Part-to-Whole Relationships

How do I determine if I should use a graph that displays part-to-whole relationships?

The simplest way to think about part-to-whole data is to determine if what you really are communicating is a ratio. And the best way to think about a ratio is usually as a percentage of something in relationship to the whole.

Consider the bag of Halloween candy. When I was a kid I would dump out my bag, organize the candy by chocolate (good) and non-chocolate (not so good). Depending on the ratio of chocolate to non-chocolate I would declare how successfully the evening had been.

Back on topic — a couple of examples of part-to-whole healthcare data are:

  • Payer Mix:
    • Commercial Insurance
    • Medicare
    • Medicaid
    • Workers Compensation
    • Self Pay
  • Medication Occurrences by Severity and Cause:
    • Severity
      1. no injury
      2. minor injury
      3. major injury
      4. permanent functional impairment or death
    • Cause
      1. incorrect prescription
      2. dispensing error
      3. transcription error
      4. administered incorrectly

Horizontal and vertical bar graphs are the best choices for displaying part-to-whole data. NOT a pie chart.

Click here to read more about displaying part-to-whole healthcare data in my newsletter “Unleash Your Inner Healthcare Data.”

Leave a Reply