Establishment of Comprehensive Indicators in TCM Pectoral-qi Case Report Based on Experts Diagnosis and Self-test Technology

RATIONALE: To establish TCM Pectoral-qi comprehensive indicators and highlight the inner structure among different variables in an objective way, the article uses Partial Least Square Second-order Latent Variable Model (PLS-SLVM) and accomplishes 3 different comprehensive indicators based on both experts diagnosis and self-test data. SLVM includes a measurement model that defines the relationship between observed variables and latent variables and a structure model that imputes relationships between latent variables. The article focuses on PLS as the estimation method. Without normal distribution and independence assumptions, PLS uses objective weighting methods based on the data. Bootstrap method (B = 200) is used to calculate the mean value and standard errors of the PLS estimates. The article chooses the percentile interval to obtain the confidence interval of PLS parameters. PATIENT CONCERNS: The patients were diagnosed by the means of experts diagnosis and self-test technology. On the one hand, the patients want to know the effect of self-test by wearing a kind of instrument. On the other hand, we want to establish TCM Pectoral-qi comprehensive indicators and highlight the inner structure among different variables in an objective way. DIAGNOSES: The group of 59 subjects are the same no matter whether they were diagnosed through TCM Pectoral-qi Assessment Questionnaire of self-test technology. INTERVENTIONS: The same group of 59 subjects keep wearing the instrument for hours and get the self-test data consequently. OUTCOMES: As one of comprehensive indicator establishing methods, PLS-SLVM highlights the structure state among variables and improves the evaluation efficiency. Furthermore, it provides a new tool and method in TCM diseases prevention and health security. LESSONS: As expected, PLS-SLVM is a useful tool due to its nonassumption of normal distribution and independence with consideration of correlation among different variables. Thus PLS-SLVM can be applied in ordinal data from assessment questionnaire and continuous data about physicochemical indexes for the same group of people. It displays that PLS-SLVM builds a connection between TCM experts diagnosis and the self-testing technology.