Skip to main content
. 2023 May 23;12(11):3625. doi: 10.3390/jcm12113625

  1. Hernan and Robins stated that the rate of protocol compliance in Holme et al. 2014 was 70% [3], but Holme et al. reported in their publication that rate as 63% [10]. Note that Bretthauer et al., 2022 [6] had five coauthors in common with Holme et al.

  2. The only very large treatment effect of TOGETHER-FLV was its ten-fold reduction in mortality by PP analysis. No ITT measure of efficacy, except for reduction in emergency room visits, even achieved the threshold of p < 0.05 for statistical significance.

  3. As detailed in Appendix B, conflicting descriptions of the TOGETHER-IVM placebo group were provided in its study paper and other study documents. These describe TOGETHER-IVM’s ITT placebo group as consisting either of 3-day and 14-day placebo patients (matching the TOGETHER trial treatment arms having the corresponding numbers of days of drug delivery); of 1-, 3-, 10- and 14-day placebo patients; or of 3-day placebo patients only. Defying any usual meaning of per-protocol, the TOGETHER-IVM study paper stated at different points that its ITT placebo group included mixed placebos (either 3-day and 14-day only or 1-, 3-, 10- and 14-day), but that the PP placebo group dropped the mixed placebos to include only 3-day placebos. No rationale for this unusual selection from the ITT to PP placebo group or breakdown of any of these variously described mixed-day placebos was specified. Furthermore, none of the other three TOGETHER trial publications mentioned excluding mixed-day placebos from the PP placebo group. The different values for the placebo ITT-to-PP decrease in Tables 2 and 3 (of the study, [15]) arise from these conflicting explanations.

  4. The 10 May letter and four follow-up emails [24] were addressed to 11 coauthors of TOGETHER-IVM (those whose email addresses could be identified), five members of its data and safety monitoring committee and the editor-in-chief of the journal that published this study (the NEJM), and copied to 39 science journalists, bioethicists and other scientists.

  5. IVM was also omitted in the ClinicalTrials.gov record for the TOGETHER trial in its versions of both 15 January 2022 and 3 July 2022 (see Appendix A, above). On 19 July 2022, three coauthors of this analysis (DES, CA, EJF) emailed the TOGETHER-IVM coauthors about the omission of IVM. Two days later, the Vivli record was changed to include it among its “intervention/treatment” agents [24]. The Vivli data have two additional major shortcomings, listed in Appendix A.

  6. The mortality rates were ≤ 4% in both the treatment and placebo groups for all but one of the ten studies included in this June 2022 meta-analysis, and therefore relative risk (RR) was very close to the odds ratio (OR), with the pooled value of the latter = 0.51. The mortality reduction for the RCTs rated as having a low risk of bias was less (log OR = −0.12), but the TOGETHER-IVM study was included in that low-risk-of-bias group, weighted to account for 63% of its pooled log OR.

  7. To the first of the emails requesting PP outcomes for deaths and hospitalizations sent on 11 April 2022 by one of us, Colleen Aldous, study coauthor Edward Mills did not disclose those outcome numbers, but answered another question; Aldous thanked him, and said she would share his response with other investigators. Mills replied “Do. You may want to also let them know the world is not flat” [23].

  8. The dating of the FTX funding was of interest to Reuters with respect to whether it was in place during the TOGETHER trial’s evaluation of IVM [78]. As determined from web archive records [79], FTX funding for the TOGETHER trial first appeared on the latter’s website sometime after 3 December 2021 and before 3 March 2022. The first FTX grant to the TOGETHER trial was in the amount of USD 3.25 million, followed by a second, USD 15 million FTX grant awarded in May 2022 [80]. After the May grant was announced, “Funded by FTX” and the FTX logo appeared near the top of the home page of the TOGETHER trial’s website, as shown in archival website copies between 12 May and 19 November 2022. All references to FTX funding of the TOGETHER trial were dropped from the latter’s website within a few days before 22 November 2022, the date of the Reuters report [79].

  9. This failure to publish any of these submitted letters may be reflective of editorials that the NEJM published in 2015 and 2016, which downplayed concerns about data sharing and conflicts of interests [81,82,83,84,85]. These editorials prompted sharp dissent by three former senior editors of the NEJM [86] and others [87]. The NEJM also published the Bretthauer et al., 2022 colonoscopy study and accompanying editorial.

  10. On 7 June, an ICODA manager emailed one of the scientists who had inquired about this data, stating: “ICODA does not hold the data and we have requested that the authors update the data sharing agreement to reflect that” [28]. These follow-up emails, which noted that ICODA never held TOGETHER trial data, were sent to recipients including 11 TOGETHER-IVM coauthors and 39 cc’s [24], as described in item 3 above.