Authorship and signature of regulatory documents
Who counts as the author of a clinical study protocol or a clinical study report, and who has to sign one, are questions the ICH guidelines answer only in part. Companies usually attribute authorship by their own policies, the signatures a report needs are set by guidance, and an investigator's signature asserts less than a journal author's endorsement of a paper. Public disclosure of reports has made some writers more reluctant to sign them, and one experienced writer holds that a regulatory writer's role is a technical one rather than the intellectual authorship a journal article carries.
Why the authorship of regulatory documents went undiscussed
Clinical study protocols, clinical study reports, investigator's brochures and informed consent forms are among the most common documents a regulatory medical writer authors. Unlike the authorship of publications, which has been under scrutiny in recent years, the authorship of these documents has not really been a topic of discussion. This is probably because they have traditionally been hidden behind the shroud of confidentiality. Now that increasing requirements for transparency have reached regulatory documents, their authorship attribution needs to be considered, and as some of them move into the public domain, the implications of disclosure for their authors and signatories need to be seen too.
Protocol authorship and signature
A protocol is frequently written by a regulatory medical writer who receives input from other functional groups, among them a biostatistician and a medical expert. In some cases investigators or clinical research scientists draft it themselves, with or without a writer's support.
ICH E6 gives no guidance on the authorship of the protocol. It does indicate who should sign the final protocol: the investigator or institution and the sponsor. By signing the protocol, or an alternative contract, they declare their commitment to follow the protocol and abide by the principles of good clinical practice.
Report authorship in ICH E3
Clinical study reports are more often than not written by regulatory medical writers.
E3 mentions authorship in two places. Section 6, on investigators and study administrative structure, recommends that a list of the people whose participation materially affected the conduct of the study be provided in appendix 16.1.4, and that this list include the author or authors of the report, including the responsible biostatisticians. Appendix 16.1.5 on signatures uses a different term: the sample signature form in Annex II refers to study authors, and these authors are not necessarily the signatories. E3 does not clearly specify the distinction, if any, between report author and study author. Its recommendations on report authorship are somewhat unclear.
If the two terms are read as interchangeable, the medical writers who drafted the text sections of the report would in theory qualify as its authors, and so would the biostatisticians who produced the statistical outputs. Other roles to consider for authorship or contributorship are the site staff, meaning investigators and subinvestigators, the sponsor, and the staff of a contract research organisation where one is used. If the terms are not interchangeable, attributing authorship might require metrics for an individual's level of involvement in and contribution to the study.
How companies attribute report authorship
In practice, authorship attribution usually depends on company policy, and one regulatory writer has met three patterns over the years.
- Authorship is attributed by default to sponsor personnel. Where writing is outsourced, the medical writing company, consultant or contract research organisation is listed in appendix 16.1.4 as responsible for writing the report, while the title page lists only the sponsor's responsible person or people, one, some or all of whom sign.
- No individual is named as an author. An institutional author, the company name, appears on the title page and in the synopsis, a responsible medical officer on the sponsor side signs, and appendix 16.1.4 may or may not list the individuals or institutions responsible for developing the report.
- The report is a shared endeavour, with responsibility spread across the functional groups. Some clients insist on naming the medical writer, the biostatistician and the medical officer as authors of the document, and usually expect these authors to be signatories in appendix 16.1.5.
Report signatories and what an investigator's signature asserts
The authors and the signatories of a report are not necessarily the same people, as the first two patterns show. Most regulatory documents are authored and approved internally, but a company must sometimes seek the signature or approval of someone outside it. One regulatory writer sets the minimum for the signature page in appendix 16.1.5 at the signature of the sponsor's responsible medical officer, with the coordinating investigator's signature also required in European studies. Another writer reads ICH guidance as requiring that a clinical study report be accompanied by the signature of the principal investigator for a single-centre trial, or of the coordinating investigator for a multicentre one. In 2016 no regulatory guideline mandated the medical writer as a co-signatory of a report, and signatories beyond those required by ICH or by the relevant country or regional guidelines remained a matter for individual sponsor consideration or policy.
The regulations are rather vague about who should be the coordinating investigator. ICH E3 states that the coordinating investigator will usually be designated by the protocol, but its wording does not make this obligatory, and in practice the protocol is often silent on the matter. Where a study has a data safety monitoring board or another oversight body, one common practice is to make its chair the coordinating investigator, and another approach might be to ask the investigator who recruited the most patients. Either way, the final decision often has a political dimension; for example, if a publication is planned from the study, the lead author and the coordinating investigator may be one and the same.
The investigator signature page essentially confirms that, to the best of the signatory's knowledge, the trial was conducted according to good clinical practice and the results presented reflect those obtained. Unlike the author of a publication, the investigator does not necessarily have to agree fully with the interpretation of the results in order to sign, only acknowledge that the study was appropriately conducted and that the results themselves are accurate.
Public disclosure and the willingness to sign
The report and the protocol are the two most important clinical trial documents affected by disclosure through the European Medicines Agency's Policy 0070. The third pattern, in which the writer is named as an author and expected to sign, is especially controversial, because some medical writers do not feel comfortable signing so important a document. The requirement to disclose reports publicly has increased their reluctance to sign. Under Policy 0070 as it stood in March 2016, sections 16.1.4 and 16.1.5 were not to be posted publicly, so attributing authorship to individuals in those sections did not amount to disclosing their personal data, though even this argument cannot allay concerns about possible legal implications. Handled appropriately, authorship attribution and sign-off on a report need not be affected by public disclosure.
The investigator's brochure and the informed consent form
As with the protocol, ICH E6 specifies no authorship attribution for the investigator's brochure or the informed consent form. Unlike the report and the protocol, neither document is posted publicly; both remain filed in the trial master file, so their authorship attribution is less likely to become an important issue in the future.
Regulatory authorship and journal authorship
When medical writers discuss authors and authorship, they usually mean documents in the public domain, such as journal articles or congress abstracts, whose primary purpose is to communicate a study's results to the scientific community and the wider world.
Although the investigators named as authors of a paper take public responsibility for its content, there is a tacit recognition, perhaps not always addressed in discussions of journal authorship, that the study design and analysis are largely done by the drug company. The investigators administered the study drug to the patients, but a team inside the company, with expertise in biostatistics, clinical science, clinical pharmacology and other areas, was responsible for the design and the administrative aspects. The company drafted the protocol, although investigators may have contributed to the design through steering committees and advisory boards. Likewise, the statistical outputs are produced within the company, and drafting the study report, along with public disclosure of the results on a clinical trial registry where appropriate, is also the company's responsibility.
Clinical development is a complex, collaborative process involving many company employees, hired external workers, and the investigators and other medical staff. A journal article describing an interventional trial generally aims to disseminate the results, and the limited number of authors who take public responsibility for its content may not accurately reflect the extent of the effort or the real intellectual input. A regulatory document authored and reviewed according to company guidelines, with input from the relevant departments, should instead provide an accurate and detailed description of the clinical development process and represent the company position.
Unlike journal articles, regulatory documents are generally prepared for submission to the health authorities and are not available in the public domain, although that may be changing with the shift towards greater transparency. Companies have well-defined standard operating procedures describing exactly who is responsible for reviewing or approving, or both, a document or a section of one. One experienced writer holds that the medical writer assigned to the project is not an intellectual author in the sense that the authors of journal articles are, or should be, but has a more technical role: mainly to compile and present data from a wide range of sources and to ensure that the regulatory documentation requirements are met. In that writer's view, there is as a result usually no need to name the writer as an author or to acknowledge their participation explicitly in a document that, if it has been through the appropriate review and approval cycles, should reflect a consensual company position.
The target audience, the health authority reviewers, is aware of the conflict of interest, since the company ultimately wants the drug approved, and so generally reads the documents with a critical eye. Audits can ensure the accuracy of the data, and the health authorities can ask further questions in many types of interaction. If a company has attempted to mislead the health authorities, the consequences, both financial and in loss of faith and credibility, can be severe. In one writer's experience, companies take their interactions with the health authorities very seriously, and "we can't be seen to be hiding anything" is a common sentiment in discussions about presenting data to them. The same writer's impression is that companies nowadays show plenty of apprehensive respect to regulators, who can make life very difficult for them.