Procurement and Technology Onboarding: How Big Pharma Picks New Tools
By Allison Proffitt
July 14, 2023 | A panel of pharma representatives talked about their tech acquisition pipelines last month at the Bio-IT World Conference & Expo. Representatives from Roche, Janssen, and Novartis outlined their strategies for acquiring new technologies and the pitfalls they look for.
Acquiring and implementing new technologies is a long and arduous process for sponsors, said Aaron Mann, founder and CEO of the Clinical Research Data Sharing Alliance, and moderator of the discussion. Across the industry, the average duration of the process—from initiation to full implementation—is 5.8 years, he reported. Each phase of the process gets progressively more challenging, and he warned tool providers that any internal pharma champions they have earned in the early stages would face more hurdles as the process progresses.
For tool developers, there is a fundamental misalignment between what their investors want and how the pharma industry behaves, he said. While investors are looking for “hockey stick growth”, that is rarely feasible for an industry that is highly regulated, making huge financial bets, and responsible for patients’ lives.
But economic and regulatory risks are not the only ones giving sponsors pause. Especially with new technical solutions, sponsors need to determine if a solution is trustworthy, if it’s the right approach to solve their problem, and if a particular vendor is the right group to partner with, Mann said. He outlined some of the questions sponsors ask themselves of new vendors: “Are you going to be around at the end of the six-year [acquisition] process? Are you going to be able to support it? Am I confident you’re not going to pivot?” Vendors must be ready to reassure potential partners that the technology is not only fit for purpose, but that their company is sustainable and equipped to deliver that technology, he said.
Power in Partnerships
Data show that sponsors are most often exposed to innovation via company collaborations, partnerships, and consortia, Mann said. “It is the leading place to develop a meaningful relationship over time, over that long process with the people who are going to be interested in what you are producing, what you’ve developed. It’s really important for you to get involved.”
He acknowledged that for small or startup companies, working alongside a competitor feels risky if not foolish. “But this is what powers innovation for sponsors. Find out what’s going to in your domain and think ‘Ecosystem’ and not, ‘I’m a point solution.’ Think about what you can do to move the ecosystem forward,” Mann recommended.
Laszlo Vasko, Senior Director of Novel Modalities Technologies, Janssen Pharmaceuticals, agreed. “One of the things we really struggle with, often, is the lack of cross-vendor discussions… We get a lot of pitches to say, ‘Hey, we’ve got this great ecosystem of all of our components or modules plugged together!’ but the challenge we have is that we can’t take all of those modules. We might only be interested in one module and we need it to work with all of our other systems. One of the things we always ask is: ‘Well, have you reached out to the other established vendors in this space?’” Unfortunately, often the answer is no.
April Bingham, Executive Director Global Medical Compliance and Governance Chapter, Roche, also echoed Mann’s advice. “I can’t really emphasize enough how industry collaboration really does allow you to achieve more, faster, than going direct [to pharma],” she said. “Really industry collaboration is a way in which we overcome barriers that we have, substantially, around regulatory requirements.”
For Roche—a pharma active in 90 markets around the world—regulatory issues are front of mind, Bingham said. Interoperability, scalability, and complexity are important, but Bingham said that her biggest current challenges are from GDPR and new regulations from China. “Historically we talk about computer software validation—and that’s actually my depth of expertise—I would say that’s less of a challenge than some of the new emerging regulations around privacy.” Bingham advised that vendors take time to understand the complexities of privacy regulations.
Peter Mesenbrink, Executive Director of Biostatistics of Novartis Pharmaceuticals, added security as well. “Security—and the legal pieces of the pipeline—are at the top of the list,” he said. “Even when people say ‘We’re big on IT security!’, sometimes they’re not.”
The intellectual property question is one aspect of security that may surprise smaller or new vendors. “IP concerns can be a surprise to the vendor side and a surprise internally because you’re using the company’s data to actually build the algorithm, which is then going to get commercialized outside,” Mann said. Bingham agreed that IP questions can be a “substantial hurdle” but said Roche has had challenges even with large vendors.
She recounted shifting to Google Drive for storage. “The Google guys said, ‘Are you kidding me, Roche? We’re Google! We’ve got a lot of IP to protect and you’re acting like your IP is more important than ours!’” Yet Roche created a security wrapper around Google’s security wrapper. “We often do—what feels like from the vendor side—exorbitant measures to protect our IP. I don’t see that changing any time soon,” she said.
All the Right Questions
Both Bingham and Mesenbrink recommended that sponsors begin the technology acquisition process by asking the right questions. When Bingham is advising teams about new tech, she warns against “Shiny Object Syndrome,” encouraging teams to ask “Should we?” not only “Can we?”. Her teams are looking for tools that reduce complexity, scale, are adaptable and operable over many systems and structures, quantify ROI, make the company more competitive, give better insight into the customer, and help achieve better regulatory compliance.
Mesenbrink proposed carefully distinguishing which customers are priority for each solution. Is a tool for exploratory research or for regulatory submission? How are data stored and accessed? Can statistical analysis be done within the tool or do data need to be transferred to another location? He encouraged considering not only the needs of today but trying to identify gaps to predict—even if imperfectly—needs in the future.
Conquering Complexity
Even when the right general technology has been chosen, narrowing the pool down to a specific solution from a specific vendor brings the complexities of the sponsor organization into play. Roche, for instance, is a decentralized organization with parallel and competing technologies within the organization. Bingham reported multiple data lakes; minimum data standardization; Microsoft, Apple, and Google environments; and multiple browsers. “This is probably one of the unique and challenging aspects of coming into a company like Roche that really puts that technology innovation at the forefront. It certainly challenges small startups to be able to build in all those different formats and to be able to interface with them,” she said. Building a tool or solution for Roche is challenging, Bingham acknowledged. To meet the needs of all the global markets and scale to more than 100,000 users is daunting.
In addition to his experience at Novartis, Mesenbrink has worked with TransCelerate for the past five years, focusing on data sharing and statistical analysis. Since the pandemic, he’s been additionally focused on assessing integrated research platforms as part of multi-sponsor platform trials. He believes that to be successful, modern computing environments need to balance processes, people, speed, reproducibility, future proofing, elasticity, standardization, and accessibility, Mesenbrink said. Even so, he acknowledged it would be virtually impossible to meet the needs of all scientific customers.
Deliberate Processes
Laszlo Vasko, Senior Director of Novel Modalities Technologies, Janssen Pharmaceuticals, gave an overview of how the entire process proceeds at Janssen. Assessing new technologies is organized into an innovation project lifecycle for new technology areas that can be illustrated with four “stage gates” between five stages of discovery, ideation, demonstration, scale-up, and deployment. A portfolio review committee (PRC) reviews the project at each stage gate to decide if it should proceed or be exited. It’s the same assessment process for new technical and non-technical new products, he noted.
The risk threshold decreases along the process, but the vendor time commitment increases as the PRC adds conference room pilots, proofs of concept, and live clinical trial pilots. For this reason, Janssen works with quite a few vendors early on, expecting many to withdraw as the process moves forward, Vasko explained.
The demonstration phase, he said, can take up to three years and in that time Sponsor stakeholders can change. New people step into the process. At the scale-up phase, he added, sometimes Janssen can surprise everyone and do an RFP. After spending a great deal of time seeking to understand the technical landscape, “We actually want to select the technology, the vendor, and the partner that we would like to scale this up,” Vasko explained. Choosing the RFP is hard on the sponsor as well as any vendors who they have been working with along the way, Vasko said. “To actually have—basically—the guts to say, ‘Well that worked pretty well, now let’s take a pause and select a different technology.’ It’s not an easy decision.” Echoing the concerns that Mann predicted in his introduction, Vasko explained that sometimes their concerns are with the technology, but sometimes they are with the viability of the vendor company.
Janssen has had failures, Vasko shared: twice for very critical capabilities. “I can say that in those two occasions we could not just simply say, ‘Oh well, the company failed. Oh well, we just have to decommission the system.’” Instead, he remembered, Janssen had to hire staff and secure the intellectual property to maintain the failing system while the procurement process began again. “The risk of failure of proceeding on with a technology that—either from an architectural reason or from the vendor perspective—potentially has large risks, could be very significant.”