OR WAIT 15 SECS
Laurent Fanichet outlines four ways pharma companies can use cognitive search and search-based applications to improve competitiveness.
Pharmaceutical companies accumulate enormous amounts of data over many years of working on drugs, from early research, via lab projects through years of clinical trials, approval processes, and ensuing pharmacovigilance. They are permanently confronted with the question of how to best leverage that data in different situations, whether presenting new drugs for FDA certification, putting together teams of experts for new projects or conducting clinical trials based sometimes on years of intense research. To exploit these data and remain viable in their competitive market, IT teams from pharmaceutical companies’ R&D, drug safety and regulatory departments can rely on cognitive search and analytics platforms and search-based applications (SBAs) that use natural language processing (NLP) and machine learning (ML) to deliver increasingly more relevant information to everyone.
Below are just four ways pharmaceutical companies can use cognitive search and analytics platforms and SBAs (based on these platforms) to improve their competitiveness.
1. Discover Networks of Experts
For a pharmaceutical company with R&D laboratories across the world, it is not always easy to select the ideal team of experts for a given project from the thousands of in-house researchers. Many companies turn to enterprise social networks for their employees to declare their competencies and focus of work, thus fostering cooperation and enhancing synergies. But these social networks usually are not kept up-to-date after a few initial months of enthusiasm. They simply lack the immediate responses of general social media, and people stop posting news of their work after a while.
Making things more difficult, in the pharma industry these experts necessarily come from different disciplines: pharmacology, medicine, biology, genetics and possibly others. So what’s the best way to discover experts on a given subject? The answer is: look at their written “footprints.”
Cognitive search and analytics solutions analyze scientific articles, project and clinical trial reports and other documents, as well as data from internal and public databases, to select the best profiles. They can sift through more than 100 million documents and billions of database records to accomplish this. Once the groundwork is done, they answer queries in less than two seconds.
In order to optimize the relevance of proposed networks of experts, these solutions can be tuned, for example, by requiring that a person be a “leading author” of a publication or report, or that names of people be mentioned in the same document or in the same sentence as the topic for which experts are required. Such a topic could be a drug to be repositioned, taking into account its many synonyms.
By expanding the field of research to experts from other institutions, pharma companies can also find potential cooperation partners in various stages of a project (R&D, lab tests, clinical trials, etc.).
2. Create R&D News Alerts
Rather than actively looking for information in a particular field or on a specific topic, people can subscribe to relevant news for regular alerts so they can keep pace with advances in their fields of study. SBAs that include NLP and ML algorithms such as clustering, classification by example and similarity calculations can perform the monitoring necessary to ensure the latest information is delivered in a timely fashion.
These SBAs cover all accessible external data sources, e.g. publications, Embase, Medline, Scopus, clinical trial reports, etc., as well as internal sources such as SharePoint and Documentum. They allow people to see, at a glance, spikes of activity in their field of expertise that potentially indicate that competitors are working on the same subjects as they are – conveying a strong degree of urgency in an industry where “the winner takes it all,” i.e. the first on the market makes most of the profits from a new drug.
3. Optimize Clinical Trials
Clinical trials going over many years generate vast amounts of data per drug and study. Over time, biostatisticians face tremendous challenges performing their analyses with the right datasets; for example, a comprehensive list of patients having certain diseases and having developed certain side effects during trials of a given drug.
Cognitive search and analytics platforms and SBAs built on top of them let people perform targeted searches on content even when merging data silos from different sources. They can use extremely precise criteria by specifying exact values or value ranges for the right variables. Pharmaceutical researchers can thus transform their growing clinical trial data into a valuable asset.
4. Simplify Compliance
Following publications of regulatory authorities around the world and making sure submissions to these authorities fulfill their requirements is no easy task. The same is true for listening to the “voice of patients” on social media. Teams of experts often manage these tasks. Unfortunately “humans don’t scale”, i.e. you can’t hire scores of people around the world to keep up with the evolution of regulations and patient feedback.
A cognitive search and analytics platform helps biopharma companies detect relevant information in regulatory texts and in “patient voice,” and pushes it to the right experts within the company for further treatment. “Classification by example” algorithms can use the corpus of hand-selected data that has been attributed to the right people by human experts over the years.
Cognitive search and analytics platforms allow pharmaceutical companies to extract valuable data from a multitude of complex and diverse sources to obtain contextualized and relevant information. Modern SBAs offer pharma research possibilities unthinkable just a few years ago, accelerating time-to-market and helping pharmaceutical companies remain competitive.
Laurent Fanichet is VP of Marketing at Sinequa.