__STYLES__

D.C. Hospitals: not the best in sight, but 6 in 10 patients think they're alright!

Tools used in this project
D.C. Hospitals: not the best in sight, but 6 in 10 patients think they're alright!

About this project

This is my exhibition entry for the Maven Healthcare challenge. I'll get more detailed about it when I get some free time, but the short version below gives a general overview.

Purpose: Patient satisfaction is an important metric that hospitals continually strive to enhance. This project aims to understand the current state of patient satisfaction in D.C. hospitals, identify the key drivers behind these scores, and suggest actionable strategies to elevate patient experiences.

Analytical approach & methodology

1. Data acquisition

Data comes from the HCAHPS survey and includes satisfaction scores across various states between 2015 and 2023.

2. Exploratory data analysis (EDA)

First, I used scatter plots to get an initial understanding of the overall hospital satisfaction against the willingness to recommend on state level. This helped me identify D.C. as a pronounced outlier with distinct characteristics and got me interested in making a deeper analysis.

3. Regression analysis

To understand the drivers behind the satisfaction scores, I conducted a regression analysis which helped me identify key areas such as are Care transition, Discharge information, and Cleanliness of hospital as primary drivers of overall patient satisfaction.

4. Comparative analysis

Then I benchmarked D.C.'s performance against national averages using various visualization techniques. This helped in pinpointing areas where D.C. excelled and areas that needed attention.

5. Playbook (the path forward)

The last section summarizes the findings and offers a core strategy how D.C.hospitals can work to improve patient satisfaction.

Discussion and feedback(3 comments)
comment-338-avatar
Enrique Ruiz
Enrique Ruiz
8 months ago
This is excellent! Loving your work with regression analysis lately. One minor suggestion worth trying out is to align the order of the measures in the "comparison or changes" chart with the

comment-341-avatar
Jasmin Jusufbegovic
Jasmin Jusufbegovic
Project owner
8 months ago
Project owner
Thank you Enrique! That's a great suggestion. The dot plot is currently sorted by the absolute difference for D.C., ascending order. Roughly, -6, -4, 0, 0, 0, 1, 3, 3.

comment-345-avatar
Chukwuemeka Chilaka
Chukwuemeka Chilaka
8 months ago
I am glad I saw your post on Linkedin, great analysis. I hope you have more time to share other insights you gained from working on this data set. I had battled with which plot pattern to vi
2000 characters remaining