應用UMLS建立線上住院病人疾病分類給碼之輔助系統

• Main Authors: Yu,Chin-Yuh, 游金玉
• Other Authors: 盧瑞山
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/xm745u

碩士 === 德明財經科技大學 === 資訊科技與管理研究所 === 100 === The Fund from Bureau of National Health Insurance (NHI) has become the main income of all Taiwanese hospitals since 1995 as the NHI is established. For controlling all medical expenses from ever growing, the Bureau of NHI executed the act of payment in Diagnosis Related Groups (DRG), which is another milestone of Taiwanese healthcare insurance system. Hospitals and Medical Doctors are responsible for expense control under the DRG system by putting the patients of similar diagnosis, age, sex, and treatment into the same Diagnosis Group , and give them a corresponding value. In short, all patients under similar conditions and treatments ought to have same expenses, and the hospital will be resoponsible for the money control. The basis of DRG is the International Classification of Diseases, or ICD-9. The ICD-9 commands the main diagnosis of a patient, the doctor must follow the main diagnosis and input ICD, then follow the DRG examine program to receive a DRG code for applying fund from the Bureau of NHI. However, the coding process and the retrieving DRG code is proceeded by a ICD-9 coding specialists after finishing the discharge record, by the mean time the patient has already left the hospital and the cost is paid, thus the hospital will never be able to control the cost. If the hospital is responsible for cost- effective control, they have to retrieve the DRG code while the patient is still in hospital under medical treatment. Therefore, to be responsible to cost management, all Medical Doctors has to be familiar with the ICD-9 coding system, and retrieve the DRG code before the patient discharge, then estimate the fund from NHI. This is a feasibility study on the possibility of writing a disease classification programme with Unified Medical Language System (UMLS). The result indicates that the DRG codes provided by the programme are 80% correct. After linking with the HIS system of the hospital, now doctors under heavy work load are able to identify the ICD-9 and DRG codes to control the expenses for the hospital. Although there is a 20% fail rate since the complicated cases might affect the programme from judging it correctly which needs the assistance from ICD-9 coding specialists , the programme itself is still effective.