.83) between flu deaths and vulnerable population.


Time series analysis illustrated the seasonality of flu data – flu season starts in September and peaks in January. It is the same in every state, and forecasts indicate the pattern will stay the same in the next year.


The staffing agency cannot hire additional staff, so determining where staff are most needed was vital to an effective plan. Since the flu season does not vary by state, it was important establish which states to prioritize since all states have high flu cases at the same time. Combining historical flu death data and vulnerable population – as opposed to one variable – allowed for a more accurate prioritization. This analysis revealed the top five priority states: Texas, California, Virginia, New York, and Florida (in that order).


RECOMMENDATIONS

Age, time of year, and location all play a role when considering where to send medical staff during flu season. Vulnerable populations are more likely to need medical attention as a result of the flu, but the key to creating a successful plan includes determining when and where to send staff. Examining multiple variables in each state established highest priority states (in order of priority), based on risk factors. When combined, it allows for a comprehensive staffing and monitoring plan.