Case Study – Flu Risk Forecasting

Project Overview

Background


Influenza is a contagious respiratory illness caused by viruses that infect the nose, throat, and lungs. Symptoms can range from mild to severe, and influenza may lead to serious complications or even death — particularly among vulnerable populations. According to the Centers for Disease Control and Prevention (CDC), these groups include adults over 65, children under five, pregnant women, and individuals with chronic health conditions such as asthma, heart disease, or diabetes.

The CDC recommends annual vaccination as the most effective way to reduce the risk of infection, particularly for those in high-risk groups.

In the United States, flu cases surge annually, placing additional pressure on healthcare systems due to an increase in hospitalisations. To meet this seasonal demand, hospitals and clinics often rely on medical staffing agencies to provide temporary healthcare professionals and ensure adequate coverage.

Scope


This analysis is designed to inform a medical staffing agency’s staffing plan for the upcoming 2018/19 influenza season, across all 50 U.S. states. The study focuses on age as the primary factor influencing vulnerability.

Goals


To support strategic staffing decisions by using age to identify high-risk population groups and analysing seasonal trends. Using these insights to inform geographic risk allocation and enable forecasting to guide when, where, and how many medical personnel to deploy.

Objectives


  1. Confirm whether age affects vulnerability to influenza and assess the relationship between population size and mortality.

    The objective explores if certain age groups are affected by influenza and establishes the foundational link between demographics and flu-related deaths.

  2. Identify when influenza season typically occurs, and how it varies in terms of severity each year.

    This section focuses on the timing and fluctuation of flu season across the United States, identifying patterns in onset and severity to support better planning.

  3. Translate demographic and seasonal patterns into state-level risk profiles and produce forecasts to guide when, where, and how many staff members should be deployed.

    Building on earlier insights, this objective applies risk classification by state and integrates it with seasonal timing to inform strategic medical staffing decisions.

Key Questions


  1. Which age groups are most affected by influenza?
  2. Is there a relationship between population size and influenza-related deaths?
  3. When does influenza season typically start and end?
  4. How does the severity of influenza vary from year to year?
  5. Which states are most at risk?
  6. Based on risk, when, where, and how many staff should be deployed?

Methodology


CDC datasets were cleaned and prepared in Excel, with suppressed values imputed where necessary. Statistical analysis and hypothesis testing were conducted to examine the relationship between population size and flu-related mortality among older adults.

Visual exploration in Tableau supported the identification of vulnerable groups, seasonal flu patterns, and regional variation. Risk was allocated manually based on the total number of deaths among individuals aged 65 and over in 2017—the most recent and relevant year for planning purposes. This risk profile was used to forecast potential pressure on healthcare systems and inform recommendations on when and where additional medical staff should be deployed.

Data


This analysis uses publicly available data sourced from the Centres for Disease Control and Prevention (CDC) and the US Census Bureau that cover the period 2009 to 2017. The data was provided by CareerFoundry as part of their Data Analytics Course.


GitHub Repository
Tableau Storyboard