| Objective: The special comment on rational drug use is an effective means to improve the level of drug treatment in medical institutions,but in the past,the special comment was completed manually by pharmacists.Due to the limited number of pharmacists and the heavy workload of special comment,it can hardly be carried out in a normal way.This study explores the construction of a data mining model to make reasonable comments on human serum albumin（HSA）prescriptions providing support for new technologies and methods for clinical evaluation of all prescriptions quickly and accurately and special comment for drugs normally.Methods: In this study,the evaluation criteria for the rational use of HSA were summarized and formulated by collecting domestic and foreign evidence-based medical evidence,and the ICD-10 codes corresponding to the indications of HSA were sorted out.The SPSS Modeler 18.0 software was used to construct a data mining model using the algorithm of cluster analysis and association rules.And the model was applied to implement special comment on the diagnosis and treatment data related to HSA of inpatients in a medical institution from January to June 2021.Results: The data mining technology was used to construct the model of drug indications,the model of drug use for the special population,the model of drug dosage,and the model of drug frequency,which could be used to make special comment on the use of HSA from multiple perspectives.By applying the medication indication model,it was found that the use of HSA in this medical institution basically had medication indications,and only 1.00% of non-indicative medication was used.Among them,there were 6 doctors with a strong correlation with unindicated drug use,involving 4 clinical departments.By applying the drug model for the special population,it was found that the special population using HSA in the medical institution was large,accounting for 62.76%,and the special population requiring special attention accounted for 1.07%.Among them,there were 4 doctors who were closely related to the special population that need to be paid attention to,involving 3 departments.By applying the drug dose model,it was found that the average daily dose of HSA used in different indications in this medical institution basically met the recommended daily dose in the evaluation standard for the rational use of HSA.The correlation analysis between overdose medication and doctors and clinical departments was carried out,and it was found that there were 3 doctors with a strong correlation with overdose medication,involving 1 department.By applying the medication frequency model,it was found that HSA was used infrequently in the medical institution,and only used frequently when the patient had a severe hypovolemic shock,severe hypoalbuminemia or serum albumin concentration ≤25 g/L and other emergencies.Through these data mining models,various aspects of the use of HSA in hospitalized patients were analyzed from the macro and micro levels,which provided data support for the precise implementation of rational drug use intervention and guidance.Conclusion: This study explored the application of information technology to build a data mining model for the prescription comment of HSA.Through accuracy verification,the model successfully realized the rapid and accurate comment on the prescription of HSA,and achieved the purpose of quickly focusing on irrational drug use,ensuring safe and rational drug use,improving the work efficiency of pharmacists,and effectively saving medical resources.It provided new ideas and technical support for prescription comment in the future. |
PDF Full Text Request