key: cord-0985051-cyljnzqw authors: Buonfrate, Dora; Bisoffi, Zeno title: Standard Dose Ivermectin for COVID-19 date: 2021-05-05 journal: Chest DOI: 10.1016/j.chest.2021.03.003 sha: 9bc0b62951d45e38c12cfd5d3838536eb92374a7 doc_id: 985051 cord_uid: cyljnzqw nan symptom onset. 2 The second is patients' health insurance coverage. In the American context, it is possible that uninsured patients, or those unable to afford hospitalization fees, would have delayed admission. 3 An important concern is that the used dose of ivermectin (200 mg/kg) reaches a plasmatic concentration much lower than severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) 50% inhibitory concentration at pulmonary tissue (<35 times inferior). 4 We highlight that the study considered all-cause mortality as the main outcome while evaluating only pulmonary severity. We consider that the severe pulmonary compromise category in the study is a heterogeneous one, grouping patients with different clinical statuses. The mortality trend through time was assessed as inconsistent, but a minimal variation in deaths or classification of severity would change statistical significance, so we believe it is relevant to show mortality trends over weeks. A caliper distance difference for propensity score matching (0.2, instead of, eg, 0.25) would risk introducing an additional selection bias. For example, a patient that received two doses of ivermectin and died was not included. If only one additional death would be included in the ivermectin treatment group, the OR for mortality would have changed from 0.47 (CI, 0.22-0.99) to 0.51 (CI, 0.25-1.05). Finally, in the study, the sex OR for mortality is nonstatistically significant, when there are studies that find a clear relationship between these two variables. 3, 5 Taking into consideration that age and sex are powerful confounders for mortality, we would have wished deaths by age and sex to be presented. To the Editor: The article by Ceplowicz Rajter et al 1 published in CHEST (January 2021), which presents a significant effect of ivermectin at standard dose on COVID-19 mortality rates, raises once again important questions on the significance of observational studies that report posttreatment outcome for COVID-19. Uncontrolled, observational studies have already created confusion in the medical community's response to the pandemics, the example of hydroxychloroquine being the most obvious one. 2 For instance, in their article, the criteria for treatment with ivermectin are not specified, and a bias due to treatment indication is not addressed completely. Moreover, despite the use of a propensity score matching aimed at reducing confounders, relevant variables might have been measured inadequately. As an example, the authors did not find a benefit associated with the use of steroids (which were given in a significantly higher proportion of patients in the ivermectin group of the unmatched cohort) and suppose that this finding, in contrast to what has been demonstrated by the RECOVERY clinical trial, 3 might be due to a propensity to save this treatment for the most critically ill patients. In addition to the limitations due to the study design, we would like to point out the concerns about the dose of ivermectin that has been used. Indeed, based on the article by Caly et al, 4 the potential drug efficacy in vitro was observed at high ivermectin concentration. The IC-50 reported (2,190 ng/mL) was at least 50 times higher than the maximal concentration achievable with the standard dose of 200 mg/kg, which is chestjournal.org the one reported by Ceplowicz Rajter et al. 1 A potential clinical efficacy of this dose was not even plausible; thus, introducing ivermectin as a treatment for COVID-19 patients was (and is) not justified. We currently are coordinating a randomized, Phase 2, double-blind clinical trial on ivermectin at high dose (600 or 1200 mg/kg for 5 consecutive days) for COVID-19 early stage in patients with mild symptoms who do not need hospitalization. 5 We did not even consider to include an arm with a standard dose, precisely because of implausibility. We humbly argue that, if reported results of an observational study are not plausible, unrecognized confounders are the most likely explanation. Ivermectin is being used already in an uncontrolled way in Latin American countries. 6 We do not need an improper use of the drug to be further promoted, unless well-designed randomized controlled trials will be able to prove an efficacy, presumably with the use of much higher doses of ivermectin. We appreciate the comments of Ortega-Guillen et al 1 and Buonfrate and Bisoffi 2 and their interest in our paper. Ortega-Guillen et al 1 ask about the time to onset of symptoms; unfortunately, because of the retrospective nature of the study, these data were not uniformly available to the authors of the Ivermectin in COVID-Nineteen (ICON) study. 3 Speculations can be made as to the importance of this variable, but it would likely have been evenly divided between the two propensitymatched groups. Regarding their comments on corticosteroid use, the RECOVERY Collaborative Group 4 only documented an absolute mortality risk reduction of 2.9% in nonventilated patients and 12.1% in mechanically ventilated patients. This obviously cannot account for the much larger absolute mortality risk reduction in our cohort. Moreover, in our propensity-matched cohort, the two populations were matched for corticosteroid use, and a mortality difference was still demonstrated. We would also note parenthetically that in the randomization section of the RECOVERY trial, the managing physician had the discretion to consider use of dexamethasone as either definitely indicated or definitely contraindicated. These patients were excluded from analysis. This may have biased patient selection in the randomization process and the intention-to-treat analysis during the recruitment of patients in the RECOVERY trial and may have skewed the outcome. Regarding their question on insurance coverage, unlike many parts of the United States and the world, the four facilities in the ICON study are subsidized by a county government program to provide care for uninsured individuals. Thus, insured as well as uninsured patients are receiving identical care, notwithstanding their insurance status. Ortega-Guillen et al 1 are indeed correct in stating that increased mortality in one group or the other would have changed the statistical outcome, but that would be true of any study. We believe that Ortega-Guillen et al 1 Use of ivermectin is associated with lower mortality in hospitalized patients with coronavirus disease 2019: The Ivermectin in COVID Nineteen Study Dexamethasone in hospitalized patients with Covid-19 Disparities in outcomes among COVID-19 patients in a large health care system in California Use of ivermectin is associated with lower mortality in hospitalized patients with coronavirus disease 2019 Retraction-Hydroxychloroquine or chloroquine with or without a macrolide for treatment of COVID-19: a multinational registry analysis Dexamethasone in hospitalized patients with Covid-19: preliminary report The FDA-approved drug ivermectin inhibits the replication of SARS-CoV-2 in vitro COVidIVERmectin: Ivermectin for treatment of Covid-19 (COVER). NCT04438850. ClinicalTrials.gov. National Institutes of Health Latin America's embrace of an unproven COVID treatment is hindering drug trials