By Thomas Cocciardi, LMK Clinical Research Consulting
The Veeva 2019 Unified Clinical Operations Survey examines the extent to which the clinical trial industry has sought to overcome its dependence on outmoded processes through the implementation of clinical applications.1 The survey’s main objective is to grow understanding of the “drivers, barriers, and benefits of a unified clinical operating model” through the exploration of clinical stakeholders’ attitudes about the clinical applications they currently use. The survey, gathering the experiences of 461 clinical research professionals, well characterizes the industry-wide frustration with the inefficiency and opacity of clinical operations and clinical applications. The survey outlines that adoption of multiple forms of clinical applications has accelerated to the point of ubiquity: 88 percent of sponsors and CROs utilize electronic data capture (EDC) systems, 69 percent utilize electronic trial master file (eTMF) applications, and 63 percent utilize randomization and clinical supplies management software (RTSM) applications. eTMF adoption has increased 12 percent since 2017 and, yet, 97 percent of respondents say they have at least one major challenge with their clinical applications, with 83 percent reporting two or more challenges. Nearly all respondents (99 percent) report the need to unify clinical applications to achieve better visibility and oversight, speed, and collaboration.
Shockingly, despite the recognized need for increased clinical operations transparency and efficiency through improved clinical applications, 81 percent of respondents continue to report the use of manual processes (like spreadsheets and emails) to manage study start-up. As cited in the Veeva survey, study start-up is no faster than it was a decade ago. Such stagnation suggests that clinical applications, despite their widespread adoption, have not fully met the needs of users and are not well integrated into the flow of information across a clinical trial. Instead, it appears clinical applications are working alongside (or even in opposition to) undocumented, but critical, manual processes.
Clinical applications, in my opinion, are successful when they are accepted by users—and they are accepted because they achieve three key criteria: speed, accessibility, and flexibility. By this measure, email and Excel, despite their obvious drawbacks and compliance concerns, still surpass the most advanced clinical applications. As a result of this acceptability gap, clinical operations professionals are not using clinical applications at all, are using them begrudgingly, or are using them in parallel with email and Excel. Unless the clinical application accessibility gap is closed, clinical applications will continue to proliferate — but fail to deliver on their lofty promises.
To meet the performance expectations of CROs and sponsors, slow simply won’t do. As the adoption of clinical applications continues to grow the availability of time-based metrics, clinical research professionals feel mounting pressure to produce more in less time. Coupled with the explosion of large multicentered trials, clinical research professionals must now process and dispatch information more quickly than ever. Because of this time crunch, clinical research professionals cannot afford to be patient with clinical applications that slow down the pace of work or add administrative tasks that do not align with their performance metrics.
Email can instantly deliver documents or generate queries as quickly as the user can type. Messages are delivered nearly instantly, and attachments can be retrieved in seconds on nearly any type of workstation. Message content from other applications can easily be copied into emails and complex problems can be tackled asynchronously or in real time by multiple stakeholders. When the information sent or received by a clinical trial stakeholder must be analyzed, Excel offers users powerful tools in each spreadsheet, and the resulting analysis can be shared easily with anyone via Microsoft Office. While other data processing tools are reserved for experts, users of Excel can intuitively and graphically process their own data to make informed decisions and stay organized without specialized IT or statistical knowledge. Most importantly, data input is usually as quick as copying from one spreadsheet (or email or other Microsoft Office document) to another.
Meanwhile, for example, a cumbersome TMF system may take minutes to both upload and apply metadata to a single document. Clinical trial management systems (CTMS), in particular, are known for requiring their users to negotiate many layers of dropdown menus, fill dozens of required, but inapplicable fields, or repeat the same administrative actions over and over again. Users, instead of spending hours interacting with an inflexible application, make their own spreadsheet trackers in a matter of minutes, copy documents into personal folders, or leave files flagged in their inbox for ”later.” For often overworked clinical professionals, the clear priority is achievement of activation, enrollment, close-out, or other stage gates. They know their performance is judged by propelling the trial forward, not by their use of the available clinical applications.
Clinical professionals are caught between decision-makers who appraise their performance based on speed and clinical systems that prize compliance above all else. The clinical applications of the future, therefore, must fight this pressure to interact with clinical applications “later” by providing clinical professionals the ability to work faster than they would using email and Excel. Smartly designed SOPs, work instructions, and processes can help align users’ goals with the use of clinical applications. Regardless of the clinical application, speed and reduced regulatory risk are a trade-off that must be continually reexamined.
No matter how effective, if users cannot get past the login page, a clinical application is useless. During a clinical trial, where days and hours matter, it is commonplace to wait weeks to access critical applications. In order to gain access, everyone from senior decision-makers to new interns must request permission from (often) distant administrators and complete the same mandatory training. While waiting for a welcome email or spending hours clicking through training, the trial has already begun and the backlog is growing. Delays like these encourage credential sharing and other noncompliant behaviors. Meanwhile, while waiting two weeks to access the CTMS, for example, users have access to email and Excel from the first day they log on to their workstations.
Access control, of course, is a central component of 21 CFR Part 11,2 the regulatory basis for the FDA’s strict oversight of clinical applications. The regulation states that applications must be designed so that privacy is assured, entries and access are attributable, and that, overall, any data generated by clinical applications are protected. All of this is underpinned by the use of unique usernames and passwords. Because of the regulatory importance of access control, options for process improvement are limited. In order to compete with the instant accessibility of email and Excel, sponsors and CROs should focus on what can be controlled: the period between when a user needs access to a system for his or her role and when he or she actually gains access. This latency period greatly reinforces the alternative manual processes clinical applications aim to replace. Access procedures should be integrated as part of a structured onboarding process. Training requirements that are prerequisites for access could also better reflect the roles of users or differentiate between those with read/write access and those with read-only access , employing a risk-based training approach. Hardware or software solutions for credential management also have the potential to eliminate much of this unnecessary barrier to user engagement. Successful implementation of any clinical system requires winning the attention and admiration of users. Access delays, lost passwords, and lockouts nudge users to quietly abandon a new platform for their old routines.
A clinical trial can quickly devolve into two separate worlds: what is and what should be. Clinical applications, unfortunately, are often designed for the latter and are configured for an ideal world that rarely unfolds during the course of a typical clinical trial. Criticizing the inflexibility of clinical applications should not disregard the challenge of designing a clinical application in the first place. Regulatory expectations, unfortunately, must take precedent over the user’s experience. Designing such systems to flexibly yield to procrastination, incompleteness, noncompliance, and the downright unexpected would be unwise, impossible, or unfeasible. Current clinical applications are rigid and, as a result, each small unexpected event slowly drives a wedge between ideal and actuality. As errors, omissions, and backlogs grow, users feel they can no longer use a clinical system at all because their current scenario (what is) is irreconcilable with the available resources.
For example, consider the TMF Reference Model, a standardized filing taxonomy that has facilitated the growth of the eTMF and TMF interoperability. The Drug Information Association, stewards of the model, caution “The Model is not intended to be taken and used off-the shelf.”3 Why might this be? Presumably, because supporters of the Reference Model have seen inflexibility institutionalized within eTMF systems. The Reference Model, as an “analog” clinical application of sorts, lacks the flexibility needed by most clinical trials. Complete adherence to the Reference Model, without considering the needs of a clinical trial, has in many circumstances prompted the filing of an overwhelming number of duplicate and extraneous documents. An abundance of duplicate documents often plays a contributing role to more serious TMF inspection findings.
This kind of inflexible adherence to the Reference Model has also inspired an increase in manual data entry required for eTMF administration and an overall explosion in eTMF system complexity. But, perhaps most damagingly, lack of flexibility in eTMF systems has habituated users to approach TMF completeness with a checklist instead of appraising the documents in the context of the needs of the trial. As a result, users default to Excel document inventories that may or may not mirror the current state of the eTMF. Even when eTMF systems provide more advanced reporting functions, users copy their contents into Excel when it is time for the “real work” to be done.
In the eTMF, and most other clinical applications, it is clear that when users encounter a situation for which their clinical applications cannot compensate, they retreat to where they are most comfortable and most empowered to develop a workable solution. Again, due to regulatory requirements, clinical applications must be structured and secure. Off-the-shelf applications are also inexpensive, but the costs of inflexibility can be daunting. An eTMF system or CTMS system should not be as insecure as an email inbox or as revisable as an Excel tracker, but users should at a minimum know who to contact when system functionality and users’ needs are no longer aligned. Flexible functionality is now part of the design of some of the latest clinical applications, but even as these features become available, utilization remains poor. Communication between the IT professionals and administrators that configure these applications and clinical operations end users continues to be unidirectional. The benefits of flexible and/or customizable clinical applications, although substantial, can be abstract, while the costs to customize off-the-shelf systems are easily calculated. Still, clinical applications are implemented to solve challenges, not immortalize them. As the challenges of clinical operations change, our applications must respond to remain relevant.
Give The People What They Want?
I recall, in particular, the first few days of a role where I was required to complete the basic training of a TMF system before gaining access to the TMFs I was assigned to evaluate. Absent-mindedly following the prompts of the online training through to completion, I sent an email to the third-party representative to request access. A day later, I received no welcome email, so I followed up with the TMF vendor’s representative. Another day passed, and I received a curious response: I was assured that I had not, in fact, completed the training. It was as if I did not exist in the system at all.
It was later discovered that, in reviewing the email instructions to complete training, I had been given somebody else’s training credentials in error. I unflinchingly logged in with these credentials (although I do remember wondering why the username seemed to refer to someone else). I then completed the training using the wrong account. Even when the error was clear, neither I nor the eTMF representative were sure who to ask to ensure I received credit for the training. My access was still delayed, and it seemed like the only solution would be to take the hours-long training again, using the right credentials. While waiting for access, documents queued in my inbox and I hastily created an Excel tracker to bridge the gap. I became part of the statistic using manual processes.
The situation was eventually resolved without retaking the training, but speed, accessibility, and flexibility were not a part of the narrative—and the backlog of documents haunted me for months. As already discussed, there are good reasons clinical applications cannot conform to our every whim. Outside of a regulated industry, convenience, consumer preference, and market forces shape the applications we use. Within regulated industries, outdated regulations cause clinical systems to focus on placating regulators rather than the needs of users.
Deprived of a place at the negotiating table, clinical research professionals have been shamed into believing their own needs are unreasonable and have stopped advocating for them. In response, they have, in the most critical moments of the trial life cycle, abandoned modern clinical applications for the tried and true tools conveniently located right on their desktop. Instead of fighting the siren song of spreadsheets and email attachments, however, clinical applications designers and those choosing to implement clinical systems should learn what makes their primary competitors so enticing. Clinical applications represent a finely balanced compromise and we simply can’t have it all. But we can make more equitable concessions that ensure all stakeholders are fully engaged. To bring the efficiency and transparency to clinical research, we may have to give clinical research professionals some of the speed, accessibility, and flexibility they have desired all along.
About The Author:
Thomas Cocciardi is a technical writer at LMK Clinical Research Consulting who is committed to expressing the human story behind each trial master file (TMF). In addition to technical writing, he also works as a regulatory writer specializing in FDA Center for Tobacco Products (CTP) submissions. Cocciardi has gained wide-ranging experience, both in and out of the pharmaceutical industry, working as a TMF-focused clinical trial associate, TMF health specialist, clinical research coordinator, preclinical data coordinator, intern medical writer, and intern brand writer. He holds an M.S. in regulatory affairs for drugs, biologics, and medical devices from Northeastern University and the RAC (US) credential.