Background & Aims

A substantial number of qualitative studies have been developed to investigate and comprehend the experiences of patients through narrative discourse. This form of discourse enables patients to create a narrative reflecting their individual, societal, and healthcare experiences. Computational linguistics, often referred to as natural language processing (NLP), represents a branch of computer science that employs computational methods to interpret, assimilate, and generate human language content. We aimed to (i) undertake a descriptive analysis of the discourse from individuals suffering from chronic low back pain using sentiment analysis (SA) and network analysis; (ii) examine the association between patients’ profiles, the intensity of pain, and disability levels in relation to SA and network analysis; and (iii) identify distinct clusters in our dataset based on linguistic patterns and SA, utilizing an unsupervised machine learning method.

Methods

We conducted a secondary evaluation of a qualitative investigation that involved participants with chronic non-specific low back pain. Our focus was on the data concerning participants’ emotional responses upon receiving their diagnosis. For the sentiment analysis (SA) and network analysis, we utilized the Valence Aware Dictionary and sEntiment Reasoner (VADER), and the Speech Graph techniques, respectively. The clustering process was executed using the K-means algorithm.

Results

In the SA, the mean composite score recorded was -0.31 (Sd. = 0.58), indicating a predominantly negative discourse among most participants (41 out of 57, or 72%). The Word Count (WC) and the Largest Strongly Connected Component (LSC) showed a positive correlation with the participants’ level of education. However, there were no statistically significant correlations found between the intensity of pain, disability levels, SA, and network analysis. Our analysis identified two distinct clusters within the sample.

Conclusions

The sentiment analysis (SA) revealed that participants expressed their emotions using negatively toned sentences when recounting their experiences at the time of diagnosis. There was no statistically significant correlation found between the severity of pain, levels of disability, SA, and network analysis outcomes. A positive correlation was observed between the level of education and both the Word Count (WC) and Largest Strongly Connected Component (LSC).

References

1. Verbeek J, Sengers MJ, Riemens L, Haafkens J. Patient expectations of treatment for back pain: a systematic review of qualitative and quantitative studies. Spine. 2004;29(20):2309–18.

2.Coelho RM, Drummond C, Mota NB, Erthal P, Bernardes G, Lima G, et al. Network analysis of narrative discourse and attention-deficit hyperactivity symptoms in adults. Plos one. 2021;16(4):e0245113.

3. Liles BZ, Duffy RJ, Merritt DD, Purcell SL. Measurement of narrative discourse ability in children with language disorders. Journal of Speech, Language, and Hearing Research. 1995;38(2):415–25.

4. Scott CM, Windsor J. General language performance measures in spoken and written narrative and expository discourse of school-age children with language learning disabilities. Journal of Speech, Language, and Hearing Research. 2000;43(2):324–39.

5. Hirschberg J, Manning CD. Advances in natural language processing. Science. 2015;349(6245):261–6.

6. Mota NB, Vasconcelos NAP, Lemos N, Pieretti AC, Kinouchi O, Cecchi GA, et al. Speech graphs provide a quantitative measure of thought disorder in psychosis. PloS one. 2012;7(4):e34928.

7. Pavlopoulos GA, Secrier M, Moschopoulos CN, Soldatos TG, Kossida S, Aerts J, et al. Using graph theory to analyze biological networks. BioData mining. 2011;4(1):1–27.

8. Zunic A, Corcoran P, Spasic I. Sentiment analysis in health and well-being: systematic review. JMIR medical informatics. 2020;8(1):e16023.

Presenting Author

Felipe J. J. Reis

Poster Authors

Felipe Reis

PhD

Federal Institute of Rio De Janeiro

Lead Author

Igor da Silva Bonfim

UNISUAM

Lead Author

Leticia Amaral Corrêa

Macquarie University

Lead Author

Leandro Calazans Nogueira

UNISUAM

Lead Author

Ney Meziat-Filho

UNISUAM

Lead Author

Renato Santos Almeida

Lead Author

Topics

  • Assessment and Diagnosis