Q: Do we have to use journal-specific informed consent forms for case reports?
Does signed authorization by the patient or representative require use of the form specific to the journal or can the patient sign a "generic" IC that has the required elements and use that for journal submission with the manuscript? Although the case report for publication is de-identitified, the editor and publisher now have the patient's name and signature. Why did the ICMJE not utilize a Physician Attestation of Informed Consent (PAIC)?
Generally speaking, it is fine to use a generic form that follows the ICMJE guidelines or a form provided by the Institutional Review Board (IRB) that has the required elements. However, some journals provide specific IC forms for case reports that require that the patient has read the article to be published. In such cases, it is generally mentioned that it is manadatory to use the form provided by the journal. If there is no such instruction, you can submit the same forms to the journal as part of the submission package. If you are in doubt, you can write to the journal editor and find out.
The ICMJE guidelines on protection of research participants instructs authors to omit all non-essential identifying details of the patients in the publication. Since you have de-identified the patients, there is nothing to worry about. You need not be concerned about the informed consent forms being with the editor as journals have very strict confidentiality policies when it comes to patient information.
Regarding your comment about the ICMJE not using a Physician Attestation of Informed Consent (PAIC), one reason for this could be that using a PAIC might leave room for manipulation. However, we can raise this point with the ICMJE and see what their view is.