September 17, 2024
It’s not cheap to develop medicines that save lives, prevent diseases and manage health conditions: big pharmaceutical companies spend around 20 percent of their revenues on R&D. And to get new drugs to market, they need to draft thousands of documents to obtain regulatory approvals — all of which must be reviewed for quality. Manually processing these documents increases the potential for errors at the end of the drug development cycle, making regulatory information management [RIM] workflows a top target for automation using generative artificial intelligence (AI).
The best way to take advantage of AI for this is to employ a fundamentally scalable framework that encompasses data classification, security and compliance, and sets the stage to execute additional use cases over time for better return on investments.
Pharmaceutical companies can expect significant benefits when they use AI for RIM workflows to:
- Classify documents related to clinical trials in electronic Trial Master File [eTMF] systems as they are automatically imported into electronic Document Management Systems [eDMS]
- Generate first drafts of submission documents from clinical trial and other pertinent data, and
- Translate these documents into various languages.
For starters, these benefits can include 40% faster regulatory submissions; a 50% improvement in cost efficiency and a 2x reduction in quality issues.
By using AI as a tool to target and eliminate low-end, repetitive and manual tasks, life sciences companies can empower their workforce as well. Staffers can be focused on performing higher-level RIM activities, such as reviewing automatically-generated first drafts for accuracy, or take on additional work related to clinical trials.
Defining metrics for success
Pharma companies want to bring better-quality drugs to patients faster and at lower cost, and it’s important that any AI automation journey they embark on aligns to this. Workflows under consideration must be able to be scaled up to compound productivity gains and savings while minimising risks. It’s good practice, then, to define proper metrics for business impact and return on investment from the start.
Pharma companies should start their efforts with a single RIM workflow, like automating submission document drafting, before expanding to others. This way they can lay a solid metrics foundation. With proper metrics, the success or failure of an AI automation effort will be apparent within 4 to 8 weeks. At least three workflows should be defined as ripe for automation at project onset and mapped down to their discrete components. Over time each successfully implemented automated workflow will compound the productivity gains and dollar savings for the organisation.
How DXC can help
DXC has decades of experience and best practices in building validated software for the pharma industry to manage regulatory information. The cloud-native AI framework FirstPoint that we developed to help life sciences companies de-risk their RIM workflows is easily translatable for use by other industries. For instance, consumer retailers and media organisations can create localised and individualised promotional materials in multiple languages at scale, and quality check content.
We support our customers across all facets of using Gen AI to automate workflows. We have developed a workshop to baseline processes, evaluate automation potential and craft methodology to ensure that use cases are defined with proper metrics, so that companies can build their initiatives on a strong foundation.
About the author
About the author
Hiranya Bharadwaj leads DXC’s Life Sciences Regulatory Business team and is focused on transforming regulatory workflow, simplifying operations and amplifying stakeholder experience while ensuring compliance.