JMIR Publications

ABSTRACT

Background:

The increasing availability of ‘real-world data’ in the form of written text hold promise for deepening our understanding of societal and health-related challenges. Textual data constitute a rich source of information which allow to capture lived experiences through a broad range of different sources of information (e.g., content, emotional tone). Many studies rely on interview techniques in order to gain such insights. However, conductance, transcription, and evaluation of interviews is time-consuming, only feasible in small samples, take place in a specific interview setting, and typically interviews require an elaborate evaluation process. While innovations in natural language processing have made large-scale text analysis more accessible, the analysis of real-world textual data is still complex and requires several consecutive steps.

Objective:

To provide guidance for applied researcher, we developed and subsequently examined the utility and scientific value of a natural language processing (NLP) pipeline for extracting real-world experiences from textual data.

Methods:

We applied the NLP pipeline to large-scale textual data collected by the Swiss Multiple Sclerosis (MS) registry. Such textual data constitute an ideal use case for the study of real-world text data. Specifically, we examined 639 text reports on the experienced impact of the first Covid-19 lockdown from the perspective of persons with MS. The pipeline has been implemented in Python and complemented by analyses of the ‘Linguistic Inquiry and Word Count’ (LIWC) software. It consists of five interconnected analysis steps: (1) Text preprocessing; (2) Sentiment analysis; (3) Descriptive text analysis; (4) Unsupervised learning - topic modelling; and (5) Results interpretation and validation.

Results:

A topic modelling analysis identified four distinct groups based on the topics participants were mainly concerned with: ‘Contacts / communication’; ‘Social environment’; ‘Work’; and ‘Errands / daily routines’. Notably, the sentiment analysis revealed that the ‘Contacts / communication’ group was characterized by a pronounced negative emotional tone underlying the text reports. This observed heterogeneity in emotional tonality underlying the reported experiences of the first Covid-19-related lockdown is likely to reflect differences in emotional burden, individual circumstances, and ways of coping with the pandemic situation, which is in line with previous research into this matter.

Conclusions:

This study illustrates the timely and efficient applicability of an NLP pipeline and thereby serves as a precedent for applied researchers. Our study thereby contributes both to the dissemination of NLP techniques in the applied health sciences as well as to tailoring MS treatment in times of the pandemic to the individual needs. Clinical Trial: https://clinicaltrials.gov/ct2/show/NCT02980640