Reader Comments

Post a new comment on this article

Conversation with lead author

Posted by SPalak on 19 Nov 2022 at 07:14 GMT

Dear all, dear authors,

We have been in contact with the leading researcher of this paper, Laura CG, who confirms that a significant revision of this paper is under its way.

When our group, who is working on a review (prospero CRD42022326540), extracted data from this study (available in the supporting information) we were not able to re-produce the findings outlined in the study. Indeed, we found results opposing the presented findings.

We urge the scientific community to download the supporting data file to arrive at their own conclusions or await the revision from the author group before drawing any conclusions about the presented data.

Kind regards,
Sunil Palak

Mail correspondance with Laura CG:
Dear Dr. Palak,

I am Laura Gomes, the first author of the paper “Levels of hepatitis B antibody titers are affected by age and doses gap time in children from a high endemic area of the western Amazon”. Dr. Rita Uchoa shared with me your questions about this paper. We are delighted with your interest in including our work in your meta-analysis. So, I hope I can clarify your questions.

Few weeks after the publication of our paper, we realized that it was missing an information in the methods section, which is described below:

“The anti-HBS variable was ln-transformed before performing the statistical analysis; and in 16 cases where the values were zero, they were replaced by 0.001 without compromising the final analysis and results.”

The reason why we replaced zero for this value is because the log of zero is 1, and this prevents a normal distribution of the data due to the disproportionality generated by it. This is one of the main limitations in log transformation of a data to normalize the distribution.

Therefore, we immediately requested a post-publication correction to the journal as soon as we realized it was missing in the paper. Despite their acceptance for correction, they reported that would take several months until it be online. Unfortunately, they did not publish this correction yet.

It is probably why you are not getting the same results as us. However, if after you replace the zeros by 0.001 you keep having troubles in the analyses, please let me know. Then I can check or share the script I used to do the analyses.

Good luck in your work.

Best wishes,
Laura Gomes

No competing interests declared.
report a concern respond to this posting