Your browser version is outdated. We recommend that you update your browser to the latest version.

Understanding patient satisfaction with received healthcare services:
A natural language processing approach

This work was done in collaboration with Dr. Mike Conway, with assistance by Danielle Mowry and Dr. Wendy Chapman.

Important information is encoded in free-text patient comments. We determine the most common topics in patient comments, design automatic topic classifiers, identify comments’ sentiment, and find new topics in negative comments. Our annotation scheme (picture 2) consisted of 28 topics, with positive and negative sentiment. Within those 28 topics, the seven most frequent accounted for 63% of annotations. For automated topic classification (picture 1), we developed vocabulary-based and Naïve Bayes’ classifiers. For sentiment analysis, another Naïve Bayes’ classifier was used. Finally, we used topic modeling to search for unexpected topics within negative comments. The seven most common topics were appointment access, appointment wait, empathy, explanation, friendliness, practice environment, and overall experience. The best F-measures from our classifier were 0.52(NB), 0.57(NB), 0.36(Vocab), 0.74(NB), 0.40(NB), and 0.44(Vocab), respectively. F- scores ranged from 0.16 to 0.74. The sentiment classification F-score was 0.84. Negative comment topic modeling revealed complaints about appointment access, appointment wait, and time spent with physician. 


System to identify topics in patient satisfaction feedbackSystem to identify topics in patient satisfaction feedback



Ontology of patient satisfaction topicsOntology of patient satisfaction topics