Drug safety risk analysis – solving the puzzle of unstructured data

Traditional information gathering throughout the pharmaceutical industry, directly from healthcare professionals, patient registries and regulatory databases, provides insightful information for drug safety risk analysis.

However, relying solely on a limited source of structured data only accounts for part of pharmacovigilance (PV). In recent years, as the digital landscape has continued to expand, patients have increasingly used non-traditional reporting channels to express concerns and seek advice on drug safety events.

As a result, there is a wealth of untapped information and direct patient feedback from patient assistance programs (PSPs), social media platforms, online forums, discussion groups, and more. To access and leverage this data, clinical research stakeholders are increasingly turning to automation and artificial intelligence (AI) to extract and interpret information from these consumer channels. To understand how these technologies are revolutionizing data collection and analysis, it is important to consider the current drug safety landscape and the use of traditional data collection methods.

The traditional landscape of AE detection
In the past, adverse events (AEs) were reported directly by patients and healthcare professionals. Traditional AE reporting relied on manual and fragmented systems and protocols, resulting in an incredibly time-consuming process. This reporting relied on structured data collection networks in a pre-approval setting or by healthcare professionals in a post-approval setting, creating a gap in information sharing and potentially hindering the early identification of safety concerns.

To make matters worse, not only do the methods for detecting and reporting AE make it difficult to structure accurate results, but the data also shows that experts still lack a good understanding of the current AE landscape.

You can read the entire article here.