What SIP Gets Right And Where It Still Falls Short
By Dan Schell, Chief Editor, Clinical Leader
After three articles about the Shared Investigator Platform (SIP), I’ve talked to some of the people who conceived it, some who helped build it, and current users who explained why the industry’s site-technology problem is bigger than SIP itself.
In Part 1, I looked at SIP’s original vision and the problem its architects were trying to solve. Part 2 explored what happened when that vision had to become an actual technology platform. And in Part 3, current users described an industry that continues to ask sites to maintain the same information across an expanding collection of systems.
For this final installment, I wanted to add two more site voices. Sophie Haynes, who works in regulatory at Hightower Clinical, has used SIP for nearly four years. Vicki Alvarez has considerably more experience. She supports independent research sites and has set up more than 10 SIP accounts. She’s become so familiar with it that when one of the sites she supports has a SIP problem, she’s usually the person who gets the call. (You can also hear more from Alvarez in her interview with Steve Wimmer on his 30 Minute Randomized podcast.)
Despite what you might expect after reading the first three articles, neither woman told me SIP is devoid of value. Instead, their experiences raise a more useful question: Does the value SIP provides justify the work it requires from sites?
SIP Does Some Things Well
For example, she likes Exostar (an identity and access management service used to access SIP)because the same login can be used across multiple sponsor systems. She has also seen SIP used to electronically sign 1572s and allow PIs to acknowledge protocol amendments. Like some other users I spoke with, she gives SIP’s help desk kudos for eventually resolving problems, including a complicated situation involving the merger of two accounts.
Alvarez also offers a more nuanced view of the burden associated with maintaining information in SIP. Once a site is established in the system, she says the ongoing upkeep isn’t necessarily enormous. GCP certificates need to be updated when they expire, study experience may need to be added, and other information occasionally changes. The much bigger challenge is setting everything up.
The PI must register, create the facility profile and designate someone to manage it. Site personnel then need to create individual profiles and upload credentials such as GCP certificates and medical licenses. “You have to send an invitation from the PI’s profile. What PI is going to take the time to do that?” Alvarez says. “None. None whatsoever.”
The TransCelerate CV can create a similar problem. An experienced PI might have dozens of studies to enter, yet login credentials aren’t supposed to be shared. The process may technically belong to the investigator, but operationally, someone else at the site still has to find a way to get it done.
That heavy up-front burden could produce a worthwhile return if — and this is an important if — sponsors consistently reused the information. Alvarez uses something as simple as a site’s local laboratory information as an example. If it is already stored in SIP and does not change, having multiple sponsors pull from the same profile could save time.
That is the version of SIP’s value proposition she can see working. The problem is that sites don’t always experience SIP that way.
One More System To Keep Up With
Unfortunately, the next word in that sentence is but.
“But, with SIP it hasn’t really made processes easier,” Haynes says. “A lot of the time, we’re still uploading the same documents into multiple systems because not every sponsor uses the same platform. So instead of replacing other portals, it just becomes one more system to keep up with.”
Yes, I know that sentiment has been expressed by multiple people in this series, and I also know that contradiction is at the center of probably every site’s SIP experience. But I think it bears repeating: A shared platform becomes considerably less valuable when it is not shared broadly enough to replace … anything.
Haynes also calls the abbreviated CV requirement “extremely repetitive” and agrees that each sponsor seems to use SIP differently. The platform may be the same, but the workflows aren’t necessarily consistent, leaving site personnel with another sponsor-specific learning curve.
One of her other examples seems minor until you think about how often these little inefficiencies occur. Haynes receives daily notifications informing her that new regulatory documents have been released. She follows the link and logs into SIP, only to find that the document usually isn’t yet available in the study’s Document Download section. Sometimes it takes a couple of days to appear.
No, that’s not going to derail a clinical trial. Neither will an extra login, another profile update or having to figure out where a sponsor placed a particular function. But multiply those small tasks across staff, studies, sponsors and systems, and you start to understand how technology burden accumulates.
Usability adds another layer. Alvarez says what sponsors see within SIP isn’t necessarily what site users see. A sponsor may tell a site that information is missing, while the site user can’t locate what the sponsor is referencing. Sometimes the two sides resort to exchanging screenshots simply to determine where the information belongs. “It’s very cumbersome, not intuitive, and there are a lot of hidden features,” she says.
What Work Actually Disappeared?
Alvarez doesn’t deny that SIP creates some efficiencies. Her comments made me question whether those efficiencies address the work sites most need technology to eliminate.
Being able to electronically sign a 1572 or acknowledge a protocol amendment is certainly convenient. But Alvarez points out that getting a PI to sign those documents wasn’t necessarily one of regulatory’s biggest operational headaches in the first place. SIP can make that task easier without making a meaningful dent in the larger workload created by startup, sponsor-specific processes and repetitive information requests.
It’s interesting how we evaluate clinical trial technology by the functions it adds. Can it digitize this document? Automate that step? Move another process online? Those are fair questions, but the experiences described by Haynes and Alvarez suggest sites need the industry to ask another one: What work did this technology actually make disappear? That’s a tough one.
For these two users, the answer appears to be: not enough.
That conclusion doesn’t mean SIP provides no value. Haynes still believes in the idea of putting everything in one place. Alvarez likes Exostar, sees the potential value of reusable site information, appreciates certain electronic-signature functions and gives the help desk credit for helping her resolve problems.
Their frustration is that those benefits coexist with a difficult setup process, repetitive CV requirements, inconsistent sponsor workflows, delayed document availability and features that can be hard to find. In other words, SIP can make certain tasks easier while still leaving sites with much of the administrative work the platform was supposed to reduce.
That may be the fairest place to end this series. The original idea behind SIP remains appealing because the underlying problem remains familiar. Sites still want fewer logins, reusable information, consistent processes and technology that reduces work instead of simply giving that work a new place to live.
More than a decade later, those users are not saying SIP does nothing for them. They are saying it has not yet done enough.