Bilirubin's Role in Sepsis Risk with Sunburst Visuals

Bilirubin is a yellow substance made when red blood cells break down. In our bodies, bilirubin is a substance that can indicate how well our liver is working. This article will use a Sunburst chart to show how measuring Total Bilirubin can help detect sepsis early.

Understanding Total Bilirubin in Human Body : Total bilirubin is the sum of all bilirubin in the blood, including both unconjugated (indirect) and conjugated (direct) bilirubin.

The normal ranges are as follows:

• Total bilirubin: Normally between 0.3 and 1.2 milligrams per deciliter (mg/dL).

• Indirect bilirubin: Normally between 0.2 and 0.8 mg/dL.

• Direct bilirubin: Normally between 0.0 and 0.3 mg/dL.
  

If bilirubin levels are higher than 1.2 mg/dL, it's considered high. Levels above 12 mg/dL are very serious and may indicate a severe health issue that needs immediate attention.

Why It Matters: High bilirubin levels usually indicate liver problems, often due to inflammation or severe infections.

Mortality Risk: In sepsis, patients with total bilirubin levels of 5 mg/dL or higher have a higher risk of dying within a year compared to those with levels below 5 mg/dL. High bilirubin levels in sepsis can damage the liver and other organs like the brain and kidneys, leading to serious health issues.

Visualizing the Sepsis Dataset in Tableau

To start visualizing the Sepsis dataset in Tableau, go to the provided link and load the dataset. Once loaded, click on "Sheet" to begin creating your visualizations.

First, create a calculation field called "Centre" using the formula Makepoint(0,0) to set a central point. Drag and drop this calculation field 'Centre' onto your sheet. You can add the Centre field multiple times by dropping it onto the Marks Layer for multiple layers. Next, remove the Longitude field from Columns and the Latitude field from Rows.

To create a layered visualization, select all the layers and set them to display as 'Pie' charts under marks. Adjust the size incrementally and use different colors for each layer. Repeat these steps for each layer, increasing the size and using distinct colors to distinguish between different layers in your visualization.

In the innermost layer of your pie chart, start by creating a calculation field called Gender, which includes male and female. Once that's set up, drag Gender into the 'Colors' under small pie chart. Next, use 'Dimension' under the Gender dropdown. Then, take the 'patient id' and place it in the 'angle' section, selecting 'distinct count' from the dropdown menu. After that, move the distinct patient id to the label and choose 'percentage of total' from the quick table options. Don't forget to label the Gender category clearly for better understanding. Finally, arrange all the labels neatly within the innermost pie chart by dragging them to their appropriate places.

In the second layer of your data analysis, start by creating a new calculation that categorizes patients into three groups: Sepsis, Non-sepsis, and Onset sepsis. Once this is done, add this calculation to your second layer visualization. Next, drag the 'patient id' to the 'angle' and choose 'distinct count' from the options provided. Then, include Gender as a dimension and set it to color by selecting the detail button. Move both the Sepsis category and the distinct patient count with their percentage of the total to the label section of the visualization. Finally, arrange all the labels neatly within the second pie chart by dragging them into the correct positions.

In the outer layer, start by creating a calculation field for Bilirubin Total. The normal range should be between 0.3 to 1.2 mg/dL. If it's above 1.2 mg/dL, mark it as abnormal, and if it exceeds 12 mg/dL, it's considered severe. After setting this up, display Bilirubin Total using a pie chart by moving it colors under marks. Include 'patient id' by showing the distinct count as a percentage, similar to what was done earlier.

Next, focus on the Sepsis category, which is the main colored category from the previous layer placing it under detail, then change it colour. Label the Bilirubin total, Sepsis category and the percentage of patients visually. Make sure to arrange these labels properly in their respective places for clarity.

Bilirubin Total Analysis

The data shows how patients are categorized based on their total bilirubin levels, whether they have sepsis, and their gender. Here's what we found:

Bilirubin Levels:

• Most patients without sepsis have normal bilirubin levels (68.8%).

• Abnormal bilirubin levels are more common in patients without sepsis (23.2%) compared to those with onset sepsis (3.7%) and sepsis (0.4%).

• Severe bilirubin levels are rare across all sepsis categories, with the highest percentage in patients without sepsis (1.4%).
  

Sepsis Status:

• Patients with normal bilirubin levels make up the highest percentage across all sepsis statuses.

• There's a small percentage of patients with severe bilirubin levels in all sepsis categories, lowest in those with sepsis (0.1%).
  

Gender Distribution:

• More males are in the group without sepsis (49.9%) compared to females (38.8%).

• In the onset sepsis and sepsis categories, males also outnumber females, with males making up 6.0% and 0.7%, respectively.

• Overall, there are more male patients (56.6%) than female patients (43.4%).
  

This analysis shows that most patients without sepsis have normal bilirubin levels, and severe bilirubin levels are uncommon across all sepsis statuses. It also reveals a gender difference, with more males affected across all groups.