Clinical Data Anonymization & Redaction - Health Canada Guidelines

With an aim to make anonymized clinical information in drug submissions and medical devices applications publicly accessible, Health Canada released a guidance document named as Health Canada's Public Release of Clinical Information (PRCI). Adhering to Canada’s Privacy Act, this document on clinical information enables independent re-analyses of data, completion of Health Canada’s Regulatory review process, fosters new research queries and aids the healthcare professionals, public, industries and other stakeholders to better understand the clinical aspects and make appropriate decisions about health.

The Scope & Application

In line with the amendments to the Food and Drug Regulations (FDR) and the Medical Devices Regulations (MDR), which came into force on February 28, 2019, this document specifies the clinical information in drug submissions and medical device applications that cease to be Confidential Business Information (CBI) following a final Regulatory decision and authorize Health Canada to publicly release this information. All the clinical information submitted to the Health Canada in the past as well as in future are applicable to these Regulatory amendments, which describe the scope of clinical information eligible for public release, the procedures to remove information that remains CBI and to protect personal information before the public release.

On a proactive basis, Health Canada intends to publish the information on drug submissions and medical device applications. As described in section 3.3 of this guidance, proactive publication of this information will be phased in four years. On request, Health Canada intend to make available the information in drug submissions and medical device applications that received a final Regulatory decision prior to the coming into force of the regulations (past submissions).

Drug Submissions Clinical Information

Clinical information of the drugs is submitted under division 8 of the FDR and provided to Health Canada in the internationally harmonized electronic Common Technical Document (eCTD) structure. Health Canada will release clinical information contained in eCTD modules 2.5 (Clinical Overviews), 2.7 (Clinical Summaries) and 5.3 (Clinical Study Reports), and in the following appendices to clinical study reports: 16.1.1 (Protocol and Protocol Amendments), 16.1.2 (Sample Case Report Forms) and 16.1.9 (Statistical Analysis Plan). The chemistry, manufacturing and non-clinical information will remain subject to the FDA’s definition of CBI, regardless of its location in a drug submission. Since 2019, Health Canada intends to proactively publish clinical information in new drug submissions and the information is not subject to proactive release and is available on request.

Clinical information may be found in the following drug submission types:

• New Drug Submissions (NDS)
• Supplemental New Drug Submissions (SNDS)
• Abbreviated New Drug Submissions (ANDS)
• Supplemental Abbreviated New Drug Submissions (SANDS)
• Extraordinary Use New Drug Submissions (EUNDS)
• Supplemental Extraordinary Use New Drug Submissions (SEUNDS)

Medical Devices Clinical Information

The clinical information for class III and IV medical devices is currently being received under section 32 (3) and (4) of the MDR. Class I and II medical devices are out of scope and the class I medical devices are exempted from device licensing. The clinical evidence for class II medical devices is not required to be submitted upon application for a medical device licence, except in certain circumstances. The manufacturing and other non-clinical information will remain subject to the FDA definition of CBI, regardless of its placement in a medical device application. Beginning in 2021, Health Canada intends to proactively publish clinical information in new medical device applications. Clinical information may be found in the following medical device application types:

• Class III Medical Device Application
• Class III Medical Device Application Amendment
• Class IV Medical Device Application
• Class IV Medical Device Application Amendment

Implementation Schedule for Proactive Disclosure

Health Canada plans to phase-in the proactive release of clinical information in drug submissions

and medical device applications that receive a final Regulatory decision following the coming into

force of these regulations (new submissions). Proactive publication of this information is expected

to be implemented on the following schedule:

Table: Implementation Schedule

Procedures

The clinical information is published under PRCI through five distinct phases such as:

Initiation: Prior to initiation of the publication of the clinical information, the sponsor may elect to attend a one-on-one Process Initiation Meeting (PIM) with the Health Canada to discuss the scope of clinical documents release and clarify Health Canada’s requirements and process. A PIM may be requested within the period of 120 calendar days prior to the final Regulatory decision, and up to 20 calendar days following the final Regulatory decision.

Submission: As mentioned in section 5 of the Health Canada’s guidance, for submission, the redaction proposal package must include annotated documents with proposed confidential business information (CBI) redactions and anonymization. The process of data anonymization should be detailed in a separate anonymization report.

Review: The justifications for each proposed redaction within the annotated documents will be reviewed by the Health Canada. Post review, the submitted documents will be accepted or rejected prior to finalization of the clinical information for public release.

Finalization: Following review, the manufacturer must submit a final version of the documents, according to the Health Canada instructions.

Publication: The final documents will be made publicly available for non-commercial purposes through Health Canada's clinical information portal.

Anonymization Process

To avoid serious possibility of identifying individual clinical trial patients, the clinical information must be adequately anonymized prior to public disclosure. Adopted from the 2016 Information and Privacy Commissioner of Ontario De-identification Guidelines, Health Canada’s guidance strongly recommends a 3-step process for anonymization. The risk of disclosing personal information can be reliably reduced by adopting an anonymization process that broadly follow the below mentioned steps:

Step 1: Identify and classify the variables

Prior to processing the anonymization of clinical information, directly and indirectly identifying variables must be classified. Directly identifying variables are commonly described as information that meets the test of being:

• Replicable - The variable is unlikely to frequently vary over time
• Distinguishable - The individual patients may have distinct recognizable values
• Knowable - Someone knows the variable associated with a certain individual

Other identifying variables that fall within the definition of 'personal information' within Canada's Privacy Act are the indirectly identifying variables. For anonymization, this variable disclosure must present a serious possibility of re-identifying an individual when combined with other available information (e.g. demographic data). These variables may be necessary for the purpose of understanding the clinical information and therefore their anonymization must be carefully justified, in line with the guiding anonymization principle number 2.

Step 2: Measure the re-identification risk

The data risk needs to be measured after the variables’ classification and this risk measurement justifies any data transformation that may follow. Variables which do not present a serious possibility of re-identifying an individual, alone or in combination with other information, are not considered personal information and should not be transformed. The overall risk of re-identification associated with the disclosure of clinical information is the product of the risk inherent to the data and the risk associated with the context of the release. For the public release of clinical information (to the general public), the calculation of risk of re-identification needs to reflect this environment; in a public release environment, the context risk is unreducible, so the overall risk of re-identification is equivalent to the risk inherent to the data (as opposed to the release of information to a small and select group of individuals, which would constitute a lower context risk, and therefore a lower risk of re-identification).

Step 3: Anonymize the data

Data Utility: The methodology used to anonymize clinical information can have a detrimental effect on its utility and the preserved data retains the greatest utility. Consequently, it is advisable to not transform (anonymize) variables that do not contribute to the risk of re-identification, and to adopt methods that have the lowest impact on data utility.

The directly identifying variables may be anonymized through the process of redaction, pseudonymization, or randomization. Examples of directly identifying variables that may be redacted include, names, initials, signatures, job titles/positions, addresses, fax numbers, email addresses, health plan beneficiary numbers, batch/serial numbers and telephone numbers.

Health Canada encourages the generalization of indirectly identifying variables, which may include, city, state/province, zip/postal code, demographic data (race, gender, etc.), medical history, serious adverse events, dates, height, weight, and BMI.

Documenting the Anonymization Process and Governance: To be equipped for necessary audit trail, the anonymization process should be thoroughly documented. Health Canada requests that the manufacturers must submit a completed anonymization report (template provided in Appendix F) with the submission of all anonymized clinical information.

The anonymization/redaction of documents as per Health Canada’s requirement will be best streamlined by Regulatory medical writing experts. Holding expertise in clinical data disclosure in Clinicaltrial.gov and EudraCT, the qualified working teams on disclosure management, information redaction of patient data for the EU region and multiple delivery models will cater successful deliverables, fill the process gaps and buffer the timelines for unforeseen delays. So, to effectively spearhead your submission and bring your content to commercialization with compliance, approach a Regulatory specialist armed with vast knowledge. Stay informed. Stay compliant.