Prevalence of non-communicable chronic diseases in rural India amongst peri- and post-menopausal women: Can artificial intelligence help in early identification?

Abstract

AimsTo identify peri- and post-menopausal women at risk of non-communicable diseases in rural India and to assess their prevalence amongst these groups via the use of artificial intelligence. Settings and designAn observational study conducted by the Indian Menopause Society in collaboration with the Government of Maharashtra. The study included rural women residents of three villages in the Latur district of Maharashtra, India. Materials and methodsAccredited social health activist workers identified 400 peri- and post-menopausal women aged 45–60 years. Specific symptoms able to predict the presence of a non-communicable disease were identified through the use of artificial intelligence. Statistical analysis usedDescriptive statistics and predictive network charts analysis. ResultsThe mean age of 316 women included in the analysis was 50.4 years and the majority of them were illiterate (68 %). The prevalence of dyslipidaemia, osteopenia, diabetes mellitus, obesity and hypertension were 58 %, 50 %, 25 %, 25 %, and 20 % respectively. None of their symptoms or laboratory reports could be significantly correlated directly with any of these non-communicable diseases. Hence, we used a cluster of symptoms to suggest the presence of hypertension, diabetes mellitus, osteoporosis and hypothyroidism via predictive network analysis charts. ConclusionsScreening of at-risk women can be done using an artificial intelligence-based screening tool for early diagnosis, timely referral and treatment of non-communicable diseases with the support of community health workers.