Skip to content
The role of data visualization in healthcare

“A picture is worth a thousand words.” We’ve all heard this old adage, but in the age of big data, it takes on new meaning. Few people have the gift of simply looking at numbers and gleaning useful information. Data visualization brings clarity to the numbers, statistics, and measurements that make up healthcare today.

What is data visualization?

Data visualization is a visual representation of information. Using graphs, charts, tables and other imagery, data visualization helps put meaning to data points. Most importantly, data visualization is a tool that helps a person analyze and make decisions. It helps to understand that data visualization is one step in a larger process of data analytics, as shown below.

data visualization pathway

The idea of data visualization isn’t new. We still use the Mercator Projection to represent the world map—an innovation that came on the scene in 1569.[1] Industries as diverse as finance and manufacturing make great use of visualization, through stock charts, income statements, productivity graphs, and more.

How is data visualization used in healthcare?

In healthcare, most people are familiar with simple data visualizations, such as a pediatric growth chart. Measurements are plotted at each visit, establishing a trend line over time. Providers use the chart to determine whether the child is growing at an expected rate in an established range. If the child falls significantly outside the acceptable range, the visualization helps the provider make decisions on treatment and care moving forward.

However, much in healthcare is not as simple or straightforward as a child’s growth over time. It has been reported that healthcare organizations have seen an 878% health data growth rate since 2016.[2] For this reason among many, data visualization has exploded in the healthcare industry. The use of infographics, dashboards, and other visual analytics can make complex datasets easier to understand and act upon. Visualizations are used in everything from government projections of healthcare costs to the simple activity tracker on a consumer’s smartphone.

Data visualization examples

Data visualization is most meaningful when it answers a question or prompts decision-making and action. Typical visualizations might include:

  • Comparisons (“How is my blood pressure today compared to yesterday?”)
  • Counts or amounts (“How many hours of sleep did I get last night?”)
  • Trends/patterns (“Each year during the holiday season I gain weight.”)

For visualizations on an individual level, such as those recorded in the HealthIO app, the data can help a user and his or her provider decide the best course of action for optimal health.

Sample HealthIO data visualization for blood pressure reading over time.

For visualizations used at a macro level with de-identified data, the data can help government agencies, provider groups, payers, or self-insured employers understand the health trends of their respective populations.

For example, the Centers for Disease Control and Prevention published its “National Diabetes Statistics Report 2020”, replete with a variety of charts and graphics. The data visualization revealed that, “Among US children and adolescents aged 10-19 years…overall incidence of type 2 diabetes significantly increased.”[3]

Source: Centers for Disease Control and Prevention. (2020). National Diabetes Statistics Report, 2020. U.S. Department of Health and Human Services.

Another approach is to combine visualization with additional analytical approaches like categorization to evaluate trends and distribution of risks across a population as seen within the HealthIO example below.

Example of an area chart showing categorized data generated from the HealthIO platform.

In this area chart, a client’s population is categorized by their risk of hypertension. Category one represents the lowest risk with each subsequent category indicating a higher level of risk. By category six, some individuals in this population will have already been diagnosed with hypertension and tasked to manage their condition.

Using this type of data visualization can help clients see fluctuations in their population’s risk over time. In this example, categories five and six (those at highest risk of developing or having hypertension) peaked around February 15th, 2021.  Although these highest risk categories declined over the next two weeks, the trend line appears to be increasing again. With such trends visualized and identified, decision-makers are equipped with the information needed to plan appropriate interventions for the at-risk population.

Potential pitfalls of data visualization

While data visualization can provide much needed clarity, caution should be exercised. Given the overabundance of data points, it is possible for a visualization to become overly complicated. Poorly designed graphics can be hard to interpret or analyze. Data visualization should always start from the viewpoint of answering a question. The type of answer sought will drive the design of the graphic and help overcome these challenges.[4]

The potential for errors in data is an additional concern. Using bad data to create a visualization could lead to incorrect trends or indicators. Bad data can include duplicate data, missing data or data that is incorrectly tabulated. While data is highly dependent on its source and the method used to record it, steps can be taken to mitigate the risk. Risk mitigation can include automating the data acquisition, cleansing data, running automated data quality tests, and utilizing human oversight to validate system results where possible. It’s just as important to trust the data processor as it is to trust the data source when evaluating data visualizations.

Will data visualization lead to better health outcomes?

The prevalence of big data has put more information into the hands of consumers and organizations than ever before. As data visualization continues to evolve and improve, adding storytelling to the data can help highlight the most important aspects of the graphic representation.[5] While data visualization is not a motivator on its own, the insights it reveals can be a step on the path to eventual behavior change.

[1]  Farnworth, R. (July 3, 2020). Towards data science. Retrieved March 18, 2021, from https://towardsdatascience.com/a-short-history-of-data-visualisation-de2f81ed0b23

[2] Donovan, F. (May 8, 2019). Organizations see 878% health data growth rate since 2016. Retrieved May 24, 2021, from https://hitinfrastructure.com/news/organizations-see-878-health-data-growth-rate-since-2016

[3] Centers for Disease Control and Prevention. (2020). National Diabetes Statistics Report, 2020. U.S. Department of Health and Human Services. Retrieved March 18, 2021, from https://www.cdc.gov/diabetes/library/features/diabetes-stat-report.html

[4] Durcevic, S. (May 2, 2019). Designing charts and graphs: How to choose the right data visualization Types. The datapine Blog. Retrieved March 18, 2021, from https://www.datapine.com/blog/how-to-choose-the-right-data-visualization-types/

[5] Meyer, M. (December 21, 2017). The rise of healthcare data visualization. Journal of AHIMA. Retrieved March 18, 2021, from https://journal.ahima.org/the-rise-of-healthcare-data-visualization/

Stay current with HealthIO by joining our email list