Artificial Intelligence – exciting tool for Pharmacovigilance professionals

The processes involved in the pharmacovigilance (PV) will be improved  by Artificial Intelligence (AI), in near future and AI will be a key player in  making PV processes faster, better and effective.

 AI governed PV processes easily meets compliance, and are much easier and effective for pharma companies and they drastically reduce the cost of running PV process. 

As an example, to showcase how AI is critical player in PV and how it practically affects all PV process, I am presenting one example below. To make things simple to understand, I have taken example of Literature review to showcase how AI helps in PV process.

AI is making literature review and its validation process very easy and automated with  reduced human  efforts.

Literature Review Process in PV:

What is Literature review?

All Pharma companies need to search, analyse and review world-wide published literature for all   there medicinal products . Followed by making decisions regarding creation of  case reports from such articles that needs to be databased as Individuals case safety reports (ICSRs). This ICSR/Case reports are literature articles that are describing safety related issues occurring in a patients taking particular type of drug products, in laymen’s term “side effect reports or articles”.

.

Why do we need to do literature review?

The medical literature is a significant source of information for the monitoring of the safety profile and maintaining the risk-benefit balance of medicinal products, particularly in relation to the detection of new safety signals or emerging safety issues. Marketing authorisation holders are therefore expected to maintain awareness of possible publications through a systematic literature review of widely used reference databases (e.g. Medline or Embase) no less frequently than twice a month preferably once a week. It is mandatory activity by health authorities worldwide for pharma companies.

Challenges in manual/current literature review process:

The literature reading and identifying the valid ICSRs is very boring and tedious work and most of the time error and non-compliance by employee is major challenge faced by pharma companies. This adds to time spent by employee, financial expenses, and efforts to run the process in complaint manner.

Quantity of task: When we do search and download the literature articles from Embase (major source of literature downloads and is widely used) for a medicinal product, search results gives us thousands of literature articles for our product. The ratios of valid safety information containing articles/ ICSR out of this 1000 articles is  merely 5-10 maximum relevant articles.

Quality of task and time spent: To identify these 5-10 relevant articles/valid ICSRs we have to read these 1000 ICSRs which is very time consuming, costly and a process that has loopholes for noncompliance. To give an idea per day an experienced PV professional read 100-150 articles and identifies valid ICSR’s. Out of this Valid ICSRs there are few ICSRs that has timeline of 15 days from its receipt’s. This adds to the effort and compliance.

How AI can help in literature review process :

What is AI?

In simple terms, AI technology works on logical process and commands given to it while developing it to provide results or do task as required.

The definition of AI from the internet is “ is the simulation of human intelligence processes by machines, especially computer systems. These processes include learning (the acquisition of information and rules for using the information), reasoning (using rules to reach approximate or definite conclusions) and self-correction”.

How AI will work in literature review?

When  we train  AI tool about Literature selection criteria for a particular medicinal product to do literature search, it runs its program with this preset combination and provides us the valid safety information articles out of 1000 non-relevant articles within one click.

AI tool also has a capability to learn, when we run 1000’s of literature article with the help of  AI tool it starts learning and becomes more  accurate and effective with time.

As AI is technology which learns and adapts, once we set it to recognize the safety information it effectively becomes faster and better than manually reading and scanning 1000 of literature articles. This will save time and efforts, which can be utilised to review and analyse only those articles that are relevant by PV professionals.

The regulatory compliance and time lines are meet even with the small team of employees working in PV once AI is implemented. AI can deliver the best quality output and to reduce the time and human efforts. It reduces the cost positively affects the price of medicines, the goal is to make medicine affordable and easily available to all patients.

Challenges of AI:

 AI has its own set of challenges, one of it is a  AI development which requires best IT technical skill as well as core expertise in  Pharmacovigilance to implement the set of instruction in AI tool.

I personally know about an exciting start-up company, which has developed AI tool, which is much powerful and can replace the traditional process in Pharmacovigilance, I am more than happy to share more details regarding same, contact me for more details about the same on my email id darshit_dra@yahoo.in.

Written by
Darshit Patel

Edited by 
Dr.Shraddha Bhange