No abstract available.
Machine Learning

Background: The coronavirus disease 2019 (COVID-19) pandemic is an emerging threat worldwide. It remains unclear how comorbidities affect the risk of infection and severity of COVID-19. Methods: This is a nationwide retrospective case-control study of 219,961 individuals, aged 18 years or older, whose medical costs for COVID-19 testing were claimed until May 15, 2020. COVID-19 diagnosis and infection severity were identified from reimbursement data using diagnosis codes and on the basis of respiratory support use, respectively. Odds ratios (ORs) were estimated using multiple logistic regression, after adjusting for age, sex, region, healthcare utilization, and insurance status. Results: The COVID-19 group (7,341 of 219,961) was young and had a high proportion of female. Overall, 13.0% (954 of 7,341) of the cases were severe. The severe COVID-19 group had older patients and a proportion of male ratio than did the non-severe group. Diabetes (odds ratio range [ORR], 1.206–1.254), osteoporosis (ORR, 1.128–1.157), rheumatoid arthritis (ORR, 1.207–1.244), substance use (ORR, 1.321–1.381), and schizophrenia (ORR, 1.614–1.721) showed significant association with COVID-19. In terms of severity, diabetes (OR, 1.247; 95% confidential interval, 1.009–1.543), hypertension (ORR, 1.245–1.317), chronic lower respiratory disease (ORR, 1.216–1.233), chronic renal failure, and end-stage renal disease (ORR, 2.052–2.178) were associated with severe COVID-19. Conclusion We identified several comorbidities associated with COVID-19. Health care workers should be more careful while diagnosing and treating COVID-19 when patients have the abovementioned comorbidities.