Tanya du Plessis, VP Data Solutions and Strategies at Bioforum and Rich Davies, VP Solutions Expert at CluePoints, explain how to apply the principles of RBQM to Data Management and the benefits of a risk-based approach.
Increased Pressures in Clinical Data Management
RBQM plays an important part in the way clinical trials are run today. It is at the core of everything from design to execution and conclusion.
This comes at a time when Clinical Data Management (CDM) is under increasing pressure from the five Vs – variety, volume, velocity, veracity and value.
We are seeing more and more systems being responsible for collecting data. Many of those data sources can also capture much more data, much faster, than we have been used to in the past.
As data managers, we are responsible for ensuring the veracity and value of that data is maintained. So how do we continue to deliver quality results under these pressures?
A risk-based approach can help us scale to meet these demands, move away from a one-size-fits-all approach and become more targeted and efficient.
RBQM – A Timely solution to Data Management Issues
RBQM provides some key opportunities for Data Management.
Currently, the focus is on tasks and their cycles. Just by going through these cycles we are able to put a tick in value-added box.
But there is an opportunity here to think about the outcome – to look at what value those tasks and cycles are actually bringing to the data.
TransCelerate BioPharma: Challenging the Value of Source Data Verification looked at around 1,000 trials and found only 3.7% of data was updated after initial entry into EDC. Source Data Verification (SDV) was shown to be responsible for even less of the changes.
The recent Stokman P, et al. Risk-based Quality Management in CDM study found only 1.7% of data is changed due to back-end DM queries.
These figures show that there is a real opportunity to pivot and focus on detecting and eliminating errors that matter within data, additionally creating an opportunity to use our existing resources to tackle newer and wider problems as outlined above.
If data managers are more engaged with the RBQM process from the start they can become more influential and help to design out issues before they occur.
Complete, Compliant, Consistent
When we think about CDM, there are three words that often describe what data managers are responsible for – complete, compliant, consistent.
We spend a tremendous amount of time ensuring we collect complete data sets across the different source systems, implementing rules laid out in Data management/validation plans and, checking for consistency.
A key benefit of RBQM for all stakeholders is focusing on data that matters, this might mean a shift away from a completeness and compliance focus for all data.
CluePoints’ SMART Engine provides a data-driven approach to look for nuanced patterns within data that can be representative of more impactful site issues.
It detects issues that matter by using the data against itself in a programmatic way and showing where you have differences developing for patients, sites, countries, or regions.
No rules are written – if data is being flagged it is because it breaks the trend of normal distribution compared to the rest of the data within your trial. This is a completely data-driven approach.
Getting Started with RBQM
Data is changing. The data you are using today could influence future trials, not just in terms of decision making but as part of the trial itself.
There are three principles we are applying when we look at RBQM – quality by design, critical data and processes and risk assessment.
We need to bring all three of these principles together. The main steps to get started are:
- Identify all critical data points expected to be collected for your study
- Determine if there are any protocol specific elements that need to be monitored
- For non-critical data, determine which data will fall under data surveillance and which reviews will only be triggered by risk activation
- Define Key Risk Indicators (KRIs)
- Define mitigations for risks as well as responsible role
- Define a company-specific quality control process
You then need to look at your edit checks. Getting the data in of a high quality is just as important as getting the data out.
Start by reviewing all your edit checks then define and discuss critical data with biostatistics.
Changing your processes can feel overwhelming but data managers are already performing targeted reviews through CRF design, edit check programming, and manual review focus. This is just the next step.
However, it does require a mind shift change. We no longer work in silos. A risk-based approach requires traditionally separate departments to work as one unit.
This might mean upskilling team members and updating review methodologies and tools.
Take a systematic approach. Go through each process and think about it carefully. Speak to companies that have already applied RBQM or vendors that can help you.
Benefits of RBQM in Summary
- Higher quality and integrity of data – the opportunity to target data with alternative, improved data cleaning techniques.
- Maximum value from available resources and shortened timelines – redundant/low value activities are removed, and more sophisticated and efficient data cleaning techniques are used.
- Financial savings– issues are likely to be detected earlier, or even avoided completely.
We know there are processes which have been embedded in companies for a very long time and these can be hard to move away from.
But, in our experience, once that lightbulb goes on, people start to see the value of RBQM and the organization-wide benefits it offers.
To listen to the full webinar presented by Tanya and Rich, click here.