Diagnostic traditional medicine criteria of Covid-19 by using a latent tree model

Abstract

Objectives: Until now, the guideline of the Vietnam Ministry of Health and several pieces of research have confirmed the effectiveness of traditional medicine in managing Covid-19, especially for mild levels. However, the research on differentiating traditional medicine disease patterns of Covid-19 to systemise diagnosis and treatment has not been implemented. Thence, this study aimed to establish diagnostic criteria for mild clinical forms of Covid-19 using latent tree models (LTMs). Materials and methods: Stage 1: Conduct studies on classic medical documents, textbooks and monographs on traditional medicine, to make questionnaires to observe symptoms in Covid-19 patients at a mild stage. Stage 2: Choosing and investigating Covid-19 patients at a mild stage by inclusion criteria. A 65-question survey was used to investigate 438 patients at Ho Chi Minh city Traditional Medicine Hospital, then included data were analysed by using the latent tree model. Results: Literature review exhibited 10 documents (5 classic medical documents, 5 textbooks) with content related to Covid-19. After modelling, 21 latent variables (Y0 to Y20) were established, referring to the groups of included symptom variables. According to ve - khi - dinh - huyet diagnosis theory, 2 traditional medical disease patterns are matching with the literature: the Ve-area disease pattern and the Khi-area disease pattern. Conclusions: The latent tree model analysis pointed out 2 traditional medical disease patterns of Covid-19 at a mild stage, including the Ve-area disease pattern and Khi-area disease pattern.