Data Visualization

Knowledge Activity: Orientation to Data Visualization IV (Informatics)

Prerequisites

1. Use of Microsoft Excel® is required to complete this activity
2. This activity is the fourth activity in a sequential 5-activity series. Completion of the activities below is required to successfully complete this activity:
   • Orientation to Data Analytics I
   • Applied Data Analytics II
   • Applied Data Analytics III

Student instructions

1. If you have questions about this activity, please contact your instructor for assistance.
2. You will review the de-identified chart that accompanies this activity. Your instructor has provided you with a link to the Orientation to Data Visualization IV (Informatics) activity. Click on 2: Launch EHR to review the patient chart and begin this activity.
3. Refer to the patient chart and any suggested resources to complete this activity.
4. Document your answers directly on this activity document as you complete the activity. When you are finished, you will save this activity document to your device and upload this activity document with your answers to your Learning Management System (LMS).

Introduction
The goal of data visualization is to extract meaningful insights from data to help with identifying trends, making decisions, and driving change. (Meyer, 2017). In the healthcare field, this may involve analyzing population health data, operational data (like claims and costs), performance data (like length of stay), and other patient health data. With the growth of EHRs, more and more data are becoming available, and the importance of accurately interpreting results is concurrently on the rise.
Methods for data visualization vary from summary tables, to graphs and charts, to sophisticated storyboards and dashboards. Regardless of the type of visual, it’s important to know the target audience and what information they value – then summarize it as clearly and concisely as possible. In general, simpler is better. (Bresnick). This activity and subsequent data visualization activities in EHR Go will use Microsoft Excel®. There are many other software options available and you are encouraged to experiment with using additional tools.
Another important aspect of data visualization is accuracy so as not to mislead the audience with how the data are presented. Ensuring your data set is comprehensive, accurate, and unbiased is the first step before beginning to analyze it. Understanding the difference between correlation and causation is critical when drawing conclusions. Correlation is a statistic that describes the relationship of two or more variables. Two sets of data may be strongly correlated but a change in one may not be the cause of a change in the other. Whereas with causation, one event is the result of the occurrence of the other event. (Statistics). A popular example of this in healthcare is that smoking is correlated with alcoholism but is not the cause of alcoholism. However, smoking does cause an increase in the risk of developing lung cancer. (iPerceptions, 2016). If you’re not careful, it’s easy for correlation to confound the results of a study.
The activity
Review the de-identified EHR that accompanies this activity under 2: Launch EHR and answer the following questions.
Questions

1. Which diagnosis(es) caused the admission? Explain.
2. Which diagnosis(es) are correlated with the admission? Explain.
   One of the patient’s problems is Coagulopathy due to NOAC (Xarelto). Review the definitions of causation and correlation, then answer the following questions.
3. Many patients who are prescribed Xarelto also have coagulopathy. Does this sentence prove causation or correlation? Explain
4. Patients who are taking Xarelto will develop coagulopathy within two weeks of starting the medication. Does this sentence prove causation or correlation? Explain
   Review the patient’s problem list and answer the following questions.
5. What was the listed cause of death?
6. Which category does this patient’s death fall into from the CDC age-adjusted leading causes of death? See the bar chart below.
   Microsoft Excel® for data visualization
   Using the leading causes of death data set provided from the Centers for Disease Control and Prevention (CDC) Data Visualization Gallery website, replicate the chart pictured below to present the results. Then explore other ways to view the data and draw conclusions.
   In the following steps, the bar chart shown below will be re-created using Excel. Follow each step carefully. (Tejada-Vera B, 2019).
   Open the resource CDC Leading Causes of Death Data Set Excerpt (found under 1: Overview & Resources along with this activity document) in Excel. A data set with 5 columns should appear as pictured below.

Making pivot tables
Click and drag to select the 5 columns of data.

Select Insert|PivotTable. Leave the default settings at the dialog box and select OK.

A new worksheet will open to begin the pivot table process.

Use the pivot table to summarize the age-adjusted death rate for each type for the year 2017. To do so:
• Click and drag the ‘Year’ field to the Filters area. This will allow each year to display.
• Click and drag the ‘Cause Name’ field to the Rows area. This will list each cause as a row in the table.
• Click and drag the ‘Age-adjusted Death Rate’ field to the ∑ Values area. It should appear as ‘Sum of Age-adjusted Date Rate’. This will display the total number of deaths for each cause.

• Next, specify the year in the resulting table. Select the dropdown arrow in cell B1 where it currently states ‘All’. Then check the Select Multiple Items box and specify only 2017. Then OK. The data table will automatically update to only show the number of deaths for the year 2017.

• Remove the data listing for ‘All Causes’ so the table only displays the specific causes of death. Select the dropdown arrow in cell A3 where it states ‘Row Labels’ and uncheck the ‘All Causes’ option followed by OK.

To create a similar bar graph as displayed by the CDC, the table needs to be sorted to display the lowest number first, instead of alphabetically.
• Select the death rate for Alzheimer’s (cell B4). Right click and select Sort followed by Smallest to Largest.

The resulting data table should look like this:

Making bar graphs
Select any cell in the pivot table then select ‘PivotTable Analyze’ in the top, center of the Excel window. Choose PivotChart. Select the Bar option then OK.

The chart will be inserted.
• Move the chart on the page and make it larger (click and drag one of the corners) to see all the data.

• Adjust the chart to make it more visually appealing.
• Click on the small legend box on the right where it says ‘Total’ and delete it.

• Next, click on the title field where it currently says ‘Total’ right above the bar graph. Type in the new title, “Age-adjusted Death Rates for the 10 Leading Causes of Death: United States, 2017”.

The chart is showing the 10 leading causes. Remove All causes.
• Select the ‘Cause Name’ dropdown and uncheck the following causes: All causes… Then OK. The chart will automatically update.

• Add some color. Right click on any of the existing bars in the chart and select Format Data Series…

• Select the Fill & Line icon (emptying paint bucket) then select the Vary colors by point option. Hold off on closing this window.

• Also use the Format Data Series options to change the thickness of each bar. Select the Series Options (column chart) icon then decrease the Gap Width field to about 10%. Close the window by clicking the small ‘x’ in the upper right. The chart will update. (Note: the colors won’t be the same as the CDC display – that’s okay).

• Next, scale the x-axis. Double-click on any of the x-axis values to open the Format Axis settings. Change the Maximum Bounds to 250. In the Number section, select ‘Number’ from the Category dropdown and select 1 for the Decimal places

• Finally, add a label to the x-axis. Click anywhere on the chart then select the Design menu at the top of the screen. Select Add Chart Element then Axis Titles and Primary Horizontal. Click on the new text box that currently says Axis Title and enter the new title “Rate per 100,000 U.S. standard population”.

• Now that the chart is complete, right click on the white space above the chart, just above the title, and choose Copy. Paste the chart into a Word document or PowerPoint and it will appear without the pivot data fields.

Communicating the data
One of the most important parts of data visualization is being able to clearly communicate the findings, both visually and through text.
Questions

7. Paste the bar graph you created here.
8. Describe the bar graph in 50 words or less.