The European Medicines Agency (EMA) issued guidance on the implementation of its Policy 0070 on the publication of clinical data for medicines, including with respect to anonymization of clinical reports for publication. As background, please see our previous briefing on the EMA’s new approach to transparency of clinical studies here. As of October 2016, all drug manufacturers that are making a marketing authorization application under the centralized procedure in Europe will be subject to the new guidance.1
The EMA & Policy 0070
Similar to the US Food & Drug Administration, the EMA ensures that medicines sold in the EU markets under a centralized marketing authorization are safe and effective. EMA Policy 0070 provides for the proactive publication by the EMA of clinical data from clinical reports submitted to it. The clinical reports will be made available to the public for general information, academic use and other non-commercial purposes. One of the underlying concerns in this shift toward transparency is the privacy of clinical study subjects.
Guidance Regarding Anonymization of Clinical Reports
Balancing subject privacy and transparency presents drug manufacturers with a difficult task—how to increase transparency of clinical studies while also attenuating the risk of subject reidentification. In its guidance, the EMA discusses three approaches to anonymization of clinical reports:
- Masking – Described as the simplest method, masking is accomplished with a redaction tool that scrubs specified information.
- Randomization – Randomization changes the data so it is less identifiable to an individual.
- Generalization – This method dilutes the “attributes of the data.” For example an individual’s name could be substituted with an age range.2
These anonymization techniques can be used separately or in combination. One industry commenter suggests that manufacturers should focus on anonymizing patient-level data to ensure that downstream uses of that data are secure, including use in clinical reports, and maintains that this method is superior to ad hoc masking of the data, which may be more prone to error. However, the EMA’s guidance does not mandate use of any specific methodology.
Whilst the first phase of implementing EMA Policy 0070 has been focused on publication of clinical data which relates to clinical reports only, the second phase will involve the EMA finding the most appropriate way to make available individual patient data in compliance with privacy and data protection laws. Pharmaceutical companies should anticipate more calls for transparency in clinical studies. At the same time, organizations should continue to be mindful of best practices when it comes to anonymizing and securing data. Future compliance may require organizations to anonymize patient-level data at an earlier stage in the clinical trial process. In anticipation of those changes, companies should proactively develop policies and procedures for implementing long-term solutions to anonymization.
Applications which must be submitted through this procedure, and not national procedures, are: certain biotechnology products; advanced therapy medicinal products; new active substances with certain therapeutic substances (such as cancer); and orphan medicines.
These techniques are consistent with the Article 29 Working Party’s Opinion 05/2014 (WP216) on Anonymisation Techniques.