Transformation is one of the most vital steps within the clinical trial submission processes. Intended to protect patient privacy while providing the highest level of utility, data anonymization can be time-intensive and difficult to handle in-house.
Historically transformation has been a manual process and there are a variety of data anonymization methods that each come with their own unique level of re-identification risk. Choosing the correct method for your team’s clinical study reports requires a thorough understanding of methodologies, as well as the resources and tools required to handle all data anonymization properly.
What is anonymization?
Clinical anonymization is the automated transformation process to mask patient identifiers in clinical study reports (CSRs) and clinical data sets that are shared both internally and externally.
Once data is collected, organizations are responsible for removing any Personal Identifiable Information (PII) from any CSR reports that are created. There are two options to do this:
Redaction, the process of eliminating PII from CSRs through redaction
Pseudonymization, the process of substituting PII from CSRs with pseudo identifiers
Both options come with their own levels of risk, particularly if the study was specific or small. Many organizations rely on a combination of methods that fall into both categories in order to ensure true patient protection and regulatory compliance. It is important to be able to quantify and adjust as needed this risk of re-identification to legally protect your company.
Why is anonymization important?
If sponsors share documents that are useless for medical innovation, even though they are compliant, the whole purpose of the activity is meaningless. Transparency is the gateway to medical research as sharing high utility data sets and documents drives targeted treatments and vaccines. Compared to redaction and qualitative assessment, anonymization leverages software to automate manual processes and provides a numeric score that ensures regulatory compliance.
Intended to promote transparency and data sharing in the clinical trial space, EMA Policy 0070, Health Canada Public Release of Clinical Information (PRCI) requires pharmaceutical companies to de-identify clinical trial findings and encourage anonymization.With retroactive requests for disclosure and proactive publication of clinical information for all new drug submissions (both NDS-NAS and those not categorized as new active substance), failure to properly protect patient data can result in serious consequences.
How do you make data anonymous?
There are a number of ways to make data anonymous. Each method has its benefits and can be used in combination with others.
Subject Date Offset involves altering identifiable dates related to an individual and applying a newly, alternative date throughout the study.
ID Scrambling uses new, randomly generated Unique Subject IDs, and applying to individual participants throughout the study.
The Hiding In Plain Sight (HIPS) Method is a method based on obfuscation, or the process of hiding information by making it ambiguous. This could include using realistic pseudonyms for names or phone numbers. Subject date offset is a means of hiding in plain sight.
Using Age Bands involves aggregating ages into singular categories of commonality, such as changing 26 to 20-29, or all ages under 65 into Age <65.
Geographic Clustering categorizes individual patients into groups based on geographical location, such as by city, state, country or continent.
Methods for K Anonymity involves the use of suppression or generalization to ensure that the Personal Identifiers released in a study cannot be distinguished from a pre-established set (k) of other participants.
Data and Document Anonymization Services by d-wise
Traditionally, many pharmaceutical companies have used redaction tactics to handle transformation. However, regulatory bodies are now pushing back on fully redacted components of CSRs and are promoting the use of sophisticated intelligent software to meet tight compliance timelines. Such software can be used to mitigate anonymization risks as well as leverage labor, cost, and time resources.
d-wise, the leader in transparency technology, enables sponsors to share data and documents on any scale with their outsourced Data & Document Anonymization Service offering. By combining a team of experts that leverage Blur, the #1 product in anonymization and quantifiable risk measurement, outsourcing transformation has never been simpler and safer.