Age Variation in COVID-19 Treatment in Mexico: A Retrospective Cohort Study} \date{August 2021

The COVID-19 pandemic has caused unprecedented stress on health systems throughout the world and the immense number of people infected has created demands for health care resources so high that rationing of medical equipment, especially ventilators, is inevitable. Administrators and physicians may follow certain guidelines or objectives and prioritize one types of patients over the other. Using data from Mexico, I study how intubation decisions vary with respect to patients' age after controlling for mortality risks. Intubation rates vary widely among patients of different characteristics; I estimate the mortality risks of patients and identify the high-risk ones using Cox proportional hazards model and develop a heteroskedastic probit model to predict the probability of a high-risk patient receiving intubation. My estimated model suggests that old patients (aged 70 - 100) have uniformly lower chance of receiving intubation across all risk levels than other patients, where at a certain risk level, middle-aged patients (aged 40 - 69) can have as much as 43\% higher chance of receiving intubation than old patients. On the other hand, there is relatively smaller variation between how younger (aged 18 - 39) and middle-aged patients are treated. The findings in this study could help administrators and physicians make more informed decision regarding triage of patients.