Resource Library

Data Anonymization with Short Timelines & Missing Datasets

Performing Quantitative Risk Assessments Without Access To Trial Datasets

The sponsor organization started the PRCI process with an initial Process Initiation Meeting (PIM) with regulators. The requirements became clear: anonymization of study documents was preferred (vs. redaction), a quantitative risk modelling approach was highly recommended, and a vendor was required to support the end to end submission in an accelerated time frame. Access to trial datasets was limited, and so Real Life Sciences (RLS) was engaged for its ability to extract patient datasets directly trial documents using Real Life Sciences' RLS Protect By doing so in a highly compressed timeframe, the Sponsor was able to submit to regulators all the required quantitative risk assessments outputs and anonymized documents below the mandated risk thresholds all while maintaining data utility.

Putting in Place a Standardized, Repeatable and Scalable Process to Meet Increased Volumes

Ensuring that all Health Canada required datasets are present is a key step in order to decrease time to submission. The forces at play in disclosure are focused on data utility to enable additional use cases. Knowing what regulators are looking for in datasets and extrapolating existing datasets to fulfill requirements at the beginning of the submission journey saves valuable time and resource. 

Several key steps were and should be considered:

  • Health Canada PIM (Process Initiation Meeting) Preparation and Regulator Communication Guidelines - questions to ask/items to clarify to set up correctly with the end in mind
  • Quantitative Risk Modeling Framework - In this case, creating Sponsor specific templates to process datasets directly from documents.
  • Anonymization Reports - creating a reusable template that is approved for senior management - easy to reuse.

Meeting Unique Submission Requirements For Generic Drugs

Meeting regulatory demand involved implementing several key components. First, a Risk Modeling plan with a path to meet a 0.09 regulatory mandated acceptable risk threshold.  Meaning the acceptable risk of clinical trial participant reidentification, according to Health Canada needed to be 9% or less.  Additionally, to achieve regulatory compliance a Similar Trials approach was implemented along with a blueprint detailing how to document the methodology and results in an anonymization report. Along the way data utility was optimized by prioritizing key identifiers for disclosure vs removal.

Results

  • Generated  required Data from documents to produce quantifiable risk model.
  • Guided Sponsors team on full set of compliance requirements.  
  • Delivered submission docs and data  based on quantified models to regulators in stated timeframe.
crossmenu linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram