AI use cases in biopharmaceutical medicine

I love winter, the days are short, yes, but walks out in the forest are wonderful and the cold makes being indoors even more enjoyable. Incredibly, however, I just realised, there are only four weeks to the end of the year.

Today’s topics:

We all agree AI will change/has changed the world. The reason this is happening now is the convergence of the following: computing power, networks, knowledge, statistics, and data availability.

The use cases are endless, and the bright new world of the future, some of which is already here, spans vignettes including accelerated clinical trials, earlier disease detection, a lower burden of chronic disease, AI-assisted operations, coaching and psychological counselling as well as more relaxed doctors who, supported by AI, can spend more time with patients. I am confident that this last one, although often discussed, will sadly not become a reality.

At last week’s “AI in biopharmaceutical medicine” event hosted by PWC, three presenters shared their knowledge: Dr Joanna Soroka, Principal at investor Hitachi Ventures, who spoke about investing in the AI/healthcare space and shared some insights on the criteria Hitachi applies when selecting companies to invest in, Sotirios Perdikeas, Predictive Modeling and Data Analytics Leader at Roche, who spoke about the impact that AI has in his area at Roche, as well as key success factors for implementing AI and Dr Andreé Bates Founder and CEO of Eularis, a consulting agency that specialises in the strategic and adapted use of AI/FutureTech to improve business outcomes, who covered an incredible amount of topics in a very short amount of time.

Many clinical trial topics were covered including the use of AI in drug discovery and clinical trial management. Dr Andreé Bates mentioned many use cases including a clinical trial where Phases I and II were run exclusively using digital twins, i.e. with no human or animal subjects. Sotirios Perdikeas shared an example where AI was used to assess the impact of standard inclusion and exclusion criteria on the hazard ratio in clinical trials. The outcome of this assessment was that 70% of the assessed inclusion and exclusion criteria had zero impact on the hazard ratio. He also mentioned that AI could be used to optimise clinical trial assessment schedules, revolutionising an area where typically the optimal number of visits and examinations is experience-based.

Data science topics were also extensively covered at the event. Dr. Joanna Soroka delved into considerations like prioritizing platforms when considering investments. She also highlighted the shifting landscape, where healthcare companies, who have started to leverage technology in their business, are now being joined by technology firms leveraging their expertise to enter the healthcare arena.

While the speakers covered many topics, three key success factors for the use of AI in the biopharmaceutical industry emerged consistently: the tailored use of the technology, the availability of high-quality data, and the importance of involving the right people from start to finish.

Often, we concentrate on AI’s potential to provide solutions without recognizing the essential role of humans in project success from design, to implementation, to post launch management.

Sotirios Perdikeas spoke to this often-overlooked point: “the launch of any new technology is an issue, deployment is an issue, we often think deployment is the end of the journey, but it is, in fact, the beginning”.

If you need help to support a deployment or someone to accompany your project from start to finish to help ensure your organisation and your people are well placed to succeed and make the most of a new solution or approach, I would be happy to explore with you how I can help.

Key takeaway: AI based solutions are changing the healthcare space at an incredible rate, but you can do nothing without the right approach, data and people.

Your rabbit changed my life – a personal data management case study

I often take notes when I am out and about so that I can use them for later newsletters on topics that are relevant to the pharmaceutical industry and for anyone interested in optimising their operations. One such note read as follows “crystal ball/factory farming/athletes/farming versus industry/mini break/medicating patients/startups/princess does not do punctures/your rabbit changed my life”.

Unfortunately, while I have vague memories of the content these snippets were anchoring, the snippets are insufficient for newsletter use in any other way than to illustrate the importance of making sure the data you capture in your systems is fit for purpose.

In a recent discussion on AI the instructor said, “you should expect to spend around 50% of your time on cleaning the data you will work with, another 25% of it analysing it and 25% on visualisation”.

In my example above, the only thing the data tells me is that my mind sometimes works in mysterious ways and covers many seemingly random topics.

Consequently, I have had to use the data to tell a different story to the one I had initially intended.  So, in this case it’s repurposed data. However, in many cases the data is not as obviously garbled as in my example above, and you might not notice. If you then use it to analyse your market you may have an issue.  

Therefore, when you are designing systems for data capture make sure you design in such a way that individual interpretations of the data that is being captured, and consequently variations in data capture, have minimal impact on the data quality. Aim for a harmonised understanding of data structure and categorisation across your business.  Ensure your teams know why this is important and how good data capture will benefit them and their projects. 

Key takeaway: You reap the data you sow, so, sow wisely.

Leveraging diversity – learning from the cosmetics industry

Douglas, a leading European cosmetics company, markets products both online and in-store. The company operates under different names such as Douglas, Nocibe, or Parfumdreams in nearly all European markets.

During a recent visit to my local store, the manager told me, “Regardless of someone’s place in society or their age, they are welcome to work here.”‘ He continued, “My team is incredibly diverse, with each member bringing unique experiences that enrich our group. For instance, the lady working at the cash register is past retirement age, but she’s a seasoned veteran in the industry. This is fantastic for us. We learn from her, and that helps us serve our customers better“.

The cosmetics industry targets everyone in the market, and, in the store I visited, everyone is serving the market. I’m not sure if this is the official ethos of Douglas or just the perspective of the manager I met, but I think it is a smart approach.

Why is this relevant to the pharmaceutical industry? It matters because, despite the industry’s official commitment to diversity, and the admittedly great changes that have been made, I often encounter individuals aged 50 and older who worry that they won’t find another job if they are affected by a reorganization. A friend with an impressive CV recently told me that she was recently told the hiring manager was “looking for someone younger”.

While the pharmaceutical sector is different from cosmetics retail, the argument for hiring for diversity, and this includes older, experienced individuals, is even more relevant. In an industry where individuals change roles often, navigate complex international and regulatory environments, and are highly educated, hiring for diversity, with a focus on knowledge transfer, will contribute to better project outcomes and enhanced effectiveness.

Key takeaways:  1) Fostering diversity, valuing experience, and promoting continuous learning are essential for improving project outcomes and overall effectiveness. 2) Every discussion, every interaction, and every industry has something valuable to teach us

Leadership – balancing the majority to drive innovation

Your most important asset is the people you work with — the individuals who hold much of the knowledge in your organization and whose creativity has to power to drive your business.  In order for a company to thrive, ideally, voices that raise creative suggestions should be heard. However, this is not easy to ensure, because often the status quo, or the majority view, triumphs over voices that bring in new suggestions.

While I am writing about people, I am using an AI example here to illustrate how a single voice can easily be overheard:

Your AI system is only as good as the data it is trained on and this data, is by necessity, historical data.

Using the example of the earth being flat or round: If you trained your AI on data from 500 BC, the consensus would generally have been that the earth is flat. Your AI tool would answer the question: “is the earth flat?”  with a clear “yes”.

If instead you train your AI on data from 200 BC and asked the same question. The answer would be “No, while up until recently people believed the earth to be flat, scientists have now shown that the earth is round”.

If you had asked your AI to respond to this question at any time point between 200 BC and 500 BC, the answer you received would have varied depending on how you trained your tool and what data you included.  

This illustrates the point that even when the data is overwhelming and everyone agrees on something, that doesn’t make it correct. 

Leaders in organisations, managing systems, are faced with this challenge all the time. We learn from the past, there is a lot of knowledge held both in systems and tacitly, and it is easy to trust the weight of the knowledge that exists and to take comfort in numbers and the majority vote.However, the issue is that ground-breaking new ideas typically originate from very few individuals within an organization.

As a leader you need to ensure that the accepted status quo doesn’t suffocate innovation. It’s important to ensure that unique, original and creative ideas are heard, and considered, regardless of the position of an individual in an organisation’s hierarchy. 

If you want to explore how to help your teams embrace innovation without fear, I would love to discuss.

Key takeaway: New ideas are often met with distrust. It is important have a process in place to  ensure they are not dismissed out of hand.

Thank you for reading. The end of the year is nigh, if you are looking to solve an issue before 31st December or to prepare for next year I’d love to discuss how I can help you either with your business strategy and operations or with your team and personal development goals.

I look forward to hearing from you,

Isabelle C. Widmer MD

Image credit: Natalija Smirnova