The Future of Artificial Intelligence (AI) in Pharmacovigilance (PV): Six Key Principles from the World Health Organization (WHO) on the Use of AI In Healthcare
Tuesday, August 24, 2021
These days, more and more companies employ automation technology in pharmacovigilance processes. The resource-intensive industry of pharmacovigilance and clinical safety is constantly looking to streamline processes with the help of software providers to reduce the costs of manual processes and increase program efficiency.
This very topic has attracted the attention of the WHO, who recognize the potential for streamlined pharmacovigilance case processing workloads, to have a computer extract relevant information from large databases to produce aggregate reports, or to have a machine decrypt handwritten reports. Artificial intelligence can indeed help to improve quality and efficiency but will remain a complement to human input.
We can compare it to a modern car which has a lot of useful automated features like cruise control, spacer, self-adjusting traveling lights, etc. Practical and important features for the safety of the passengers indeed.
However, occasionally you will still find yourself struggling against your own car. Sometimes it breaks too early when the other car is still miles away, sometimes it overreacts in harmless situations, and sometimes the lights do not adjust fast enough.
This automation technology is helpful but human correction is still needed. This is the same for automation with AI in pharmacovigilance. It still requires human experts to provide their assessment and final opinion.
This is confirmed by the six key principles in the new guidance from the WHO on the use of AI in the health industry: The Ethics & Governance of Artificial Intelligence for Health report.
- Humans should remain in control of health care systems and medical decisions.
- AI products should be required to meet standards for safety, accuracy, and efficacy within well-defined use cases.
- AI developers should be transparent about how products are designed and function before they are used.
- Health care businesses that rely on AI should ensure they are used under appropriate conditions by trained personnel.
- AI must be designed to encourage inclusiveness and equality.
- The performance of AI applications should be continuously and transparently assessed during actual use.
A comprehensive PV strategy requires a depth of understanding not only in terms of the complex and rapidly evolving safety landscape but also in new tools and approaches that may significantly streamline and simplify workflows, increase the capacity of your team, lower your cost to deliver safety services, and, most importantly, help you differentiate and maximize the value of your products in the marketplace.
Pharmaceutical and biotechnology companies benefit from working with external partners to manage PV activities. The breadth and scope of required services can depend on the mix of investigational and marketed products, the size of the company, and their comfort with outsourcing safety. Organizations realize a variety of benefits when collaborating with an experienced and innovative technology partner for safety services.
Are you PV ready? Evaluate your program using our complimentary Safety Readiness Checklist here and ensure the proper oversight is in place for all of your safety reporting.
About The Author
Dr. Alexandra Thoenes, MD, Senior Project Safety Physician, is experienced in general and traditional medicine, neurology, and surgery in Berlin, Paris, and Shanghai. Dr. Thoenes brings four and a half years of experience in the pharmaceutical industry as a Medical Advisor in the Neuroscience and Diabetes Department as well as seven years of experience in Pharmacovigilance (Benefit-Risk Assessments, Signal detection, SOVs (safety observations), SSARs (safety signal assessment reports), PSURs/PBRERs, HA requests, Clinical Overviews and Medical Monitoring.