Jun 26, 2019
by Brian Royer | June 26,2019
I have had the opportunity to attend recent life sciences and healthcare conferences and learned from their early adopters how Cloud is being applied by their organizations to accelerate research, enable clinical collaborations and get new drug therapies to market faster. I also learned that, in life sciences, Cloud is gradually removing computing and application delivery bottlenecks, streamlining workflows and reducing infrastructure costs.
Still, there’s plenty of trial and error.
The biggest takeaway is that Cloud is not plug-and-play. Its learning curve, at least in cases where the technology is either foreign to or routinely dismissed as “not for us,” can be significant. In many cases, it requires new skills, some of which may have to be acquired by the internal IT staff tasked with migrating applications or systems to the cloud. For those organizations that choose to partner with a third-party vendor to help them to successfully achieve cloud migration, there needs to be an “alignment” between internal IT, research scientists as well as those vendors on many levels, otherwise categorized as “the vernacular.”
This last aspect has some very practical implications: life science researchers are not IT experts. In fact, scientists can’t tell you what they want. They first need to see it. To test it. To understand its properties and predict how something will behave, and to achieve a level of confidence on what outcomes can be expected of it in the real world. And, of course, their bias to learning about a subject often trades on discussions around genomes, therapies, and trials, not the Cloud. In other words, you can explain cloud to them on a 50,000-foot level, but in their realm, it’s only a means to an end, not how the drug therapy they’re working on actually gets to market, especially where adherence to rigorous testing and strict regulations is their primary concern.
As a result, in helping to relate Cloud to these particular stakeholders, IT must become (or perceived to be) a strategic partner to R&D. This includes:
Developing a standard data model to facilitate successful innovation and trials
Defining R&D processes to enable data-driven decisions
Focusing resources on what is actually strategic to your organization’s R&D methodology. (In other words, don’t waste time building fundamental infrastructure; instead, choose a path that is easy to configure, integrate and extend)
Demonstrating Cloud’s ability to provide an efficient way to enable clinical trial sites (albeit virtually) and for R&D groups to collaborate across the globe
Additionally, just because you’re a Cloud evangelist for your life sciences company and you’ve been able to win over your CTO, your work isn’t done. In fact, it’s only begun. For example, there’s certain to be initiatives around protecting sensitive and confidential data. That is, dealing with the lack of coordinated authorization for usage, especially where internal authentication and access credentials were left unmanaged, never mind aspects such as PII and NIH related security requirements. Cloud doesn’t make any of this go away. In fact, it only underscores the approach you take must be deliberate and defined well ahead of implementation.
There’s also the 800-pound gorilla in the room, that Cloud is not free; although over time, as a result of reduced overhead (e.g., administration), labor, scale back infrastructure build-outs globally, and so on, its ROI (return on investment) can be substantial.
Still, its cost needs to come out of someone’s budget even as resources are reallocated to other projects. Also, since this data is now hosted external to the organization and no longer under its complete control, more than one presenter recommended “you also need to get legal involved” especially in worst case scenarios where offsite data has been breached and is now exposed to be sold to the highest bidder or even used criminally. All of these costs, opportunities and yes, even threats, must be accounted for.
In fact, omnipresent in these myriad session tracks was an almost over-sharing of information that highlighted both outcomes. Interestingly, while the circumstances that made Cloud such an attractive option was unique to each organization, in many cases the lessons learned, as well as the takeaways, were remarkably similar.
So, as an influencer, you’ve decided to invest in cloud and you’ve now convinced your CTO of its benefits. How do you get started? What are the key areas in which you need to spend money up front?
The consensus was to use a carrot, rather than a stick approach.
For “bean counters”: Cloud preserves capital dollars, turning more of IT into a flexible operating budget line item. For operations and data crunchers: helping them to recognize that manually searching data is time-consuming and the older the results, the harder they are to track. Show them as the company (and its data) grow, data capture — using the cloud — can scale alongside it. Finally, for technically inclined users, pledge that any data generated can be easily converted and pushed to a data warehouse, (either through a Web-based UI or direct data upload via an API), allowing the organization to integrate data from multiple applications and sources into one location.
As a result, software development processes are streamlined; so too centralized decision making. And maybe even your first hire in data science to help your organization interpret data for the purpose of decision making and identifying opportunities for digital transformation that better deliver your brand promise and brand experience.
However, while we’ve seen that cloud is finding traction in these kinds of organizations, there're trade-offs to be aware of, including:
Less flexibility to change how and where data capture takes place
Changes taking more time due to expanded data access restrictions
A higher learning curve that can result in increased user restlessness and frustration
More troubleshooting when new additions are made to the upload system
In fact, when preparing for cloud migration, there is more to it than identifying a cloud provider brand to partner with, including consideration (e.g., due diligence) of all the following:
Application assessment — which applications are candidates for migration, and why;
Deploying a cloud-specific team — after all, this is NOT a traditional infrastructure team so set expectations accordingly;
Management buy-in is a must — there will be many new and unfamiliar operational processes that will be coming online;
Make certain departments not directly involved in cloud migration are aware of these changes so they understand how they may (and probably will) impact their KPIs;
Get finance and legal involved as soon as possible;
If you need a Cloud infrastructure provider to partner with, do your homework, and
If you need help with any of it, go out and find an experienced vendor to partner with
Additionally, when it comes to IT (which will generally be responsible for all the “heavy lifting” associated with cloud migration), keep in mind all of the following considerations that may directly impact this process, including:
Network & Connectivity
All of these cloud migration “hacks” aside, one of the more memorable presenters had perhaps the most concise guidance on how life sciences companies should approach cloud migration and it bears repeating here: Start Small, Get Experience, Get Training. (To which I might add “Get Buy-In”).
As many if not all of these sessions revealed, there’s no doubt that when it comes to expediting collaborative research, speeding the result of that research to market and, ultimately, to the patients who would benefit most from it, Cloud in life sciences is increasingly the rule, rather than its exception.