Project Overview

Patient satisfaction is the foremost factor that influences hospital business across the globe. A prominent hospital chain in US was faced with the challenge of declining patient satisfaction and hence wanted to use the power of machine learning to understand what patients has to say about the facility. Pearl helped to enable opportunities of business intelligence and innovation through the effective use of advanced analytics to understand the root cause of the problem and created a mechanism for monitoring the same and hence enabled timely corrections to maximise customer satisfaction

Business Scenario

With increasing number of facilities operating under the hospital chain, standardisation and effective tracking of operations became increasingly challenging. With patient satisfaction showing an alarming fall thus leading to higher attrition, management wanted to understand the root cause that is leading to patient dissatisfaction and address the same

Business Requirements

Comments and reviews shared by patients across various platforms like social media, hospital websites, in patient review forms etc were collected and consolidated. The management wanted to analyse these comments to understand the major areas of concerns and develop a strategy to address those concerns. Also create a standard monitoring mechanism to keep a track of all patient comments across platforms and generate alerts in case of any issues spotted.

Challenges

• Mining unstructured data from various online platforms
• Duplicated comments across platforms
• Identifying fake comments/reviews

Solution

Patient comments were collected via conventional & unconventional methods like forms/emails, social media websites, blogs, review websites etc, these comments were filtered for legitimacy, clarity and relevance and then subjected to different techniques of Natural language processing. The results of these analysis were converted to scores that could be visualized using dashboards.

These dashboards helped management to get broad sense of how each facility/ department is currently performing with respect to customer satisfaction.Word clouds were generated to get a broad sense of what most patients are talking about the facility and how they associate positivity or negativity with these entities.Also, Sankey diagrams were generated to highlight the pressing issues in the facility and also to identify best practices.

Results

Areas that needed immediate attention were identified and changes were implemented in the operations hence leading to increased customer satisfaction. Weekly monitoring of the facility operations was streamlined, and best practices were identified across business and standardized.