In a nutshell, if someone has lost their raw data, workup data and laboratory books (so that in effect their data cannot be checked/queried/verified/substantiated) what would be the implications of submitting his/her results to a journal? I have a very clear view. I would not do it. However, others seem to think that if you cannot prove that the results are wrong, then they must be accepted on trust. I am hoping that you could point me towards some official ruling.
The peer review process would not be of any help in such a situation as it is unlikely that reviewers would ask to see the original calculations or even imagine a story of such careless custody of data.
The Forum was unanimous in the view that if an editor is made aware that the raw data are missing, they would not publish the paper. However, the Forum agreed that it is not practical to ask authors to submit their raw data, although some journals do ask for certain types of data to be deposited in a public database and some journals ask for original digital images. But if an editor asks an author for raw data and the author cannot produce the data, then the editor should not publish the paper. The Forum did however caution that you should ask the author how the data were lost, as there are some genuine cases of data being lost after a fire or flood.