Blur

Blur™ is a standalone, cloud-based SaaS platform developed by d-wise (an Instem company) for de-identifying patient information from clinical trials. Built in collaboration with enterprise pharmaceutical sponsors, Blur combines dedicated role assignments, natural language processing (NLP), and simulation-based risk assessment to ensure patient privacy, regulatory compliance, and high data usability. It is designed for non-programmers, making it accessible to a broad range of clinical trial professionals including data managers, biostatistics professionals, and study leaders who need a defensible, scalable anonymization solution.

Traditional approaches to de-identification are often slow, error-prone, and ill-equipped to handle the volume and complexity of modern clinical trial data. Blur addresses these challenges by automating manual tasks, providing comprehensive PHI detection and redaction, and delivering audit-ready documentation — all from a single, user-friendly interface. The platform is purpose-built to reduce regulatory risk, accelerate trial timelines, and protect both patients and organisations from the consequences of non-compliance or data breaches.

Key Benefits

• Boosts Efficiency: Blur provides a centralised platform for concurrent de-identification, allowing users to manage all tasks from a single interface and reduce time spent switching between tools.
• Saves Time: Automated detection and redaction of patient health information (PHI) accelerates document preparation and reduces the time required for regulatory submissions.
• Removes Risk: Automation and NLP ensure comprehensive coverage for de-identification, protecting the most sensitive information. Robust user logs and documentation promote easy auditability and legal defensibility.
• Improves Compliance: Blur delivers compatibility with global regulatory standards including Health Canada, US FDA, EMA, EU CTR, and GDPR, ensuring each trial meets strict privacy laws.
• Maintains Data Usability: Users can transform data into other values and forms, preserving data utility without revealing confidential personal information.
• Simplified Collaboration: Role assignment capabilities simplify task distribution among team members and ensure clear accountability across concurrent workflows.

Core Modules

• Blur Data: Efficiently removes or masks PHI in clinical datasets, ensuring HIPAA, GDPR, and global regulatory compliance. Enables researchers to achieve efficient de-identification with confidence.
• Blur Risk: Calculates re-identification risk using simulation-based scoring to help users choose the safest, most data-preserving anonymization methods. Blur Risk performs one million simulations in ten minutes, enabling quick decisions that balance privacy with data utility.
• Blur CSR: Leverages natural language processing to anonymize Clinical Study Reports (CSRs) for submission under EMA Policy 0070 and Health Canada PRCI, including text, tables, and embedded images. Reduces the complexity and stress typically associated with regulatory submissions.

How Blur Works

1. Data Integration: Blur enables de-identification of data across different formats, including CSRs imported using Blur CSR.
2. Assign Roles: Assign roles such as De-Identifier, Reviewer, or Inspector, and tune settings to match the requirements of specific regulators.
3. De-identification: Blur CSR ML uses natural language processing to de-identify and anonymize data and documents, with capability for bulk de-identification and anonymization projects.
4. Risk Assessment: Blur Risk enables rapid risk identification to optimise de-identification strategies.
5. Log Data: Securely store all changes and user activity logs for audit readiness.
6. Generate Reports: Export clean, anonymized, and audit-ready files for regulatory submission with Blur CSR.

Customer Success: Anonymizing Legacy Clinical Trial Data

• A client project consisted of 9 datasets, more than half containing over 10,000 variables each, totalling more than 60,000 variables to inspect — far exceeding the typical maximum of 1,800 variables seen in a single de-identification project.
• The client was storing data in a legacy format created by their proprietary system, making it initially difficult to locate demographic data during the anonymization process.
• Using Blur, d-wise worked with the client to identify and clean up excess data, inspect variables, and fully anonymize the dataset. The project was completed within one month, extended from the typical one-to-two-week timeline due to the scale and cleanup required.
• The client now has anonymized data ready for academic sharing, enabling them to share data with other researchers, clinical research data sharing platforms, or fulfil retroactive regulatory requests.

Blur is deployed as a cloud-based SaaS platform requiring no installation; users can securely access it from their work network. It supports concurrent workflows with role-based permissions, and clients have access to expert support through Instem's Clinical Trial Transparency services. Blur sits within Instem's broader Clinical Trial Analytics and Transparency solution portfolio, complementing tools such as Accel, Aspire, and Clinical Trial Transparency (CTT) Services.