Trustworthy data science and good clinical practice

Responsible conduct and considerations of safety and ethics is essential in all of data science that has an impact on human lives.
Abzu® routinely helps customers maximize the success of their clinical trials. Common clinical use cases consist of optimizing the design of inclusion/exclusion criteria, understanding why patients respond to treatment, and identifying safe and accessible biomarkers for measuring response or remission.

Responsible data science for clinical studies

Responsible conduct and considerations regarding safety and ethics is essential for data science, especially when it has an impact on human lives. This is particularly the case for clinical studies when it comes to safety and confidentiality of the participants and ensuring the statistical validity of any analysis.

The QLattice® is designed from the ground-up to meet the quality requirements and concerns of responsible scientists in high-impact domains. The technology we have built is focused on creating reproducible and interpretable results that accommodate the needs of highly regulated environments, and we’re proud to contribute to scientific discovery by being part of Forskermaskinen and being approved by the UK Biobank.
Likewise, our team of qualified experts across the pharmaceutical value chain offer in-depth statistical and biological analyses and scientific support with a high degree of integrity. We provide numerous services across the pharmaceutical value chain from target identification to lead discovery and optimization of RNA-based medicines — all the way to pre-clinical and clinical.
This results in a holistic understanding of how to more effectively bring better and safer drugs to market at every step of the way in a safe and ethical manner.

For any clinical data analysis project, we particularly consider the following aspects of the GCP (Good Clinical Practice) guidelines. 

Good clinical practice

Data management

The clinical data stays on the machine where the clinical data analysis is conducted, whether you want to perform the analyses yourself or have us support you.

Therefore, the modeling can be carried out on a machine that adheres to the regulatory requirements for the individual project.

Depending on your needs and requirements, we have different options to choose from:

  • The computers of your scientists or the secure computers of our scientists.
  • Dedicated machines at our offices or in our ISO-27001 certified and GDPR compliant server cluster located in Germany.
  • A physical and dedicated GxP validated machine provided by you, located in one of our offices with secure access provided to specified team members.
  • A computer at your location, where specified team members either have physical or secure remote VPN access.

In short, you are in control of the access and level of confidentiality that should be considered for your data.

If you have requirements for a bespoke solution, reach out to us.

Participant confidentiality

In any cases where information is uncovered that could be used to identify individuals, we make sure to follow GCP guidelines and report the leakage. We also have internal capabilities to safely anonymize the data if necessary.

Data quality assurance

The outcome of QLattice models can be readily understood by clinicians and patients. The QLattice models have been classified as highly trustworthy by external assessors composed of disease experts and key opinion leaders. The models provide reproducible, transparent and highly interpretable results that can help identify inclusion/exclusion criteria for statistical analysis plans for clinical trials.

Additionally, we refer to CHP ICH E9 + E9(R1) (Statistical Principles for Clinical Trials) and every project with clinical data will be staffed by at least one GCP certified team member and supervised by one of our data protection officers.

Author

Tom Jelen

Tom Jelen

Tom is an experienced engineer in architecture and systems design. He has 15+ years’ experience in software development and distributed systems. In addition, Tom is certified in GCP and has 6 years of experience in building GxP validated software for various pharmaceutical companies in Denmark. Tom is addicted to bicycling. His bicycle is named Mia.

Share this article

Contact Abzu

It’s easy to get in touch with an Abzoid.

Contact form

Fill out the form below, and an Abzoid will be in touch in 24 hours.

Email us

Contact Sales at sales@abzu.ai or a scientist at science@abzu.ai.

Call us

Reach us during standard CET business hours at +45 31 23 47 64.