AboutEventsBecome A MemberResourcesGet Involved

Welcome

This login popup box will appear when a user tries to access content that is for members only. If you are not a member yet, please register for a membership. We’d love to have you be part of our community.

  • Basic/free members will have access to a limited amount of site content.
  • Premium/paid members will have access to all site content.

All users will have to be logged in to access the content. Become a member and gain access to exclusive features.

Login

 

Letting Patient Narratives Guide the Future of Healthcare

Monday, December 21, 2015
Senem Guney, PhD, Founder, Narrative Dx
In order to improve patient experience, we must first understand the patients’ experiences from their perspectives in all aspects of their care. Were their questions answered during their stay? Were the facilities clean? Did they get the information they needed at discharge? The medical staff, the facilities, and even behind-the-scenes administration contribute to each patient’s experience with a healthcare provider.   

Patients draw from all of these experiences when they talk about the care we provide them in satisfaction surveys, on social media, on doctor review sights, and in compliment and complaint letters. Across these distinct platforms, patients give us narratives of their experiences. If we need consistent measures of care to improve future iterations of care (Greenhalgh, 1999), we need to build measures that are grounded in these narratives to drive patient-centered care forward in a highly reproducible way.

Check-box surveys of patient satisfaction were an important beginning for us to understand and improve patient experiences. Patient narratives, though, provide context to these check boxes. They allow us to truly hear the voice of patients and understand their needs and expectations.


Learning from Patient Narratives


Hospitals know that happy patients improve their bottom line. The question is how to effectively analyze patient experience data (Coulter, Locock, Ziebland, & Calabrese, 2014), especially patient narratives (Cognette-Rieke & Guney, 2014) in order to know what makes patients happy or unhappy from their perspective.

We need to draw actionable insights from what patients say in their own words about their experiences, wherever they say it. The most basic approach to organizing and analyzing narrative data is doing keyword searches. We need to think twice, though, before we claim we can hear the voice of our patients by doing keyword searches on what they tell us. When we do keyword searches, we can only find what we know to look for in patient narratives. If we are interested in hearing what patients are really telling us, we need to analyze their narratives so that we can learn what is important to them, without imposing on their narratives keywords that indicate what we think is important.

Methods based on natural language processing (NLP) allow us to find what we do not necessarily know to look for in patient narratives. These NLP-based methods offer the analytical capability to see key themes emerge from large volumes of patient narratives. With this capability, we can actually hear what patients say about their experiences as they say it and we learn about the key drivers of good or bad experiences from the patients’ perspective.

As NLP-based methods become more commonly used in analyzing (narrative) patient experience data, we can also encourage the collection of more comments in patient satisfaction surveys (currently only 20% of patients leave a comment in surveys [Smith & Hoppertz, 2014]).  Healthcare providers who systematically solicit narrative feedback on their care delivery will benefit from good outcomes (Luxford, Safran, & Delblanco, 2011). Because, they will be the ones to know the root causes of good and bad experiences and act swiftly on what their patients tell, in their own words, what needs to change for better, higher quality patient-centered care.

Tags: patient feedback, patient narratives, survey data
Comments
Blog post currently doesn't have any comments.