We all keep hearing about Artificial Intelligence everywhere, and it's no surprise that it has also consumed the Pharmacovigilance industry. While I am no AI expert, I tried to talk about a few things that, as a safety physician found fascinating, we have achieved with the use of AI, and a few things that I am eagerly awaiting to be implemented in PV.
The technological revolution that helped us achieve tremendous transformation, some examples below, not just in PV but in the healthcare industry, as noted below.
1. Identification of disease has really become faster, cost-efficient, and more cost-effective.
2. Predicting the disease progression and at-risk population has also become easier with the help of technology, which allows patients to know the early signs, symptoms, and Healthcare providers to have tech-assisted platforms to then also utilize faster diagnosis.
3. We are using diagnostic algorithms, telemedicine, virtual patient monitoring, and robotic surgeries.
4. AI-powered tools assist in drug discovery, automation in manufacturing and distribution, virtual and decentralized trials, and large data volumes are managed by machine learning.
5. Data collection has become easier due to wearable devices.
In the pharmacovigilance world, we have also made some good strides with respect to embracing automation, and I have noted some well-known examples below.
1. ICSR: Automation has been implemented for duplicate check during case intake and triage, making it less time-consuming and cost-efficient not to manually check during intake if it's an initial case or follow-up.
2. Literature monitoring: An AI-powered literature database that allows for excluding and including articles based on the four minimum criteria and the parameters added at the product level (brand name, ingredient, MAH territory).
3. Signal detection: Many signal tools are available that allow to use of statistical models for signals and unified signal outcomes from various sources.
4. RMP and Aggregate report management:
Various tools and databases are available to track, schedule, and submit aggregate reports.
RMP tracking and RMP commitments are also tracked via tools that allow for ensuring compliance.
There are multiple advantages of technology integration, and those we are already aware of; however, to ensure the summary below includes some most critical advantages.
1. Cost efficiency - Intelligent and smart cost utilization.
2. Speed and real-time analysis - continuous monitoring of data from diverse sources enables quick and timely identification of new signals and hence provides risk management.
3. Improvised and consistent quality: Automation and AI integration have led us to move from volume-based to value-based.
4. Increased efficiency - Automate labor-intensive tasks, allowing faster processing and analysis of large volumes of data.
5. Scalability- Automation across global databases and platforms has led to facilitating signal detection for large databases. It has also allowed to use of AI/Automation to generate scientific outputs.
However, there are a lot of opportunities still existing in pharmacovigilance for not only embracing technology but integrating it for pioneering patient safety, and I have noted a few that I could pick from various sources.
1. Structured and regulated processes
2. Time-driven assessments to be performed regarding the implementation of technology.
3. Patients' outcomes are monitored closely after the implementation of technology.
4. Data privacy and confidentiality are unified to have simpler and effective implementation
5. Use of robust database, human scientific intelligence for validation, security, transparency, and validity. Dataset standardization with collaboration and communication.
6. Implementation of technology has to be in parallel to training human resources on technology, collaboration
7. Implement technologies that use cloud-based solutions, open-source platforms to improve reach, knowledge sharing, and collaboration.
Some of the examples about integrating technology are AI-driven case processing, allowing only relevant cases to come into the medical workflow, and Natural language processing (NLP) for literature mining, AI-driven signal detection for classification of signal, and addressing the impact of signal.
Many avenues exist for integration of AI in technology, and if we are open-minded to accept them, we will be able to achieve the patient safety outcome we all work for i.e. safe and effective drugs in the market.
Written by:
Written by:
Dr.Shraddha Bhange.
Connect with me Via comments below. (I do not respond to Facebook messages)
Support the cause of better rural education with me: ThinkSharp Foundation http://thinksharpfoundation.org/#home
VigiAccess: Promoting public access to VigiBase - PubMed
(PDF) Artificial Intelligence in Pharmacovigilance: a regulatory perspective on explainability
No comments:
Post a Comment
Please share your suggestions, they are most welcome!