|
Activity Number:
|
90
- Novel Statistical Methods for COVID Pandemic and Other Current Health Policy Issues
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 9, 2021 : 10:00 AM to 11:50 AM
|
|
Sponsor:
|
Health Policy Statistics Section
|
|
Abstract #318688
|
|
|
Title:
|
Using Text Mining, Natural Language Processing, and Text Networks to Describe Content and Sentiment of Organization Emails from a Children’s Hospital
|
|
Author(s):
|
Figaro Loresto* and Nadia Shive and Lindsey Tarasenko
|
|
Companies:
|
Children's Hospital Colorado and University of Colorado and Children's Hospital Colorado
|
|
Keywords:
|
Natural Language Processing;
Text Mining;
Nursing;
Organizational Emails;
Text Networks
|
|
Abstract:
|
COVID-19 has caused healthcare organizations to adapt and= respond. Integral to this response are organizational emails to communicate operational information to pediatric nursing staff. Text mining and natural language processing (NLP), information for future strategic communication planning or triangulation of nurse experiences can be derived. This work aims to describe the content of the organizational emails from a pediatric hospital in the Western United States using text mining and NLP. Organization emails around COVID-19 were collected from March 2020 through March 2021 (N = 336). Sentiment analysis and topic derivation using text networks were conducted on the emails monthly. Polarity scores from the sentiment analysis were averaged by month. Three topics were extracted for the LDA modeling. Sentiment polarity scores ranged from 0.18 to 0.25. Infection control topics primarily were main topics across the pandemic timeline. Results suggest the organization utilizing positive polarity in messaging around essential organizational communication. Topics changed as the pandemic evolved and national, state, and organizational events happened.
|
Authors who are presenting talks have a * after their name.
