http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Predicted Model of Hospitalization and Intensive Care Unit Use in Korean COVID-19 Patient
( Mindong Sung ),( Sungho Won ),( Bora Lee ),( Jonghoon Yang ),( Young Sam Kim ),( Su Hwan Lee ) 대한결핵 및 호흡기학회 2020 대한결핵 및 호흡기학회 추계학술대회 초록집 Vol.128 No.-
Background Severe acute respiratory syndrome coronavirus 2 is causing a pandemic of Coronavirus disease 2019 (COVID-19) with serious clinical manifestations, including death. However, medical resources are limited. For efficient use of medical resources, a model is needed to predict patients who need oxygen therapy and intensive care unit (ICU). This study was conducted to find this predictive model. Methods This study retrospectively reviewed 5,628 confirmed COVID-19 patients using data of Korea Disease Control and Prevention Agency. Enrolled patients were divided into patients who do not need oxygen therapy vs. those who need oxygen therapy and general ward vs. ICU. Results Of the enrolled 5,601 patients, 1,776 (31.7%) were over 60 years old. 2310 (41.2%) were male, and the mortality was 252 (4.5%). The 4785 (85.4%) patients improved without oxygen therapy. Among the 816 patients requiring oxygen therapy, 304 patients received treatment in the ICU. There were significant differences in age, sex, fever, dyspnea, altered consciousness, presence of diabetes mellitus, presence of chronic kidney disease, presence of cancer, and presence of dementia between patients who needed oxygen and those who did not. Predicted model of the need for oxygen therapy including these variables showed 0.879 the area under the curve (AUC) on receiver operating characteristic (ROC) curve. The analysis between patients who treated in general ward and those who treated in ICU showed statistical significant differences in age (≥60), sex, dyspnea, altered consciousness, presence of cancer, and presence of dementia. Predicted model of need for ICU including these variable showed 0.772 the AUC on ROC curve. Conclusions The predictive models of oxygen therapy and ICU admission will be helpful in utilizing limited medical resources. Acknowledgement: We thank Korea Disease Control & Prevention Agency, National Medical Center and the Health Information Manager in hospitals for their effort in collecting the medical records.