Critical appraisal of multidrug therapy in the ambulatory management of patients with COVID-19 and hypoxemia

This critical appraisal is focused on three published case series of a total of 119 COVID-19 patients with hypoxemia who were successfully treated in the United States, Zimbabwe, and Nigeria with similar off-label multidrug treatments that may include ivermectin, nanosilver, doxycycline, zinc, and vitamins C and D, resulting in rapid recovery of oxygen levels. We investigate whether these treatment protocols were successful in preventing hospitalizations and deaths. We use a simplified self-controlled case series method to investigate the association of treatment with the existence of a hospitalization rate reduction effect. To show the association of treatment with the existence of a mortality rate reduction effect, we make conservative comparisons of the treatment case series with several external control groups using the exact Fisher test. A novel statistical technique, based on the Sterne interval and the Bayesian factor, is used to assess the resilience of these results with respect to selection bias. The existence of statistically significant hospitalization rate reduction is shown for two of the three case series with the most aggressive treatments, and it is resilient against both random and systemic selection bias. Combining all three case series allows us to show the existence of statistically significant mortality rate reduction, and it is more likely than not that random selection bias does not overturn this finding. These results, combined with an extensive literature review, show that the efficacy of these multidrug treatments is supported by the Bradford Hill criteria of strength of association, temporality, biological gradient, consistency, and biological plausibility.