Hospital Readmission tool

Hospital Readmission Dashboard Guide


Below is the link to access the tool via Tableau Public:

 Patient Readmission Risk tool

The dashboard consists of 4 graphs with two metrics: Combined Readmission Rate and Combined Hospital Time along the X and Y axis. Plotting along these axis’ are the aggregated instances of readmission filtered through the variables. These variables being: Gender, Diabetes, and Age range of the patient.

Patient hospital readmission rate. Tableau dashboard to predict hospital readmission rates based on several variables.



The first graph is Combined Gender and Readmission, which displays a vertical bar chart of the aggregated readmission rates per gender and a user selected ten-year age range.

 

The bars are blue (darker color) for male and orange (lighter color) for female. Filters located on the far right upper corner of dashboard, allow the user to dictate the combination of variables and watch the visualizations populate with a different data if the variables are connected. For example, the default setting for the filters will be set to “All”; on the Combined Gender and Readmission, the Combined Readmission Rate for males is 40.845%. When the Age(bin) filter is selected to 30, the Combined Readmission Rate changes to 40.208%, same with the Combined Hospital Time. The user can drill down in the data with these filters in 27 different combinations across all graphs (in which the variable apply).

 

The next graph, Gender, Age and Diabetes, displays four vertical bar charts side by side. The blue stack represents the WGU hospital dataset and the orange, the Kaggle dataset. When the user hovers the cursor over the bar graphs, the following will populate on from the cursor tip: Kaggle dataset (representing “True” for the Kaggle set or “False” for the WGU set), Gender, Time in Hospital (days), and Combined Readmission Rate. changing the inputs of the filters on the right-hand side of the dashboard will change the data visualizations accordingly.

The next graph is Combined Diabetes and Age, which is a scatterplot chart located in the lower left-hand corner this chart has a line best fit or trend line drawn through it when the cursor is hovered over one of the residual points it will populate a pop-up box indicating whether the patient has diabetes, average time in hospital, and Combined Readmission Rate.

 The last graph, Combined Age, Diabetes and Readmission, takes the instances of patients with diabetes and bins them by age. Also, the graphs is a dual axis against the Combined Non-Readmission Rate. There is a total of four trendlines, one for non-readmitted patients and the other for readmitted patients. Within that, if the patients had diabetes. The dual axis effect creates an X mark around 37 days and there is a divergent path of residuals. when the cursor is hovered over one of the residuals what will be displayed is the age band of the patient as well as the Combined Readmission Rate. if the residual reflects a patient that did not re-admit, the tooltip will only reflect the binned age of the patient but no readmission rate. When the cursor is hovered over one of the trendlines, the linear formula populates.

 

Michael Segaline

A Data Scientist and Search Engine Optimization Expert.

https://www.bloomingbiz.marketing
Previous
Previous

Logistic Regression in Python

Next
Next

Data Analytics and Police Funding