About this project
Introduction
The report was made for Maven Analytics’ Healthcare challenge.
Participants were tasked to play the role of a Data Analyst for the American Hospital Association (AHA), a national organization that represents hospitals and their patients, and acts as a source of information on health care issues and trends.
The main objective of the challenge was to analyze the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey results for the past 9 years and evaluate whether it has been successful in creating incentives for hospitals to improve their quality of care.
Dataset
All of the data used in this challenge were provided by Maven Analytics. It comprised 7 CSV tables, namely: measures, national_results, questions, reports, responses, state_results, states. A separate CSV table was also provided which serves as a dictionary for the description of each column in each of the 7 CSV tables.
Layout
The primary concept of the report is to showcase the trends and impacts of the HCAHPS survey over the years.
Given that the analysis revolves around survey results in the U.S.A., I opted to use the country's flag colors and made slight adjustments to enhance the report's visual appeal.
Analysis
I chose to focus solely on respondents who selected the top-box answers in each survey measure. My goal was to determine if the number or proportion of these respondents increased or decreased over the past 9 years. Furthermore, I adopted a top-down approach (starting from the most general to the most specific measures) in analyzing the survey results.
Starting from the most general:
Hospitals' Overall Rating
I was able to discover that the average proportion of respondents who opted for the top-box answers when evaluating the overall rating of hospitals on a national level experienced an upward trend since 2014 but declined from 2021 onwards. I used a line chart to visualize this insight in order to give the audience a clearer view of the trend.
On the other hand, I also wanted to highlight the U.S.A.'s average proportion of respondents who selected the top-box answers for the given measure. Hence, the filled map visualization. And to dig deeper, I decided to also identify which states were above, match, or fell below the national level average.
Respondents' Willingness to Recommend the Hospital
This is another measure of great importance as it correlates with hospitals' overall ratings. In this part of the report, I decided to categorize the respondents to provide the audience a more descriptive analysis. It revealed that the proportion of Promoters declined just like the overall rating of hospitals on a national level from 2014 to 2022.
A stacked column chart was employed for this analysis to better visualize the trend and proportion of respondents for this measure.
To the more specific:
Specific Patient Experience Measures
I delved deeper and excluded the two measures mentioned earlier from the other measures that are available in the data.
I applied the same analysis, which is the trend and proportion of respondents who selected the top-box answer for each measure for the past 9 years. The insights were the same, most of the measures followed the trends of both the hospitals' overall rating and the respondents' willingness to recommend.
A matrix is the most suitable visualization for the analysis since it is composed of many categories of data.
Note:
I only used the free version of Power BI desktop and do not have the access to publish the report online.
