In the rapidly evolving landscape of artificial intelligence (AI) and machine learning (ML), maintaining data integrity is more critical than ever, especially in highly regulated life sciences industries. As AI and ML technologies are used more frequently, the accuracy and reliability of data become fundamental to ensuring safe and effective outcomes. 

This article explores examples of AI and ML use cases in life sciences, and tips for how organizations can apply ALCOA principles to help provide a foundation for achieving data integrity, and ultimately improve decision-making in the age of AI.

AI and ML Data Processing in Drug Development

AI and ML technologies, adept at processing large datasets more quickly and accurately than human capability, have already significantly influenced drug development. The U.S. National Library of Medicine cites: “Artificial Intelligence (AI) has recently started to gear-up its application in various sectors…with the pharmaceutical industry as the front-runner beneficiary.”

AI and Chromatography: Enhancing Chemical Analysis

As one example, a statistical technique for purifying and analyzing chemical mixtures is a process called “chromatography”. Traditionally, analyses of chemical mixtures were dependent upon manual interpretation. AI can now automate the integration of chromatographic peaks, which greatly enhances the accuracy of this analysis which ensures purity and appropriate concentrations. Among many other uses in the pharmaceutical industry, AI also accelerates drug discovery via computational simulations that can help predict interactions between certain molecules and target proteins. 

AI’s already-prevalent and increasing use in streamlining drug discovery, quality control, clinical trials and post-market surveillance should translate into faster availability of, and yet still safe, treatments for us as patient-consumers. At the same time, these AI applications are hugely dependent on vast amounts of data. By now it’s common knowledge that the data for AI applications (regardless of  the industry in which they are used) should be complete, accurate and unbiased, otherwise decisions and predictions made using flawed data could have serious consequences, and especially for drug manufacturers and the patients they ultimately serve.

Moreover, AI, combined with cloud computing, is also reshaping data management by enabling advanced process monitoring and control. As with all change, opportunities are often accompanied by complexity. While the adoption of these advanced technologies brings significant benefits, it also presents challenges concerning data integrity and security.

FDA AI/ML Regulation and Industry Engagement

Regulatory bodies like the FDA and EMA require that data be reliable and accurate, and that companies implement effective strategies to mitigate data integrity risks. FDA programs such as the Emerging Technology program, and the Advanced Technologies program actively engage with the pharmaceutical industry on AI and ML use cases in pharmaceutical manufacturing, to keep pace with and track their deployment. 

As well, the FDA has organized to work with industry to create AI/ML regulations that will impact both pharmaceuticals and medical devices, as it notes in its whitepaper on using artificial intelligence and machine learning: “To fulfill its mission of protecting, promoting, and advancing public health, the Food and Drug Administration’s (FDA’s) Center for Drug Evaluation and Research (CDER), in collaboration with the Center for Biologics Evaluation and Research (CBER) and the Center for Devices and Radiological Health (CDRH), including the Digital Health Center of Excellence (DHCoE), has published a document to facilitate a discussion with stakeholders on the use of artificial intelligence (AI) and machine learning (ML) in drug development, including in the development of medical devices intended to be used with drugs, to help inform the regulatory landscape…”

In addition to working with industry, the FDA is also reaching out to collaborate with patient groups and international regulators to formulate a patient-centered regulatory approach for AI, soliciting input on topics like cybersecurity and quality assurance. These initiatives are outlined in the FDA’s whitepaper, “Artificial Intelligence & Medical Products: How CBER, CDER, CDRH, and OCP are Working Together”.

ALCOA Framework is Still Ideal for Achieving Data Integrity for Today’s Applications

In today’s environment, where data integrity is exceptionally crucial – and not just for highly regulated industries – the principles of ALCOA provide a great framework for ensuring data integrity. The food and beverage industry has also considered use of ALCOA practices. ALCOA is an acronym which suggests the attributes ideal for achieving data integrity, and it stands for “Attributable, Legible, Contemporaneous, Original, and Accurate”.

ALCOA was first used within the context of Good Manufacturing Documentation practices for the pharmaceutical industry in the early 1990s. This was to ensure the reliability and authenticity of data, as the industry grew in its reliance on digital records and automated processes. Data authenticity was required by regulatory bodies then, as it is today. Over time, ALCOA was later expanded to ALCOA+ which includes the principles of  “Complete, Consistent, Enduring, and Available”.

ALCOA+ principles are now used in the healthcare, biotechnology, medical device industries, and its use is also being advocated for in the food and beverage industry.

Tips for Using ALCOA Principles to Enhance Data Quality in Your Organization

The ALCOA framework’s focus on data quality makes it useful for any industry where data integrity is especially needed and/or required – which given the proliferation of AI/ML uses, means quality data is the way to properly ensure the achievement of desired outcomes based on accurate inputs and for making quality decisions.

Here are some recommendations for implementing ALCOA principles in any business:

1. Train Employees: Educating staff on the importance of data integrity and the ALCOA principles is an important first step, and should be the foundation for establishing a culture of ALCOA and ALCOA+ practice. There is a lot of information available that powerfully illustrates the kinds of false or inaccurate results from the use of incomplete data. Not having a broad enough data set can be especially challenging for pharmaceutical companies, for use cases such as clinical trials. One good read on the topic is, “Unmasking AI: My Mission to Protect What is Human” by Dr. Joy Buolamwini. A book to help teams organize data integrity practices in alignment with regulatory requirements is, “Data Integrity in Pharmaceutical and Medical Devices Regulation Operations: Best Practices Guide to Electronic Records Compliance” by Orlando Lopez.
2. Good Documentation Practices: Establish robust documentation practices that align with ALCOA principles.
3. Regular Audits: Conduct regular internal audits to ensure compliance with data integrity standards.
4. Digital Solutions for Data Collection: Utilize digital tools and technologies to automate data collection and ensure accuracy and legibility.
5. Data Management Systems: Implement comprehensive data management systems that support the ALCOA+ principles, ensuring data is complete, consistent, enduring and available.
6. Quality Management Systems: PubMed, from the U.S. National Library of Medicine advises life sciences companies to integrate data integrity into quality management systems, and notes the advantages of  electronic systems over “paper-based” systems. A good quality management system ensures that the data used in the development of pharmaceuticals is properly maintained, and easily accessible for the day to day running of the manufacturing business as well as for any internal or regulatory body audits. Visit our case study library for examples of companies using QAD EQMS (Enterprise Quality Management System) to support their quality management processes.

These are some initial steps any business can use to enhance their data integrity practices, ensuring reliable and trustworthy data.  

Key Observation and Best Practice Sharing

Interestingly, PubMed highlighted that in an observation of findings from FDA warning letters that an organization’s culture was a significant factor in the majority of cases where data integrity failed to meet expectations in the pharmaceutical industry.

Does your organization establish a culture based on ALCOA and ALCOA+ principles? Share your thoughts below and let’s start a conversation about best practices!