The Impact of Predictive Analytics and CI on Patient Satisfaction

Patient Satisfaction vs. Patient Experience: What's the Difference?

Brainstorming Patient Satisfaction Contributors

If you find yourself in the curiosity stage, where you want to explore statistics without diving in fully, there are many different videos, articles and resources out there for you to explore.

Patient Satisfaction Survey: Taking a Closer Look

Patient satisfaction survey results provide a deeper understanding of why your patients may or may not be satisfied with your service. Let's review an example.

 
By looking at our sample patient data set, we can see that patients were asked to rate their overall satisfaction with a healthcare provider. They were also asked to rate other important factors of their care such as nurse and doctor empathy, room appearance, on-time appointments, and amenities, which we will revisit later in this blog.

Our sample survey results show that 55% of patients were satisfied with their experience, which tells us that overall, most patients are satisfied with their healthcare provider's experience.

Patient Satisfaction: A Predictive Approach

By leveraging the predictive analytics module in Minitab Statistical Software, the healthcare provider can easily identify the key drivers of patient satisfaction. For our example, we'll use CART®.

CART®, or Classification and Regression Trees, is a decision tree algorithm used to find important patterns and relationships in data variables. If the question or challenge you're facing has a binomial or multinomial categorical response, use CART Classification, while anything that has a continuous response with many categorical or continuous predictors should use CART Regression.

In our sample survey, we are categorizing customers into two groups, whether they are satisfied or not satisfied with this healthcare provider, so we will use CART Classification. Minitab Statistical Software automatically finds the best decision tree for you and provides model statistics, so you can understand if the model is useful for your analysis.

As you can see below, nurse empathy and keeping a patient informed are the most important variables when predicting patient satisfaction, followed by doctor empathy and outcome of procedure, which also ranked highly as important.

Using Tree Diagrams to Understand Patient Data

To start, we'd like to mention that nurse empathy is measured on a 5-point scale, where 5 indicates a very positive evaluation. Looking at the tree in greater detail, we can see that when nurse empathy was rated greater than 3.5, approximately 82% of patients rate their experience as satisfied. We can also see that when patients rated their nurse empathy less than 3.5, they were more satisfied if they were better informed by the provider, but much less satisfied if they were not informed.

By looking at the tree above, healthcare providers can see that their patients want an empathetic nurse and expect to be informed throughout their visit — but knowing that even if their nurse does not show empathy, they can keep patients happy by keeping them informed — is an important insight.

Conclusion

Patient satisfaction is only a portion of the overall patient experience. Data-powered insights from predictive analytics, along with brainstorming tools, can help healthcare providers reach optimal patient care.