Automation in Pharmacovigilance Data Processing – Do You Trust AI?

Dec 13, 2017

On November 28, 2017, ArisGlobal held a webinar with DIA, “How Automation Can Deliver the Holy Grail in Safety: Achieve Productivity, Compliance and Quality”.  In this webinar, we posed a few questions to the audience, and came up with some interesting information.  We took this and spent some thinking about what this means to our clients in the PV ecosystem.

Automation technologies are having a major impact on performance and productivity across industries. In pharmacovigilance, these technologies also have the potential to reshape several activities in case processing, signal management and benefit-risk evaluation, resulting in a highly efficient pharmacovigilance system. While there’s a lot of talk about how these automation technologies will affect jobs, where it will have the most powerful impact is on the quality of work and the speed at which manual tasks can be accomplished.

As per our survey with 44 participants across 30 leading pharmaceutical companies and CROs from seven countries[1], organizations have already started adopting these automation technologies into their pharmacovigilance systems:

  • 39% say they have already adopted some degree of automation
  • 23% would like to adopt automation but are unsure how to proceed
  • 18% say automation is non-existent

[1] Canada, Germany, India, Netherlands, Spain, UK, USA (Survey based on DIA Learning webinar “Productivity, Compliance and Quality: The Holy Grail in Pharmacovigilance-by Dr. Vivek Ahuja;

Automation promises to be a game-changer for pharmacovigilance, decreasing the cost of case reporting and improving data quality to allow area experts to focus on where they truly add value, including signal detection in drug safety, pharmacovigilance analytics, and benefit-risk assessment.

Repetitive and routine manual tasks in adverse event case processing can be automated and tackled in a very sophisticated and seamless way. These tasks, for which we have coined the term ‘pharmaco-diligence‘— in other words, the due diligence of collecting and processing data — are not strategic in the same way as activities such as benefit-risk assessment. Automation can play a key role in decreasing the manual effort in data collection, management, duplicate detection, case validation and follow-up. As evident from our survey, the participants think that the surveillance activities that involve scanning through huge databases, both structured and unstructured, identification and extraction of relevant elements for pharmacovigilance can be hugely impacted by automation.

Real-time, comprehensive literature review can be conducted through, robotic automation, natural language processing (NLP), machine learning and data mining, information retrieval, classification and filtering systems. The machine learning algorithms (both supervised and unsupervised) can be applied for performing complex statistical assessments for detecting ADR reporting patterns and combining PV data from various sources. In fact, our survey shows that majority of participants would consider adopting NLP and machine learning-based tools to solve their pharmacovigilance related problems.

By building automation into the adverse event case processing and allowing pharmacovigilance resources to focus on strategic activities, life sciences companies will realize better outcomes. That’s because safety departments can work smarter and faster with reliable data at their fingertips.

That, in turn, means more time can be freed up for important activities such as benefit-risk assessment and more resources can be committed to higher human intelligence activities that are vital for the scientific interpretation of patient data.

Building Trust in Pharmacovigilance

One of the barriers to adopting AI and automation has been the ‘trust factor.’ Life sciences companies like the idea of AI, but they are concerned that automation won’t produce the same outcomes as manual case processing.

In reality, however, automation has come a very long way and is well entrenched in our lives in ways we may not realize. These systems are being used in facial recognition in mobile phones, in chatbots for customer service, for estimation of real estate prices, and for providing movie recommendations on Netflix. In medicine, they are being used in cancer diagnosis and treatment, and have been proven to be better than humans alone in doing certain tasks.

It’s time life sciences industry realizes the massive potential of automation technologies and embrace these into pharmacovigilance and other related systems for conducting day-today activities.  We encourage you to listen to the full webcast broadcast for more comparisons and insights to how automation is coming to life in pharmacovigilance today.  Click here to view the webcast.


About ArisGlobal

ArisGlobal is transforming the way today’s most successful Life Sciences companies develop breakthroughs and bring new products to market. Our end-to-end drug development technology platform, LifeSphere®, integrates our proprietary Nava® cognitive computing engine to automate all core functions of the drug development lifecycle. Designed with deep expertise and a long-term perspective that spans more than 30 years, LifeSphere® is a unified platform that boosts efficiency, ensures compliance, delivers actionable insights, and lowers total cost of ownership through multi-tenant SaaS architecture.

Headquartered in the United States, ArisGlobal has regional offices in Europe, India, Japan and China. For more updates, follow ArisGlobal on LinkedIn and Twitter.

Additional Information

Connect with ArisGlobal on LinkedIn:
Follow @Aris_Global on Twitter:


Erika Thomas
+305-726-6601 |