JMIR Publications

ABSTRACT

Background:

Experience-based knowledge and value considerations of health professionals, citizens and patients are essential to formulate public health and clinical guideline recommendations that are relevant and applicable at the medical frontline. During a pandemic this is even more important as well as more difficult to achieve.

Objective:

To explore the potential of AI-based methods to harvest experience-based knowledge and value considerations regarding COVID-19 vaccination from existing data channels to improve public health guidelines.

Methods:

We developed and examined these methods in relation to the COVID-19 vaccination guideline development in the Netherlands. We drew on social media as well as databases from the Netherlands Institute for Public Health and the Environment where experiences and questions regarding COVID-19 vaccination are shared. First, natural language processing (NLP) filtering techniques and a first approach for a supervised machine learning model were developed to identify this type of knowledge in a large dataset. Subsequently, structural topic modelling was performed to discern thematic patterns related to experiences with COVID-19 vaccination.

Results:

NLP methods proved able to identify and analyze experience-based knowledge in large datasets. They provide insights into a variety of experiential knowledge that is difficult to obtain in other ways for rapid guideline development. Some topics addressed by citizens, health professionals and patients concern subjects already covered in the vaccination guideline, such as potential contraindications, vaccination of high-risk groups and administrative and organizational matters. These may help identify problems with the guideline’s practice application as well as frequently occurring exceptions which might initiate a revision of the guideline text. Other topics are not yet reflected in the guideline, but are central concerns to citizens and professionals and may be equally important to consider.

Conclusions:

This proof-of-concept study presents NLP methods as viable tools to access and utilize experience-based knowledge and value considerations, possibly contributing to robust and applicable guidelines. The methods presented provide a way to broaden the evidence and knowledge base available for (rapid) guideline development.