Email
Password
Remember meForgot password?
    Log in with Twitter

article imageQ&A: Outpacing vaccine disinformation with AI surveillance Special

By Tim Sandle     Jan 30, 2021 in Health
Automation tools and technology like artificial intelligence and natural language processing can be used mine real-world data on a real-time basis for safety and efficacy insights and maintain public trust in relation to the coronavirus vaccines.
As the vaccine becomes a reality, social media sites are having to address a deluge of mis and disinformation about safety, efficacy, and side effects. Former Facebook security chief Alex Stamos recently warned that post the 2020 U.S. presidential election, the COVID-19 vaccine would be the single biggest disinformation threat. This could have significant implications, potentially disillusioning elements of the public who are already skeptical about taking the vaccine that could help end the pandemic.
Pharma companies and drug regulators are also working to maintain real-time understanding of the vaccine’s safety and efficacy as people begin to receive it so that they can quickly disseminate that information to doctors and the general public. As examples, the U.S. FDA and CDC have set out plans to deploy near-real time surveillance of vaccine outcomes data, monitoring 10-20 safety outcomes of interest.
There are technological solutions that can help with this. To learn about this, Digital Journal spoke with Joe Rymsza, vice president of Pharmacovigilance and Regulatory Technology Solutions of IQVIA. Rymsza is a well-cited expert in FDA technology initiatives.
Digital Journal: How would you define pharmaceutical disinformation, and how can AI-powered monitoring help outpace disinformation surrounding the vaccine?
Joe Rymsza: Pharmaceutical disinformation is any published content, found online or in physical form, that presents falsified information surrounding a drug’s clinical research, intended effect, and side effects. AI-powered monitoring helps regulators and pharmaceutical researchers identify potentially false information in real-time through the simultaneous monitoring of a vast array of blogs, social media sites, and other digital forums.
DJ: Has disinformation surrounding vaccine safety increased in recent years? Why is this?
Rymsza: Why does there seem to be an increased risk of disinformation around the coronavirus vaccine? Today there are more channels than ever—from social media, to YouTube, to online forums—for ordinary consumers without any pharmaceutical or medical expertise to either intentionally or unintentionally publish and share their misinformation and disinformation around the vaccine. These channels have become a key part of how people receive news so it’s quite common for someone to consume and spread information via these channels that they perceive as factual, which hasn’t been accurately fact-checked and validated.
The COVID-19 vaccine was developed and approved by health authorities in record time relative to the approval timeline for many other vaccines. This is due to the unprecedented global collaboration between industry and health authorities to share information and execute processes in parallel vs. what can often be a highly siloed and linear fashion. Unfortunately, many people who aren’t actively involved in vaccine research and development or the execution of clinical trials leverage social media and other channels to express genuine fear and deliberate disinformation about vaccine safety and efficacy.
This is despite the fact that industry professionals and health authorities around the world who are qualified to assess the underlying clinical trial data have verified that this is not the case. Still, the public and political nature of the COVID-19 vaccine effort has made many skeptical.
DJ: Is social media the main source of disinformation, or does the established media have a role too?
Rymsza: There are many sources of disinformation, but social media platforms are the most accessible source for laypeople. Established media outlets are more likely to vet their sources and require people to go on record when they make claims such as those relating to the safety of a vaccine. However, with the rise of consumer-run news sites over the past several years, it is getting harder and harder for the general public to determine which news sources are reliable. Additionally, it’s not uncommon for consumers today to skim headlines rather than review full articles, and people don’t always do their own research. The real-time influx of information available about the vaccine could lead people to hold feelings of doubt toward that vaccine that simply are not accurate.
DJ: How can automation tools and technology like artificial intelligence be used to mine real-world data for pharma safety information?
Rymsza:What role can natural language processing play in this process? The pharmaceutical industry must have its finger on the pulse of exactly how people are reacting to vaccines by monitoring how safe and effective they are in real-time as they are distributed to the public. This unprecedented imperative to collect and share accurate information and dispel any misinformation is complicated by the sheer amount of data that will be collected on the millions of people planning to take the vaccines. AI-powered solutions and automation can help pharmacovigilance teams monitor for pharmaceutical safety information in a much more timely and accurate way than manual monitoring and data analysis.
AI-powered automation enables pharmaceutical companies to gather insights from diverse data sources, including digital channels like social media platforms and forums, electronic health records, medical call centers and patient registries. AI technology can help teams sift through data more quickly and efficiently to look for patterns of anomalies that indicate an adverse event, also known as an unfavorable or unintended side effect of a medication, to flag and examine in real-time.
Natural language processing (NLP) is a recent advancement that expedites the codification of information submitted by consumers. For example, it enables the grouping of “stomachache” with “belly hurts” even if those words are in entirely different languages. This expedites the data sourcing process dramatically, helping teams to make sense of unstructured formats, in which 80 percent of health data resides.
DJ: How can these types of tools assist governments and health agencies in monitoring safety and efficacy and maintaining public trust in relation to the vaccine?
Rymsza:Encouraging public trust in vaccines is a critical step towards society achieving herd immunity and ultimately ending the pandemic that has taken hundreds of thousands of lives. Consumers are much more likely to trust the safety of the vaccines if they understand the way in which governments and regulators are transparently and efficiently identifying and responding to vaccine disinformation with reliable, up-to-date data.
Pharmacovigilance with AI, ML, and NLP will help pharmaceutical companies and regulators keep pace with the post-market patient safety monitoring needed for a global vaccine to understand the risks and benefits in real-time. This feedback loop of accurate inbound information can be instrumental in eliciting enough trust in the vaccines from consumers to ideally bring about an end to this horrific pandemic.
More about Artificial intelligence, Vaccine, coronavirus, Safety
More news from
Latest News
Top News