Uncovering Diagnostic Correlations Between Traditional Chinese and Western Medicine Using Instrumental Variable Models in Cardiometabolic Patients: Evidence from 1.2 Million Records

利用工具变量模型揭示心血管代谢患者中西医诊断相关性:来自120万份记录的证据

阅读:1

Abstract

BACKGROUND: Little is known about the relationship between syndromes in Traditional Chinese Medicine (TCM) and chronic diseases coded by Western Medicine (WM). TCM hospitals where both WM and TCM are practiced offer an opportunity to assess this relationship. TCM, based on syndrome differentiation and treatment, may aid in guiding treatment and predicting length of stay and healthcare costs. However, inconsistent coding of TCM syndromes arises due to variations in diagnostic interpretation, subjective assessment, and lack of standardized coding practices. The objective was to assess the correlation between WM diagnoses and TCM syndromes, accounting for diagnostic miscoding in the data. METHODS: We examined discharge data from 1,218,224 records for patients aged 45 and above, diagnosed with cardiometabolic diseases and admitted to TCM hospitals between 2017 and 2022, stays ranging from 24-hours to 90 days. Latent class analysis (LCA) was used to measure the correlation between TCM syndromes and WM. To address potential diagnostic errors, we applied bivariate probit models with instrumental variable (IV). RESULTS: There were 580,698 (47.67%) records for males, while 989,702 (81.24%) records from Tertiary-A hospitals. The LCA and probit models revealed that TCM syndrome diagnoses had a high ratio of noise to signal. After correcting for diagnostic errors, significant associations were found between WM diagnoses and TCM syndromes. Notably, diabetes mellitus was strongly associated with syndrome/pattern of qi and yin deficiency (coefficient = 2.711); cerebrovascular diseases showed strong associations with syndrome/pattern of qi deficiency with blood stasis (coefficient = 2.433) and syndrome/pattern of wind and phlegm blocking collaterals (coefficient = 3.176). All patterns had strong marginal effects (P < 0.001). CONCLUSION: This large-scale study quantitatively maps the relationship between TCM and WM diagnoses. It introduces a new statistical approach to understanding the correlation between these two diagnostic systems, offering insights into integrated medicine for secondary prevention.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。