Blog

The Evolution of CluePoints’ SMART Engine

By December 11, 2020 No Comments
Olivia Quinet, R&D Architect

How C++ Revolutionized the Core of CluePoints

With software playing an integral part of almost every device that makes daily life easier, the same is true for clinical trials. The pharmaceutical and drug development industry relies on intelligent software to make the most of technology, making clinical trials better, faster and cheaper for sponsors and CROs – and safer for patients.

At the heart of every CluePoints project lies the patented SMART™ engine, diligently ensuring the accuracy and validity of every single result. It is the core of CluePoints – without the SMART engine there is no CluePoints.

Data is knowledge and clinical trials rely on data. With so much riding on the validity and accuracy of results, ensuring that data quality is high has never been more important. High quality data means clinical research teams are provided with evidence-based facts to confidently move forward and make the best, most informed decisions. But how can data integrity be guaranteed?

CluePoints’ R&D Architect, Olivia Quinet, shares some of the secrets behind the evolution of CluePoints’ SMART engine and why this technology is the linchpin to data integrity.

Continuous Innovation

I started working for CluePoints over seven years ago and have assumed many roles, including python testing, performance testing and proof of concept, now working alongside the R&D team. I see myself as a “free electron”, given the freedom to develop and work on what we call the SMART engine “evolution” project.

As a one-of-its-kind solution, the patented SMART engine not only provides clients with assurance that they are working with the only method of its type, but ensures that CluePoints are firmly at the forefront of Centralized Statistical Monitoring (CSM) within the Risk Based Monitoring (RBM) and Risk-Based Quality Management (RBQM) space.

Over the years, the R&D team at CluePoints have refined and honed the SMART engine, responding to client need and meeting industry demand. It took a considerable amount of time to build the CluePoints solution due to the complex array of statistical algorithms used. The result is a powerful engine that can be used in all clinical trials.

 The SMART engine is central to CluePoints’ CSM solution which uses statistical methodology to identify unexpected or unusual patterns in clinical trial databases. The approach uses a large number of statistical tests and scoring algorithms to identify investigative centres whose data are inconsistent with data from other centres taking part in the same trial. The patented method is proven to be sensitive to a wide-range of errors in data, including inaccuracies due to equipment mis-calibration, lack of staff training and non-adherence to a protocol, as well as data falsification and fabrication. As such, CluePoints’ solution can support RBQM and quality assurance strategies, ensuring corrective measures and monitoring efforts are focused on errant sites efficiently, resulting in both reduced regulatory submission risk and costs.

 Porting the R Code to C++

A crucial part of software engineering is being diligent and analytical, able to trouble-shoot, and solve problems on the spot. These attributes were crucial as the team and I delved deeper into the workings of the SMART engine. We have been transforming the SMART engine a statistical function at a time, conducting feasibility studies and working with the analytic team to develop a complete C++ SMART engine. My scientific education gave me a unique advantage to explore the different facets of complex concepts within the SMART engine to propose innovative, “out of the box” solutions.

The SMART engine was originally developed in the R language, which provided many of the same benefits realised today, including precise numerical precision. However, several key aspects were identified that could be enhanced by utilising C++:

  • C++ produces a compiled executable which is specifically targeted to the server whereas R and python are interpreted languages. R and python code are executed through an interpreter. This has several impacts on performance.  C++ is about 10-1000 times faster than R or python.
  • C++ allows exceptional control of memory management which results in a much lower memory consumption.
  • Very efficient and reliable code can be produced via code testing.
  • There is no source code in production, only the compiled executable. This is crucial for protecting the intellectual property of CluePoints in case of a security breach. R or python would require the source code to be present in production.

Enhanced processing offers clients key benefits including speed and efficiency gains, high reliability, high response time and high availability. Analysing clinical data now takes less time – all tests can be performed in a few hours instead of days. The robust cloud-based system uses less memory and can cope with greater variability of data. All of this guarantees that clinical trial databases are reliably handled with precision, speed and accuracy, ensuring data integrity and offering confidence in the metrics that matter most for our clients’ studies.

 A Passion for Software (and the freedom to innovate)

Crucially the success of the C++ evolution project is down to my team of scientific programmers. They are the ones designing the technology. Applying our technical skills to the invention and innovation of software, along with the development, maintenance, and management of it, we were able to use our love for technology to make a difference to the SMART engine, and ultimately impact crucial clinical trials with worldwide implications.

CluePoints sees the value of making sure everybody clearly understands their role and how they contribute to the bigger picture. It is a very inclusive and accepting company where all team members proactively contribute and interact. Ultimately it is a shared passion for ensuring quality data, which is the key to successful clinical studies.

CluePoints is constantly evolving and adapting the SMART engine. New modules are frequently being developed and the C++ code is growing. Our solutions, all of which are underpinned by the SMART engine, are deployed to drive an RBQM strategy, as advocated by ICH and FDA. When we live in such a rapidly evolving digital era, shaped by a continuous stream of pioneering technological advances, we need to be ready to evolve. Powered by the agile SMART engine, CluePoints is primed to deliver what its clients and the industry need, now and in the future. It pays to be SMART.

 

Olivia Quinet

Olivia Quinet

Dr Olivia Quinet has been using the C++ language intensively in different scientific fields for over 2 decades. She holds a PhD in Quantum Chemistry from the University of Namur, Belgium. For 10 years, she did research on non-linear optics and also collaborated with the Quantum Theory Project, University of Florida. Afterward, she switched to the private sector as a C++ software architect to develop bioinformatics programs. She is currently working at CluePoints since 2013 and is responsible for the industrialization of R&D projects.