With increasingly digital supply chains, the amount of global data is rapidly increasing. For pharma and biotech companies, data determines the safety, efficacy, and availability of life-saving drugs — which is why it’s so important to have confidence in your data at any point throughout its lifecycle.
That’s where data integrity comes in.
Data integrity is having the confidence and assurance that data — from its acquisition to destruction — has maintained all its contextual information, including any changes or modifications.
Data is vital to patient safety. For example, if you release a drug that is more or less potent than it should be, there can be adverse (or even deadly) health effects for patients. Data can also impact drug shortages, making it difficult for patients to access life-saving medications when they need them. Furthermore, regulatory authorities need to trust your data is accurate, transparent, and contextually understood. Repeat data integrity issues can break trust with health authorities and even inhibit your ability to apply for drug approvals in the future.
For drug companies, your reputation always traces back to data integrity. Often, reputational damage can’t be recovered. Developing and maintaining a robust data integrity program assures patients, medical professionals, and health authorities that your product is safe, effective, and available.
Understanding the lifecycle of your data is vital to data integrity, ensuring that your data retains accuracy and context at every stage of its journey.
The first stage of the data lifecycle is its creation. Life science data can come from a variety of sources, including clinical trials, patient testimonials, and adverse events. It can also be acquired from other systems into a larger data system. Data can be collected by manual entry, directly from a device, or transmitted from a different system, either electronically or on paper.
After data is created, it moves to an active stage where it’s utilized and maintained. In this stage, your data is used for decision-making based on its purpose or intent. This could mean using your data for release decisions, to determine safety and efficacy in a trial, or to establish if you had a robust manufacturing run. While your data is active, it can be modified, compiled, or moved. It’s important that any modifications are tracked, so there is a record of context behind each change.
After data is used, it must be stored in one or more data management systems with the appropriate security to ensure traceability and retrievability. During this stage, your data becomes passive, but it’s essential that it’s protected and able to be accessed later.
With more than 10,000 data centers worldwide, data destruction is an essential piece of the data lifecycle. It doesn’t always make sense to store an ever growing repository of data. That’s why it’s important to articulate why, when, and how you destroy certain data.
There are several factors that impact the integrity of your data. Regardless of whether you use a paper or electronic system, every stage of the data lifecycle relies on how your people, processes, and technology intersect.
Most data issues don’t stem from intentional falsification. Instead, many issues come from human error, or how your people intersect with your processes and technology. For example, data goes through a natural review process for corrections, either manually or through audit trails. It’s important that you capture who is making each correction and why, so you have contextual information down the road.
When refining your data integrity program, it’s important to pay attention to how an operator or frontline person interfaces with your data. That may be through equipment or patient interactions. Understand how data acquisition happens, and ensuring that all supporting systems actually enable accurate data collection and processing. In an electronic system, you can create automatic checks, drop down selections, and parameters to reduce the chance of human error, improve consistency, and support grouping and sorting later on. If you use a paper system, consider your user from the beginning so you can design the flow of data in a way that makes sense.
It’s vital that you can control and explain any data modifications or changes that happen throughout your data’s lifecycle. Regulatory bodies will want to see where modifications happened, who made them, and why. In an electronic system, you can use audit trails to track data modifications. In a paper system, you must rely on people to not only make corrections, but also to explain the context around changes in a way that can be understood years later. In this case, provide context to your team on how to structure an informative comment.
Storage space is no longer a premium cost, so we tend to keep more data than is necessary. But what happens when you’re done with your data? It’s important to define parameters around data destruction. How much do you keep? When do you purge data from your systems? At what point is data no longer useful? If your data is 10-15 years old, it’s likely that your processes or product have changed and it’s no longer relevant.
With electronic systems, there’s an additional element of data to consider in your destruction plans: metadata, or the data around your data. Metadata captures when and how your data was acquired, and provides checks and balances that data hasn’t been manipulated across systems.
While many drug companies are shifting to electronic systems, historical paper systems often require a hybrid approach. Inherently, hybrid systems run the risk of paper and electronic systems getting out of sync. For example, paper records could lag behind what is live in the electronic system. When building your data integrity program, consider the risks of operating in a hybrid space vs. going all in on digital. Think about where data could lose its integrity throughout its lifecycle and determine what countermeasures are necessary to mitigate risk.
Increasingly, life science companies are using artificial intelligence (AI) and machine learning (ML) to augment and automate various processes. Moving forward, the concept of data integrity will have to evolve to include how data is processed through ML and AI. In 2023, we saw health authorities issue guidance and white papers on AI and ML for both drug and device development. While many of these are still fairly high level, these types of documents provide insight into a health authority’s current thinking with respect to the application of AI and ML in our drug development process. Further exploration of AI and ML technology could reveal unfoeaeen threats to the quality and integrity of our data, which will require countermeasures to ensure data integrity.
At Pharmatech Associates, we work with companies to take a step back and architect a data integrity program from a data governance perspective, where there are clear owners, data controls, and holistic processes in place.
Many companies make the mistake of approaching data integrity by looking at how to avoid data integrity observation or data integrity issues. We look beyond a “check the box” mindset to see how a streamlined data integrity program can improve operational efficiency and increase the reliability of your product.
No two data programs are the same. While we have proven data governance policies and structures, Pharmatech approaches each client with a tailored model to offer an unbiased, third-party look at your program. We focus on the intersection of your people, processes, and technology to help you see the forest from the trees.
Contact us today to learn more about how we can streamline your data integrity program, improve operational efficiencies, and increase the reliability of your data throughout its lifecycle.
Our unique approach can help your company navigate the complexities of launching a drug, biologic or medical device into multiple markets. The first step? Connecting. Simply fill out our form and a representative will follow-up shortly.
1.877.787.0177
510.732.0177
22320 Foothill Blvd. Suite 330, Hayward CA 94541
