There’s no denying that the future of healthcare is digital. It’s almost impossible to argue its past and present are otherwise. Patients have long been scouring the internet in search of material on conditions and treatments, and 2020 saw employers and employees turning en masse to cloud-based technologies specifically to get through a public health crisis, to take just two examples. The cliché is to call this change a ‘journey’, but it’s also a race. Research and development departments that don’t hurry up and adapt will soon be choking in the clouds of dust left by their competition.

And there’s no shortage of dust to kick up. Developing a new medicine is a long and complex process. Measuring from initial inception through testing in clinical trials to final approval, it can take upwards of a decade for one to reach the market, with weeks at a time taken up with tedious manual tasks. As the costs of navigating that path continue to rise, price pressures on the final products increase and failure rates remain high, it’s no wonder that the prospect of ‘transformation’ is so appealing.

But clinical trials have never not been about the data. In the right light, the dust that accrues as drugs are developed is golden. As Novartis vice-president Luca Finelli and CEO Vas Narasimhan write in their report on the company’s digital strategy, ‘Leading a Digital Transformation in the Pharmaceutical Industry: Reimagining the Way We Work in Global Drug Development’ in Clinical Pharmacology and Therapeutics, “All along the value chain, from research to market, we wrap our molecules around with data: to prove their mode of action, their safety and efficacy, and eventually to inform healthcare practitioners on how to best use them to improve or extend lives.” In more senses than one, transformation is about knowing oneself.

The difficulty, as Finelli and Narasimhan make clear, is that, unlike communicative, explanatory molecule data, operational data – siloed across diffuse systems, functions and formats – is not typically generated for any greater purpose than supporting temporary tasks and transactions. As a result, even providing an update on one study’s status usually requires intensive examinations of national, regional and global spreadsheets, and countless one-time extractions of data points recorded and reorganised according to different assumptions and rules. At the same time, the pair write that operational data “carries a quantitative and precise representation of our ‘experience’ with many of the tasks supporting drug development”. They give the example of historical trial management data, which can indicate exactly how successful a particular medical centre was at recruiting patients and how well clinical drug supply followed the initial plans. “For an organisation like [Novartis] Global Drug Development, the wealth of operational data collected over the past 20 years is a true gold mine that can be explored with the tools offered by data science to build predictive models for how future activities will go.”

In Finelli and Narasimhan’s experience, the laborious manual work of preparing and reprocessing that data to make it suitable for analytics is the foundation of a digital transformation. At Novartis, the result has been a new analytics platform that allows development staff to “access ‘experience’, create ‘intelligence’, and unlock ‘value’”. To start, Nerve Live, as it’s called, automatically “ingests” data from different sources, cleans it, links it, and prepares it for analytics. From there, it builds sophisticated machine-learning algorithms to analyse the data and generate actionable insights, which are presented to employees through dedicated web-based front-end user modules – ‘solutions’ or ‘applications’ – to help them make better decisions.

AstraZeneca – another early leader in the race to digitise – has also spun decades of clinical trial operations experience into the data and analytics hubs ‘Merlin’ and ‘Control Tower’. The former is a study design companion that, according to digital health heads Cristina Duran, Matt Bonam and Alicyn Campbell, can create a study cost estimate in under five minutes and select sites for trials 70% faster than before, as well as helping to ensure trial populations are more representative of real-world patient groups. The more grittily titled Control Tower is a clinical trials visualisation tool that provides real-time information on the status of studies on a global, national and regional level, and is also capable of predicting recruitment. On top of that, Duran, Bonam and Campbell point to a range of digital clinical supply chain enhancements that the company estimates could cut costs by $100m and reduce waste associated with trials by 40%. In fact, they have already helped AstraZeneca navigate and fast-track research through Covid-19 by monitoring global stocks of medicines and their movements through the supply chain in real time.

Keep on track

The variety of names AstraZeneca has dreamt up for its various apps and insight hubs obscures the fact that these tools exist to unify previously fragmented operational functions. By giving clinical supply teams instant insights into patient recruitment and vice-versa, properly integrated digital platforms make it far easier to keep everything aligned and on track.

Moreover, by combining these tools with advanced manufacturing and monitoring technologies, companies can begin to multiply their productivity levels. To enhance the robustness of its manufacturing and logistics, Janssen has invested in state-of-the-art data science tools and adopted a novel labelling technology to increase its agility and make itself more responsive to customer and patient demands.

“For an organisation like [Novartis] Global Drug Development, the wealth of operational data collected over the past 20 years is a true gold mine that can be explored with the tools offered by data science.”

Luca Finelli and Vas Narasimhan, Novartis

“One of the challenges of clinical supply chains is the variability in enrolment,” says Luca Russo, the global head of the clinical supply chain for Janssen Research & Development. “Beyond the investment in better forecasting and planning technology, we invested in just in time [JIT] labelling technology. In traditional clinical supply chains, booklets covering language for all participating countries are packed with each clinical kit at the time of packaging. In the new JIT process, clinical kits are labelled with a two-dimensional barcode and only at the time of order the system prints a country specific label. This makes our supply chain significantly more responsive and agile.”

That’s just the beginning. As part of its investment in process analytical technology, Janssen is also deploying sensors and systems to support the development of predictive control models, says Russo. “Among others, this technology is implemented as an integral part of our continuous manufacturing line. We are also investing in electronic batch-record technology. Our ultimate intent is to conduct real-time release of our products.”

The individual technologies that comprise these enhancements are hardly new. Sensors are already widely used within the supply chain to monitor the conditions products are subjected to on their physical journey from one place to another. “Sensors can offer better tracking quality and better control,” explains Lisheng Gao, an analyst at Lux Research, and author of the study ‘Sensing for Modern Logistics’. “Many things can happen when you ship samples from point A to point B. Sensors can monitor the samples’ temperatures, locations and events during the journey. By tracking these critical parameters, researchers would immediately know their valuable samples’ status and [could] take timely actions to prevent unpredictable events. Whenever results go wrong, data collected by sensors also offer additional information to help researchers to identify the root causes.”

By routing those sensor outputs into digital platforms like Nerve Live – specifically the ‘SENSE Insight Centre’, a mission-control-like room with a trial dashboard that offers real-time visibility across Novartis’s network of manufacturing and planning teams, as well as the ‘Resource cockpit’ and ‘Nucleus’ modules – companies can effectively automate the process of communicating and responding to them on a local, regional and global scale. Novartis reports that the SENSE Insight Centre provides predictive oversight of its more than 500 clinical studies worldwide, using machine learning and real-time information to identify potential risks affecting the process of each trial, and support action to prevent problems before they arise.

“Real-time information offers stakeholders the capability to become agile and more capable of making timely decisions, reducing potential loss and improving efficiency,” continues Gao. More prosaically, combining sensors with legible, user-friendly platform interfaces can also improve asset utilisation. “Despite the increasing demand for containers, many empty containers are sitting in the port and being forgotten by stakeholders,” he adds. “Universally using sensors would improve the utilisation of assets for different stakeholders across industries.”

“Instead of spending energy telling each other stories about the pieces of data one does have access to,” sum up Finelli and Narasimhan, “everyone can combine forces to address the strategic questions that emerge from a full, comprehensive and shared perspective. This is a powerful change and likely to be a disruptive change. It challenges not only work methods and processes, but also existing hierarchies, power-systems, and mindsets.”

Less work, more skill

Most pharmaceutical companies are now fully invested in their own transformation journeys, aspiring to bring digital capabilities to clinical trials as soon as possible to create a more proactive approach to healthcare in an increasingly technology-enabled world. Russo wants to be at the very forefront of that development, which means he needs to be flexible.

“At Janssen, we see the number and complexity of trials increasing, their size reducing and new therapies like cell and gene therapies or radiopharmaceuticals being introduced,” he says. By way of example, he contrasts the technology required to supply traditional small molecule tablets with the ‘vein-to-vein’ tracking capabilities needed for a CAR-T trial. “Radiopharmaceuticals have an end-to-end supply chain from manufacturing to dosing measured in hours, not weeks,” he says. “Our future clinical supply chain must, therefore, be ‘multimodal’ and supported by a range of fit-for-purpose digital technologies.”

It’s not the only thing. The merciful decrease in the tedious, repetitive and inefficient elements of clinical supply work made possible by digitisation is counterbalanced by the need for professionals like Russo to expand their competencies into entirely new areas. Indeed, Finelli and Narasimhan stress that machine learning based on historical data allows Novartis to “identify” rather than “prescribe” the “putative drivers” behind process elements like patient enrolment. Rather than holding their own hard-won experiential models in their heads, then, Novartis’s digital ‘product owners’ have to be adaptable in combining the data science expertise to translate business problems into computational terms; a focused attention to users’ needs; a strong understanding of how to navigate the company’s IT process governance; and a command of leadership techniques like professional project management, ‘design thinking’ and agile software development.


Although Novartis has only just begun to quantify the Nerve Live programme’s impacts, Finelli and Narasimhan report that productivity gains in the order of 10% are already achievable across the full portfolio of drug development activities. The company is already measuring fewer inactive sites and faster enrolment periods, lower study budgets, and more efficient resource allocation. Its trials are healthier than ever. The next frontier, of course, is patient data. Novartis and AstraZeneca avoided basing their transformations on human-origin data that could be subject to privacy and protection laws, but it’s the checkpoint everyone’s racing to reach. Fittingly, though, the fact remains that most clinical trials make less of the data they produce than the average family car – and that’s about more than going fast. By digitising the operations that support their trials, Novartis, AstraZeneca and others are looking to prove themselves worthy of the same levels of consumer trust.


Estimated reduction in trial-associated waste that can be achieved with AstraZeneca’s digital clinical supply chain enhancements, saving the company $100m.



Productivity gains Novartis believes its digital transformation is making possible across drug development.