The one Pharma data revolution you may not hear about
The Pharma sector is bristling with initiatives and discussions about how data analytics and AI are being explored to help it succeed in the coming decades.
A recent PharmaVoice article outlined how digital technologies were impacting every part of our sector's range of activities. Not surprisingly, many of these were enhancing drug discovery. For example - using automation and data to map most known drugs against every Covid-19 disease mechanism; decoding protein folding, neural networks to analyse millions of known molecules.
Different digital technologies are impacting diagnostics. Pattern recognition in medical imaging; facial, speech and even gait analysis to diagnose neurological conditions.
Translational work is an obvious candidate for AI and analytics. Stratifying patients by finding mechanism correlations with genomic analyses. predicting dynamic and kinetic properties of thousands of lead molecules.
Clinical work will also benefit. Voice technologies to help trial patients from home; monitoring adherence; analysis of many past trials to better design future ones.
Commercial / marketing is embracing analytics, in creating a clearer picture of customers; assessing the success of various mixes of sales channels, across the varied regions of the world.
Since 2013, our sector has invested, in addition to its traditional innovation work, some $8 - 9 billion on digital solutions. Will these bring new advances and transform the declining productivity of our sector? Perhaps yes. But there may be something missing. In the article mentioned at the beginning, as well as in other publications on this topic, there's not a single mention of what could also be a natural opportunity for analytics and AI enhancement - project management.
Cross-functional working, across many assets in simultaneous development, requires serious project management. In recent decades, system-driven project and portfolio management in our sector has indeed developed into a thorough discipline, mirroring developments in other sectors such as construction and energy. 'Agile' project working, pioneered by the IT industry, is being tried and tested actively in Pharma.
However, some industry observers maintain that the success rate of projects across industries remains dismal. In Pharma, they point to many systemic inefficiencies, additional to the oft-cited risk and attrition rate of our drug projects. Also, that despite Agile developments, our project management practices are in dire need of a different kind of innovation: using artificial intelligence to guide our planning, decision-making and team working, based on analysis of data from thousands of historical projects.
There are some reasons why this proposition is challenging:
The Pharma R&D projects world generally favours optimism about new projects, at the expense of spending a lot of time capturing (and acting on) lessons learned from old ones!
If they did capture project data, there is generally not enough of it generated from one organisation for AI to generate meaningful conclusions. In addition, Pharma companies are reticent to share lots of data with others, even if it might be non-competitive information.
There is currently a lack of published discussion and experience about using analytics and AI in project management, as opposed to the many growing applications mentioned at the start of this article.
The Pharmaceutical Industry Project Management Group (PIPMG), which represents project managers in our sector, is looking to influence this situation. To start spreading some more knowledge it's mounting a free event on 18th November, featuring two expert speakers. Going forward, the group is looking at further opportunities. Firstly, hosting training programmes for practitioners in the industry, to develop skills in building datasets and AI agents.
Secondly, exploring the possibilities of establishing a UK Pharma Data trust. A Data Trust is generally a non-profit body that seeks and catalogs project-related information across an industry sector, making it available to sector organisations.
There is a long way to go on this journey and as yet, the potential benefits are as yet unrecognised by most all organisations in our sector. But watch this space!
Reference
'Driving Drug Innovation With AI' - PharmaVoice Magazine October 2020