Deciding what is commercially confidential in a disclosed document
Commercially confidential information is the part of a clinical trial or marketing authorisation document that is not public and whose release could harm its owner's economic interest or competitive position, and a sponsor may propose to redact it from the version that is published. Regulators set out what they will accept as commercially confidential and how a redaction of it has to be made, and they have refused most of what sponsors proposed. Deciding what qualifies is work a sponsor settles inside the company, ideally before its documents are written.
What counts as commercially confidential
Health Canada's Public Release of Clinical Information, the EU Clinical Trials Regulation and EMA Policy 0070 require the disclosure of clinical trial application data, much of which contains sensitive and confidential information about a company's intellectual property, including its manufacturing methods, drug composition, suppliers' names and future development plans. These regimes require a significant amount of information about the trial itself to be disclosed, its results included, but draw a line at information that remains proprietary to the manufacturer.
Confidential information is information in the clinical trial that is not in the public domain or publicly available and whose disclosure may undermine the legitimate economic interest or competitive position of its owner. Health Canada calls it confidential business information, and the EMA calls it commercially confidential information. The EMA considers commercially confidential any information contained in a clinical trial application or provided during the trial's life cycle that is not in the public domain or publicly available and whose disclosure may undermine the legitimate economic interest or competitive position of the entities concerned, such as clinical trial sponsors, marketing authorisation applicants and holders, or service providers. Health Canada's confidential business information is, subject to the regulations, business information that is not publicly available, that the person it concerns has taken measures reasonable in the circumstances to keep from becoming publicly available, and that has actual or potential economic value to that person or their competitors because it is not publicly available, and whose disclosure would result in a material financial loss to the person or a material financial gain to their competitors.
The EMA lists what it will consider commercially confidential, and anything beyond that list is not considered so; for certain documents that go into the Clinical Trials Information System, it states that it does not expect them to contain any.
Development stage and the public domain
Less information is assumed to qualify as commercially confidential as a drug's development progresses, and once a decision on a marketing authorisation application has been made, whether approval, rejection or withdrawal, no information in the clinical study report should be considered commercially confidential. Information already in the public domain cannot be considered commercially confidential, so a literature search for publicly available information is a standard part of the redaction process. With trials transitioning to the Clinical Trials Regulation, data are more likely to be published prematurely, particularly on sponsor websites, through conference presentations and in the sponsors' own articles, and information even partially in the public domain can affect the planned redaction strategy for information-dense documents such as the protocol. A dose-escalation scheme for one cohort presented at a scientific conference and made public as slides, for example, could compromise a sponsor's intention to protect the overall dose-escalation plan as commercially confidential. The transparency rules introduced with the Clinical Trials Information System have heightened awareness of the importance of appropriate timing in disclosing any full or segmented information related to an active clinical trial.
Vendors' intellectual property
Standard contractual clauses between sponsors and vendors usually approve the use of copies of the vendor's intellectual property in regulatory submissions. Some transitioning trials signed their vendor contracts before they were aware that all patient-facing material related to study endpoints would become public once the trial was posted. Their vendors agreed that regulatory authorities would review their intellectual property, but may not have been informed of, or consented to, its release to the public. Because public disclosure could undermine a vendor's economic interest or competitive position, the vendor's standing has to be reassessed in these circumstances.
In the marketing authorisation dossier
With the guidance of the Heads of Medicines Agencies and the EMA on identifying personal data and commercially confidential information within the structure of the marketing authorisation application dossier, first adopted in 2012, European regulatory authorities agreed for the first time a common and consistent approach to which parts of a dossier of a medicinal product for human use can or cannot be released to the public, whether the medicine was authorised by the centralised, mutual-recognition or decentralised procedure. A draft update of that guidance was put out for public consultation in April 2024. By 18 December 2024 the EMA and the Heads of Medicines Agencies had published a comprehensive overhaul of the guidance.
As a general rule, the overwhelming majority of the data in an application is not commercially confidential; the exceptions mainly concern the manufacture of the medicine, facilities and equipment, and some contractual arrangements between companies, and general quality information, considered confidential under the initial guidance, is now mostly releasable. Instead of a yes-or-no rule on whether a whole section can be released, the updated guidance treats information as releasable by default and gives practical orientation on which specific points within each section could be redacted or anonymised, with an updated annex of examples of what may be commercially confidential or protected personal data. It also sets out how personal data that can lead to the identification of a person will be protected, taking into account the General Data Protection Regulation and the corresponding regulation for EU institutions, bodies, offices and agencies (EUDPR), and how to identify the personal data of experts, staff or patients that should be anonymised.
How a redaction is proposed
Health Canada lets a sponsor request, with justification, that information be redacted if the sponsor believes it meets the criteria for confidentiality. Because a justification may have to draw on internal corporate plans, such as a new indication being developed from secondary outcome data, Health Canada asks for an annotated version of all the clinical information in scope of publication, with every proposed redaction highlighted for its review. Text proposed for redaction must remain legible, and each proposed redaction should be justified in detail on a Proposed Redaction Control Sheet in CSV format.
Under the EU Clinical Trials Regulation a sponsor may mark commercially confidential information in the non-public versions of documents for the reviewers' awareness, and the EMA may ask for more information to support the justification for confidentiality. The sponsor may redact commercially confidential information from the "for publication" versions of documents posted to the Clinical Trials Information System, while the "not for publication" version remains unredacted. The revised transparency rules of the Clinical Trials Information System, adopted by the EMA Management Board on 5 October 2023 and implemented on 18 June 2024, removed the option, available until then, of deferring a document's disclosure to protect commercially confidential information. Under those rules, redaction is recommended to protect personal data and commercially confidential information. The relaunch of Policy 0070, rolled out gradually from September 2023, includes proactive notification of expected confidential information before the draft submission.
The EMA's rules on how a redaction is made are strict. Commercially confidential information may not be redacted as a whole sentence or a whole paragraph: where a dose has to be protected, the numbers are redacted and the units are not. One sponsor took these rules into account when it wrote its investigator's brochures and protocols.
How proposals have fared
Rejection rates
After one year of Policy 0070 clinical data publication, proposed commercially confidential information was rejected in 76% of instances, the most frequent reasons being insufficient justification, followed by the information being available in the public domain. Submitting a justification that the EMA accepts is extremely challenging. On the first year's experience, the likelihood that the EMA accepts proposed commercially confidential information is very low, with a success rate of 0.01% of all published pages, a figure counted in pages rather than in instances.
Health Canada has historically rejected 65% of the confidential information sponsors proposed, and the EMA 60%, because the information marked as confidential was already in the public domain, was not innovative, or was in the public interest.
What the EMA accepted
The EMA's first report on Policy 0070 clinical data publication, covering October 2016 to October 2017 and published on 16 July 2018, shows that the EMA accepted a quarter of the proposals of commercially confidential information that applicants made, so redaction of that information was allowed. Everything accepted related to pharmaceutical development, and the reasons fell into two broad categories: quality, with 49 instances accepted, and clinical, with 60. In the quality category, 34 of the 49 accepted instances concerned the quantitative composition of the excipients of the finished or investigational product, and the remainder covered the quantitative composition of the active, the quantitative and qualitative composition of the excipients, manufacturing process items including quantitative specification and acceptance criteria and batch size and manufacturing facility, and dissolution-profile acceptance criteria relating to the comparability of different formulations. In the clinical category, 24 of the 60 accepted instances concerned detailed information on analytical methods and assays; the other categories included several kinds of future development plans, among them other products, future indications, new treatments, results and plans not yet validated, as well as financial items concerning study subjects or healthy volunteers and post-marketing exposure per country. Understanding which types of information might be accepted for redaction helps a writer better judge which commercially confidential information, if included in their reports, might reasonably be expected to be accepted for redaction.
The standard of justification
Policy 0070 states that, in general, clinical data cannot be considered commercially confidential. What may be accepted as commercially confidential is a matter of considerable debate and remains a case-by-case assessment, particularly given the lack of a legal definition of it. In three Policy 0043 cases, the EU General Court decided that marketing authorisation holders need to provide concrete evidence of how the release of the contested documents would undermine their commercial interests.
On 6 February 2018 the EU General Court delivered three rulings upholding the EMA's decisions to release documents requested under Regulation (EC) No 1049/2001, the Transparency Regulation, the first occasion on which the Court of Justice of the European Union had pronounced on that regulation's application to documents the EMA holds. The three concerned the disclosure of similarity and superiority reports on an orphan medicine prepared by the EMA's human medicines committee, the disclosure of a clinical study report, and five toxicology study reports for a veterinary medicine; in each, the company challenged the EMA's decision to release the documents in accordance with the Transparency Regulation and the EMA's 2010 policy on access to documents, Policy 0043. The Court noted that the companies had failed to give any concrete evidence of how releasing the contested documents would undermine their commercial interests, and rejected their claims on that ground; the EMA said it would continue to assess each request individually under the regulation and its own policy.
Identifying it inside the company
Ownership and the agreed list
A sponsor has to decide internally, with its legal department, what it considers commercially confidential, and different compounds or assets within one company may carry different sets. Each team or compound comes up with its list of commercially confidential information, working with the legal department, and the list is distributed top down to the writers, so that writers already take care not to include it while writing a document rather than going back afterwards to find it. The discussion of commercially confidential information started with the Clinical Trials Information System, where the investigator's brochure was very contentious; the brochure later dropped out of the picture, but the protocol remains, so the patent and intellectual property specialists need to be involved.
Reaching a consensus decision on confidential references in a trial document for an asset, programme or study is a common challenge in global pharmaceutical teams. The identification of confidential references is typically shared among intellectual-property lawyers, medical writing, regulatory, safety, clinical and nonclinical teams, and transparency and disclosure; with no single owner, decisions have proved difficult in some organisations, which have to balance protecting the information against the risk of delay if a regulator pushes back. Tracking confidential information accurately can be challenging for three reasons: a classification changes over time, awareness is needed of publications in the public domain that may refer to something once deemed confidential, and the global trial registries have to stay consistent in what each discloses.
A repeatable process
Many manufacturers are examining their processes to minimise references to confidential data while documents are authored, and to track and identify confidential information throughout the study lifecycle. This has created a catalyst for medical writing, transparency, legal and regulatory teams to collaborate and enhance their processes, minimising disruption to trial approval while embracing the opportunities to share more with the public.
Some operational considerations may help set up a repeatable, scalable process for identifying, managing and tracking confidential information consistently across studies, especially for sponsors submitting through the Clinical Trials Information System. First, procedures defined for the organisation and its team structure: an authoring minimisation initiative that keeps the references a document needs and cuts their repetition; a consistent deferral strategy and decision process; documented hand-offs between the internal teams, the contract research organisation and the redaction provider; set turnaround times for redaction, review, approval and final versions; clear roles for reviewing and approving confidential information; and resourced plans for answering requests for information. Second, a centralised library listing the confidential information for each asset, programme or study, reviewed and approved with leads from safety, manufacturing, intellectual-property law, the corporate librarian and transparency, maintained because what is confidential today may not be tomorrow, and consulted while documents are drafted. Third, authoring standards: templates revised to carry the minimisation expectations, current confidential information kept in view throughout drafting, and documents grouped for prioritisation, review and approval, with the dense documents such as the protocol and the investigator's brochure in one group and those with few references, such as the principal investigator's CV, proof of insurance and proof of payment, in another, and with use scenarios defined for the regulatory document management system.
Minimising it at drafting
Minimisation techniques at drafting are recommended to prevent the unnecessary or excessive redactions that may expose inadvertent errors or missed redactions. Include only the content the document's objective requires; highlight confidential information at the time of writing for the transparency or redaction team; do not duplicate content but cross-refer, as in "refer to Section A", so that each item appears fewer times; and do not reference commercially confidential information in the table of contents, body, headers and footers, or table titles. A company-wide shared understanding of which information is commercially confidential is essential, and cross-functional colleagues in regulatory, safety, manufacturing, intellectual-property law and transparency must share a common understanding of what is and is not confidential. That understanding is built from practical examples in earlier public versions of documents, with the justification and decision process explained and with negative examples as well as positive ones and the reasons for each, and by reviewing the current library with the cross-functional team leads.
Clinical trials carry their own publication requirements on the Clinical Trials Information System portal, independent of any Policy 0070 procedure, so one view is that protocols and clinical study reports unrelated to a Policy 0070 procedure are worth writing with disclosure in mind from early on, limiting information that does not really have to be disclosed. Where there is commercially confidential information to process, one recommendation is to keep, from the early stages, a tracker of what is really relevant to protect and of where that information is found in the documents produced.
The justification table
Identifying and justifying commercially confidential information is a complex, multidisciplinary task that needs review by and input from many subject-matter experts. All clinical reports in scope of Policy 0070 have to be screened for it, and the number of pages to review can be substantial, because cross-referenced clinical trial reports submitted with previous applications may also be in scope. For each file that contains it, a sufficient and relevant justification has to be given in a justification table, and the information proposed for redaction has to be checked and confirmed as not yet publicly available.
One sponsor ran the identification in two ways, for two dossiers. For the first dossier, it consulted subject-matter experts from clinical development and pharmacology, intellectual property, bioassays and immunology, chemistry manufacturing and control, regulatory affairs, statistics and non-clinical development. The transparency and disclosure team briefed them first on three tests: that the information is covered in the policy's own annex, that it is not listed in the external guidance as information not considered confidential, and that it does not meet any of the EMA's five rejection codes. For each item, the experts were asked to supply a specific, pertinent, relevant, not overstated and appropriate justification explaining how release would damage the company's commercial interest. They highlighted their suggestions in the PDFs and put the justifications inside the highlights. The transparency and disclosure team then worked with them to verify that each item was not public, to shorten lengthy suggestions to a succinct word or number, and to improve the justifications, with quality control checks through and across documents aimed at marking items consistently. As a final, time-consuming step, the team created the justification tables with the vendor and converted the highlights into correctly formatted redaction proposals.
Earlier decisions on commercially confidential information for documents requested under Policy 0043 but not in scope of Policy 0070 may be relevant to a Policy 0070 dossier where the same content appears in documents in its scope. To keep redactions rejected under Policy 0043 out of its Policy 0070 proposal package, the sponsor created a master list of the items accepted or rejected under Policy 0043 as a reference. The EMA's decisions on one relevant Policy 0043 request arrived too late to be considered for all applicable documents, and this, with the large number of scanned pages and the number and complexity of the items the experts suggested, prevented full consistency of the proposed items across the documentation. The EMA's decisions on what was acceptable were also not consistent between the two policies, so further discussions with the EMA and modifications were needed after the redaction conclusion and the submission of the final redacted package.
For the second dossier a modified process was tested. Subject-matter experts added their suggestions and justifications to a single master justification table for the whole dossier, the verification was performed against that table, and only the remaining items were then marked for redaction in the PDFs. This process was much more manageable than the first, but the second dossier was small, with few suggested items, and the real test would be the company's next large dossier for an innovative product.
Many of the sponsor's proposed redactions were rejected, mainly because the justifications were not considered sufficient, because the information was in the public domain, or because it was considered common knowledge. Even so, in most of the first dossier's 30 documents some items were accepted, most of them manufacturing details and immunological bioassay specifications. The external guidance led the team to expect a consultation with a chance to clarify or elaborate on particular justifications, but apart from a request for further information on two items, the EMA proceeded straight to the redaction conclusion.
Redaction tools
Redaction of confidential information from regulatory documents may be laborious and error-prone without the proper tools, and study teams may struggle with tight turnaround times when many documents have to be processed for submission deadlines and the trial approval schedule. A purpose-built redaction solution can help expedite and streamline redaction, review and approval, with capabilities such as study-specific saved searches for terms and references, applied to one document or several at once; collaboration inside the document instead of non-secure messaging and emailing of sensitive files, which also avoids version-control problems; a systemic audit trail of every redaction mark and justification; and integrated project tracking of progress and expected completion.