Data mining to explore the medication patterns of Dr. Zhou Zequan’s cough treatments

Abstract

Objective: To use data mining and statistical techniques to analyze the characteristics of combination prescriptions and medications used in the treatment of cough by Dr. Zhou Zequan, the famous and experienced chief physician of Chinese medicine in Xiangtan. Methods: The data obtained from patients with cough disease who were treated by Dr. Zhou Zequan were collected from our hospital information system (HIS). These data were analyzed with Excel 2016 and R language software to assess the frequency of drug use, the properties, flavors, and meridian of drugs, and the association rules of the drugs. SPSS v22.0 was used for drug hierarchical clustering analysis. Results: A total of 126 medical cases were included, comprising 100 drug prescriptions and 190 drugs. The drugs that were prescribed with high frequency contained licorice, Platycodon grandiflorum, Scutellaria baicalensis, Semen armeniacae amarae, Pericarpium citri reticulatae, and other substances. The flavors of the drugs were mainly bitter, pungent, and sweet; the properties of the drugs were mainly cold and warm; and the target meridians were mainly of the lungs, liver, and spleen. Sixty drug pairs were obtained by correlation analysis, and six new prescriptions were identified by hierarchical clustering analysis. Conclusion: The prescriptions had precise efficacy for coughs. The prescription medication patterns were subjected to data mining technology to discover information related to the treatment of cough at the clinical front line. The new prescriptions that were identified with clustering analysis could be used as alternative prescriptions in hospital preparations.

Keywords

• Data mining
• Cough
• Medication pattern