JMIR Publications

ABSTRACT

Background:

Ischemic heart disease (IHD) is a leading cause of death globally with a disproportionate burden in low- and middle-income countries (LMICs). Natural Language Processing (NLP) allows for data-enrichment in large datasets to facilitate key clinical research.

Objective:

The primary objective of this study was to use NLP to assess gender differences in symptoms and management of patients hospitalized with acute myocardial infarction (AMI) at Aga Khan University Hospital-Pakistan.

Methods:

We used NLP to extract symptoms and medications for AMI from 5,358 discharge summaries. The model was trained iteratively, using batches of 50 documents, until an accuracy of 0.90 F-score was achieved. F-score, sensitivity and specificity were calculated at each stage.

Results:

NLP model specificity was maintained (100%), sensitivity improved from 78% to 83%, and an F-score of 99% obtained. Among 1,768 women and 3,590 men with AMI, women had higher odds of presenting with shortness of breath (OR 1.34; 95% CI 1.15-1.56) and lower odds of presenting with chest pain (OR 0.68; 95% CI 0.58-0.80), even after adjustment for diabetes and age. Presentation with abdominal pain, nausea/vomiting was much less frequent but consistently more common in women (p<0.01). “Ghabrahat”, a culturally distinct term for a feeling of impending doom was used by 4.1% of women and 2.9% of men as presenting symptom for AMI (p=0.09). First-line medication prescription was lower in women.

Conclusions:

Use of NLP for identification of culturally nuanced clinical characteristics and management is feasible in LMICs and could be used as a tool to understand gender disparities and address key clinical priorities in LMICs.