Background: Pneumonia is the hallmark of severe COVID-19, with supplemental oxygen requirement being the main indication for hospitalization. Refractory hypoxemia in these patients requires invasive mechanical ventilation (IMV) otherwise, death is imminent. In places with a high disease burden, availability of critical care experts, beds, or resources is challenged and many patients could die without receiving them.
Methods: We performed a retrospective cohort study using open databases from Mexico City about suspected or confirmed COVID-19 patients, health system saturation, and deaths between May 8th, 2020, and January 5th, 2021. After building a directed acyclic graph, we performed a binary logistic regression to identify the association between proposed causal variables and dying without receiving IMV (the outcome).
Results: We included 33 805 hospitalized patients with suspected or confirmed COVID-19, of which 19 820 (58.6%) did not require IMV and survived, 5416 (16.1%) required and received IMV, and 8569 (25.3%) required IMV but died without receiving it. Saturation of IMV-capable beds did not increase the odds of the outcome (odds ratio 1.07, 95% confidence interval 0.94-1.22 of 90%vs50% occupancy), while general bed saturation (2, 1.86-2.14 of 90%vs50% occupancy) and IMV-capable to general bed ratio (1.64, 1.52-1.77 for a ratio of 2vs0.5) did. Private healthcare decreased the odds of the outcome (0.12, 0.08-0.17) and dyspnea increased them (1.33, 1.19-1.9).
Conclusions: In Mexico City, increased general hospital bed saturation and IMV-capable to general bed ratio were associated with a higher risk of dying without receiving IMV. Private healthcare was the most protective factor.