Systemic unreported protocol violations in key ivermectin study

Systemic unreported protocol violations in key ivermectin study

The study of Soto-Becerra et al is cited by the NIH COVID-19 Treatment Guidelines in forming its recommendation on ivermectin. The Guidelines state:

“Some clinical studies showed no benefits or worsening of disease after ivermectin use…”

Soto-Becerra et al is the only study cited by the NIH showing worsening of outcomes in patients with COVID-19 treated with ivermectin. That study deviated from the prescribed clinical protocol in a multitude of respects. Furthermore, those violations were never reported in the publication.

The clinical protocol is not publicly available, to our knowledge, but an excerpt from the protocol was released in a document, Memorando №1843 from Instituto de Evaluación de Technologias en Salud e Investigación (IETSI), the institute that employed the authors of the study. Other information about the clinical protocol is disclosed in another document, Nota № 38, released from the same organization. An administrator at the IETSI expressed concern about the differences between the study protocol and the publication in Memorando №1843:

“The study protocol shows differences in name and content with the distributed and pre-published versions.”

The protocol violations are listed below. None of the violations is mentioned in Soto-Becerra et al.

Change to first secondary endpoint. From Memorando №1843, the first secondary endpoint was: “Determine if the mentioned treatment regimens are effective in preventing the admission to the ICU evaluated as the difference in the instantaneous rates of incidence of entering the ICU in a 30-day followup period between the treatment groups.” The corresponding endpoint in the publication of the study was “all-cause in-hospital death and/or transfer to ICU”. Also, there was no 30-day limit on patient followup.

Change in the second secondary endpoint. From Memorando №1843, the second secondary endpoint was: “Determine if the mentioned treatment regimens are effective in preventing the development of acute respiratory distress syndrome (ARDS) evaluated as the difference in the instantaneous rates of incidence of ARDS in a 30-day followup period between the treatment groups.” The corresponding endpoint in the publication of the study was “all-cause in-hospital death and/or oxygen prescription.” Again, there was no 30-day limit on patient followup.

Absence of the third secondary endpoint. From Memorandum №1843 “Determine if the mentioned treatment regimens increase the frequency of serious adverse events (SAE) evaluated as the proportion of SAE and number of SAE per subject. There is no corresponding endpoint in the publication of the study. No data on adverse events is provided in the publication of the study.

Change to the data collection. From Nota №38: “Such variables [hemogram, dimer D, transferrin, PCR among others] were placed in the study protocol…” No data was presented in the published study for any laboratory data.

The impact on the findings of the study are unknown. One effect, however, is known. The use of “all-cause in-hospital death and/or oxygen prescription.” rather than “incidence of ARDS” had an impact on patient selection. From Soto-Becerra et al: “We excluded patients who were prescribed oxygen at admission since it is part of one of the outcomes.” They also state: “ …we only excluded patients who developed any of the outcomes within 24 hours of admission…” Thus, the effect of the protocol violation was to exclude patients hospitalized for COVID-19 who had significant pulmonary involvement.

The credibility of the NIH COVID-19 Treatment Guidelines Panel was brought into question by its inability to state whether a vote was held on its recommendation for ivermectin and to disclose the names of the Panel members who made the recommendation. That was reported earlier. This report raises serious questions about the validity of the recommendation itself.

Responses

This site uses Akismet to reduce spam. Learn how your comment data is processed.