How Artificial Intelligence will Transform Medical Affairs and Commercial Teams

minutes to read

New Applications of AI in Life Science Marketing and Communications

Commercial and medical affairs teams in life sciences are undergoing a massive industry transformation. The magnitude of content being produced has increased substantially over the past few years thanks to omnichannel campaigns, amongst other reasons. This has put more pressure on teams to produce much more quality content while maintaining compliance. This explosion of content is causing solutions of the past to grind to a halt. Hiring more people or great agency partners is no longer enough. Medical Legal Regulatory (MLR) promotional material reviews must be able to handle peak times when the volume of content is greater. 

There are two causes to these problems: 

  1. Frustrating manual processes are choking productivity and can be solved with technology. 
  2. The second is what I call the “human context” problem. It’s not that humans make mistakes as much as they don’t know what they don’t know. Newly hired employees or MLR reviewers can’t possibly have the context of everything that happened before it.  

Artificial Intelligence is the way that repetitive tasks can be learned and executed in a way that gives context to team members. Here are 6 ways medical affairs and life science promotions will be transformed with artificial intelligence. 


References

Creating promotional content in life sciences requires significant checks and balances. Any claim about a drug or medical device must be thoroughly vetted for accuracy. Failing to provide evidence for claims can result in fines from regulators or potential harm to patients. The process includes highlighting text in a document and linking it to a reference text highlight in another document. There are dozens, and at times hundreds, of these referenced medical claims in a single document. AI and machine learning can automatically detect medical claims based on past projects and ask for a human to verify that they are correct. This results in lower risk that something will be missed and reduces the repetitive, manual process of referencing other documents. 

Consistent Messaging

Writing about product benefits with fair balance in a consistent way across different pieces is critical to effective communications. Artificial intelligence can read, understand and flag when something is similar to another phrase or sentence in another piece of content. It helps distributed teams keep track of how to say things. It also helps in the review process as well. If a claim or section of content was previously approved in another piece, it avoids arbitrating things that already have consensus.  

Remove Language Barriers

Localizing content from other jurisdictions means that the authors or reviewers may not speak the language. Automatic translation services like Google Translate have been around for at least a decade, but just recently they are actually becoming reliable. You need to be able to trust the nuances in the writing will read the same when translated. Artificial Intelligence and specifically natural language processing (NLP) can understand things without getting lost in translation. This allows a reviewer to approve content, or a medical writer to author content who can’t communicate in the published language. This reduces coordination and complexity to creating localized content. 

Classifying documents

There is a famous story about how AI was used at a cucumber farm to classify the gourds into 9 different categories based on shape, length, colour and the number of prickles. The matriarch of the family farm spent 8 hours a day manually sorting cucumbers during peak harvest season. This can be done with documents as well. Artificial intelligence can read the document and determine what the drug or product is, whether the audience is public or HCP, and just about anything else you can imagine. This means you’ll never go looking for a document that was put in the wrong group by accident, and remove the manual process of filling in properties about a document. 

Extract Structured Data

Documents are commonly viewed as “unstructured data”. Think of Microsoft Word vs Excel. In a text document, you can’t sort or understand associations between data points, but in Excel, data is organized into columns and rows that have meaning. In a recent conversation with a client I asked them if it would be helpful to search for medical journals by author, year, journal etc. She said “Yes of course - but I don’t want to type all this stuff in.” Artificial Intelligence can extract information from documents that would otherwise be unreadable by computers. Important dates, names, companies and tabular data can be used to augment search or combined across documents to tell a bigger picture. 

Regulatory Authority Correspondence

A key source of documents being securely stored is correspondence with health authorities such as the FDA. In those letters will be feedback and commitments that need to be tracked. This might include assigning people to address specific things, and ensuring that they are done by a specific date. Currently, this is done manually, and most likely stored in an Excel spreadsheet, or worse, manually entered into some type of system. Artificial intelligence pulls that correspondence from plain text documents and structures the information in an actionable way. 

Conclusion

This shift isn’t just a prediction, it’s happening right now. At Papercurve we are working on solving these problems – so stay tuned. With so much content being produced, it’s more important than ever to have systems to reduce risk in real-time without relying on manual processes. In addition to reducing risk, we work every day to unlock employee efficiency and give time back to do your most valuable work. 

By Ryan Whitham, CEO @ Papercurve

If you have any questions or would like a demonstration, book some time with us.