Adam de Neergaard contemplates why the pharmaceutical industry is resistant to technological innovation and reflects on Vas Narashimhan’s interview in Forbes.
The Technology Gap
Working at home due to the global Covid-19 pandemic has given me some more time to reflect. I read the article from Forbes with Vas Narashimhan (CEO of Novartis, explaining “why it is so hard to bring tech into pharma”. It made me think about the gap between the IT buzz and real-life challenges that still exist within Life Sciences.
My Technology Journey
I have been working in the IT industry since the mid 90’s and have grown up with all the different buzz words the IT industry has invented:
- All the way from the technical specifications in the early days like floppy disk size, diskspace size, CPU power (Hz), RAM memory, redundant power supply, tape backup etc.
- to specific hardware components like mainframes, intel servers, desktop computing, mobile computers, mobile phones, different network technologies, etc.
- to the last 10 or more years moving into concepts like “High Performance Computing”, “Advanced Analytics”, “Internet of Things”, “Big Data/Data Lakes”, “Digital Disruption”, “Digital Transformation”, “Machine Learning (ML)” and “Artificial Intelligence (AI)”.
Data as Oil
I am sure I haven’t mentioned all the buzz words over the last 25 years, but the one thing that still holds is, as the Economist wrote in a leader in May 2017:
“The world’s most valuable resource is no longer oil, but data”
With new technologies emerging such as Internet of Things (IoT) supported by G5 networks we will see an even more dramatic increase in available data. Managing this quantity of data quickly and efficiently will provide new business opportunities. It will provide vital insights and potentially build new business models. We have already seen the impact of organisations who jumped quickly onto the ‘Big Data’ bandwagon selling their software and systems to deal with the new ‘oil.’
But what of data quality? Having vast amounts of data doesn’t always result in high quality data. In fact, the data seems to be getting lost in different siloed systems. It quickly gets separated from its metadata and even those definitions are somehow diminished in quality. What we end up with is a mess: poor data quality with disparate definitions, housed in different systems that can’t easily be combined. How do you get great insights from your data, if the connections are lost? How do you go back to a data legacy set, if its definitions have been stripped away in another silo?
What about a real world example from most industries. Defining and understanding your customer is vital in order to segment, focus and communicate our best prospects. As we collect data in the different siloed systems such as Excel, Customer Relationship Management systems, Enterprise Resource Planning systems, Web shops and Marketing Automation Systems etc., being able to do the most simple of segmentation becomes a real challenge: where do you start? Of course, siloed data exists across many industries. Some have adapted to this more readily and are more mature than others (ie. Banking and Retail). Pharma is facing similar data quality challenges within R&D, which leads me to the article from Forbes (with Vas Narashimhan).
The Technology Challenge in Pharma
Vas Narashimhan is the 43 year old CEO of Novartis, known for his informal dress code and focus on technology:
“We need to become a focused medicines company that’s powered by data and digital technologies”
In the article, Vas points out the ‘miserable attrition rate the industry is facing, ‘of the 20 drugs that enter clinical studies, only 1 makes it.’ Even worse, this rate has not changed in the last 15 years, while costs have exploded. Vas suggests that the increasing cost reflects the increased complexity of clinical trials. He specifically comments on the increased amount of work to collect data to satisfy regulatory bodies for submission purposes and addressing scientific and market access questions.
The article touches on the disconnect between the buzz being communicated on the Digital Transformation conferences and what is actually delivered (referred to as the “innovation bubble”). Being asked about AI (Artificial Intelligence) and ML (Machine Learning) Vas responds, ‘as we’ve gotten quite scaled and working on digital health and data science, we’ve learned there’s a lot of talk and very little in terms of actual delivery of impact.’
He continues and reflects on the importance of having clean and linked data:
“I think people underestimate how little clean data there is out there, and how hard it is to clean and link the data.”
“The Holy Grail of having unstructured machine learning go into big clinical data lakes and then suddenly finding new insights – we’ve not been able to crack, mostly because the data…to link it up..….We are spending a lot of our energy just trying to get all of our data harmonized, so that some algorithm could maybe find anything of use.”
Technology Revolution Needed!
After 25 years, what’s changed? The pharma industry still has siloed data, resulting in reduced data quality. Innovations in IT, that other industries have embraced to improve data quality, are sadly lacking in our industry.
The A3 team at S-cubed are working hard every day to solve this problem. We have combined well proven linked data/graph technology (known from Google, Twitter, Facebook etc.) with our long-term experience working with clinical data and CDISC data standards. The vision: to provide fully linked data from data collection to submission in Clinical Development. In the future, this will lead to automation, end-to-end traceability and high data quality, enabling pharma to apply AI and Machine Learning to their data.
Every day we get closer to our goal/vision working together with the industry and different industry initiatives such as CDISC360 and TransCelerate. This month we will submit our solution to the TransCelerate DDF Hackathon with focus on improving and automating the protocol design process and getting one step closer to full end-to-end linked data.
Find out more
If you want to discuss our innovative technology or how we can help you manage your CDISC standards, get in touch.