| Objective Taking hypertension disease as an example,the Traditional Chinese Medicine(TCM)syndromes combined with syndrome classifications and statistical methods for syndrome classification and individual diagnosis are explained.The study can be used as a reference for the individual diagnosis of TCM syndromes of other diseases,laying the foundation for scientifically evaluating the efficacy of individual diagnosis and treatment,and improving the level of diagnosis,and providing valuable reference for the quality of TCM clinical research and the authenticity of research results.Method Through the multi-center epidemiological study,1499 cases of hypertensive patients who met the inclusion criteria and not met the exclusion were collected from 5 grade A comprehensive chinese medicine hospitals.Five main syndromes of hypertension were Liverfire Hyperactivity Syndrome,Liver and Kidney Yin deficiency Syndrome,Phlegm and Blood Stasis Syndrome,Heart and kidney deficiency Syndrome,Liver Stagnacy Injuring Spirit Syndrome.15 indicators were included in the Liver-fire Hyperactivity Syndrome,and 15 indicators were included in the syndrome of liver and kidney yin deficiency by confirmatory factor analysis and other methods.The Phlegm and Blood Stasis Syndrome contains 16 indicators,the Heart and kidney deficiency Syndrome contains 17 indicators and 12 indicators were included in the Liver Stagnacy Injuring Spirit Syndrome.These indicators are divided into two categories and multiple classifications.Finally,the latent class model and cognitive diagnosis model are used to identify classifications of individual syndromes in patients with hypertension.Results1.Latent Class Model(1)Liver-fire Hyperactivity SyndromeThe two-category data of Liver-fire Hyperactivity Syndrome classify hypertensive patients into four clasters,which are defined as "typical syndrome","former syndrome","syndrome" and "climacteric syndrome".The latent claster probabilities corresponding to these four clasters respectively are 0.2853,0.2709,0.2659 and 0.1779.For all 1499 patients with hypertension,417 cases,416 cases,406 cases,and 260 cases were eventually classified as hypertension in each claster.The ordered and multi-classified data of Liver-fire Hyperactivity Syndrome classify patients into 3 categories,which are inconsistent with the clinical practice of TCM,but can be defined in terms of the severity of syndromes,and can be defined as "climacteric syndrome","typical syndrome" and "former syndrome" respectively.The corresponding latent claster of the corresponding clasters of probability were 0.6017,0.2418 and 0.1565.For all 1499 hypertensive patients,there were 700,509,and 290 hypertensive patients who were eventually assigned to each claster.The latent class probabilities and case classification results of the two categories and multi-category data were basically in line with the law of gradual reduction from light to heavy.(2)Liver and Kidney Yin deficiency SyndromeThe two-category data of Liver and Kidney Yin deficiency Syndrome classify hypertensive patients into four clasters,which are defined as "former syndrome","climacteric syndrome","syndrome" and "typical syndrome".The latent claster probabilities corresponding to these four clasters respectively are 0.5323,0.3000,0.1179 and 0.0497.For all 1499 patients with hypertension,810 cases,450 cases,173 cases and 66 cases were eventually classified as hypertension in each claster.The multi-category data of Liver and Kidney Yin deficiency Syndrome classify hypertensive patients into five clasters,symptoms from cluster1 to cluster5 gradually rise in severity.The latent claster probabilities corresponding to these five clasters respectively are 0.4029,0.3797,0.1261,0.0726 and 0.0188.For all 1499 patients with hypertension,500 cases,675 cases,195 cases,103 cases and 26 cases were eventually classified as hypertension in each claster.The latent class probabilities and case classification results of the two categories and multi-category data were basically in line with the law of gradual reduction from light to heavy.(3)Phlegm and Blood Stasis SyndromeThe two-category data of Phlegm and Blood Stasis Syndrome classify hypertensive patients into four clasters,which can be defined in terms of the severity of syndromes,and can be defined as "former syndrome","climacteric syndrome","syndrome" and "typical syndrome" respectively.The difference between the thick and greasy mosses and slippery veins in claster2 and claster2 was relatively large,indicating that the thick moss,greasy moss,and slippery veins have a greater influence on the aggravation of the phlegm and blood stasis syndrome.The latent claster probabilities corresponding to these four clasters respectively are 0.4940,0.2352,0.2324 and 0.0384.For all 1499 patients with hypertension,787 cases,316 cases,341 cases and 55 cases were eventually classified as hypertension in each claster.The multi-category data of Phlegm and Blood Stasis Syndrome classify hypertensive patients into 3 clasters,which can be defined in terms of the severity of syndromes,and can be defined as "former syndrome","climacteric syndrome" and "typical syndrome" respectively.The latent claster probabilities corresponding to these four clasters respectively are 0.5899,0.3749 and 0.0352.For all 1499 patients with hypertension,922 cases,525 cases and 52 cases were eventually classified as hypertension in each claster.The latent class probabilities and case classification results of the two categories and multi-category data were basically in line with the law of gradual reduction from light to heavy.(4)Heart and Kidney deficiency SyndromeThe two-category data of Heart and Kidney deficiency Syndrome classify hypertensive patients into four clasters,which are can be defined as "former syndrome","climacteric syndrome","syndrome" and "typical syndrome" respectively.The latent claster probabilities corresponding to these four clasters respectively are 0.3258,0.2973,0.2074 and 0.1695.For all 1499 patients with hypertension,516 cases,420 cases,297 cases and 266 cases were eventually classified as hypertension in each claster.The multi-category data of Heart and Kidney deficiency Syndrome are the same as the two-category data classification results.The definition of each category is also the same,but the potential category probability corresponding to each category is different from the number of hypertension patients.The potential category probabilities are 0.3528,0.2689,0.2415,and 0.1368.There were 541 cases,390 cases,365 cases,and 203 cases respectively among the patients with hypertension.The latent class probabilities and case classification results of the two categories and multi-category data were basically in line with the law of gradual reduction from light to heavy.(5)Liver Stagnacy Injuring Spirit SyndromeThe two-category data of Liver Stagnacy Injuring Spirit Syndrome classify hypertensive patients into four clasters,which are defined as "former syndrome","climacteric syndrome","syndrome" and "typical syndrome".The latent claster probabilities corresponding to these four clasters respectively are 0.4484,0.3204,0.1512 and 0.0800.For all 1499 patients with hypertension,704 cases,459 cases,222 cases and 114 cases were eventually classified as hypertension in each claster.The multi-category data of Liver Stagnacy Injuring Spirit Syndrome classify hypertensive patients into five clasters,symptoms from cluster1 to cluster5 gradually rise in severity.The latent claster probabilities corresponding to these five clasters respectively are 0.4025,0.1862,0.1630,0.1515 and 0.0788.For all 1499 patients with hypertension,681 cases,266 cases,234 cases,202 cases and 116 cases were eventually classified as hypertension in each claster.The latent class probabilities and case classification results of the two categories and multi-category data were basically in line with the law of gradual reduction from light to heavy.2.Diagnostic Classification Model(1)Liver-fire Hyperactivity SyndromeThe category with the highest probabilities in the 4 clasters under the liver-fire hyperactivity syndrome is the group with no two properties(no hepatic fire or fire),and the category probability is 0.5073;the followed group is “hepatic fire and fire”,the category probability is 0.2750.At the same time the most hypertensive patients is classified into these 2 groups,the number was 791 cases and 433 cases respectively.(2)Liver and Kidney Yin deficiency SyndromeThe main categories of liver and kidney yin deficiency syndromes were the group with no two attributes(no liver yin deficiency or kidney yin deficiency),and the category probability was 0.5319.There is a small difference in the category probability of liver yin deficiency,kidney yin deficiency and liver and kidney yin deficiency,which is 0.1939,0.1262 and 0.1480 respectively.The number of hypertensive patients was classified into the group of no two categories(no liver yin deficiency or kidney yin deficiency)in 837 cases,and the other three groups were 284,172,and 206.(3)Phlegm and Blood Stasis SyndromeThe main two categories in the 8 categories of the phlegm and blood Stasis syndrome were the group of without any attributes(no wet,dampness,or sputum),the category probability was 0.4749;for the “wet” group,the category probability was 0.3425.The other six groups have a probability of less than 0.1.In the classification of hypertensive patients,the group of no three attributes(no wet,dampness or sputum)and “no wet” group had the most patients,the number is 788 and 478.(4)Heart and Kidney deficiency SyndromeThe main two categories in the 8 categories of heart and kidney deficiency syndrome were the group of without any attributes(no kidney yin deficiency,heart yin deficiency or yin deficiency)and the group of with all three attributes(kidney yin deficiency,heart yin deficiency or yin deficiency),the probabilities of the two groups were 0.4569 and 0.3103 respectively,and the hypertensive patients classified into 2 groups were 770 and 568.(5)Liver Stagnacy Injuring Spirit SyndromeThe main category of the four categories of liver stagnacy injuring spirit syndrome was the group of without any attributes(no liver fever or liver Stagnacy),the category probability was 0.5513;the probability of the other 3 groups was similar,the category probability were 0.1808,0.1036 and 0.1644 respectively.The final number of hypertensive patients classified in these 4 groups were 870,252,128,and 249 respectively.Conclusion The latent class model is a mathematical model that describes the complex relationship between multiple manifest categorical variables and can be applied to the individual diagnosis of TCM syndromes.According to the distribution of conditional probabilities of various factors in the potential categories,the classification of the population is conducted.Then the model is used to individualized diagnosis and classification by comparing the posterior probabilities of different groups of people.The latent class model are performed to provide basis for individualized treatment of TCM.The cognitive diagnosis model analyzes the responses of the hypertensive patients under test to a set of carefully designed questions,and infers the type of syndromes involved in the test on the subjects involved in the test.The model can provide the reference about Individualized diagnosis and classification.It is the same as the latent class model and has application value in individualized diagnosis and treatment of TCM. |
PDF Full Text Request