Applied Clinical Trials
Have you seen many paradigm shifts lately?
How many times have you read about the urgent need to shift the clinical trial paradigm or claims that say that Wonder Tool X or Super System Y will do this for you? Have you ever attended a clinical trial “Innovation” or “Disruption” conference and left wondering in all the years the conference has been held, what has really been disrupted?
Clinical trials are deceptively complex, and many are becoming more sophisticated in their design and demanding in their conduct. It is clear there have been advances in many aspects of the ways we perform clinical trials, particularly the benefits brought by the application of technology and standards. There are also many initiatives and ideas, new technologies, and approaches being proposed. So, as we head toward the end of another year and approach the 2020s, are we confident we are on the right path to more effective and cheaper clinical development and clinical trials? Based on various publications, trade press, conferences, and similar, there seems to be widespread acknowledgment that doubts remain. There is anxiety, if not worse, some disillusionment.
Over many years in management, when faced with situations or information to review, I have learned to ask myself some simple questions. For example, why, so what, what can we do? I don’t pretend to have all the answers, but when invited to consider this topic, I used this questioning approach. These are some personal thoughts and views.
Accepting that studies can be large and complex, have we, nevertheless, overcomplicated the problem? A clinical trial can be considered as basically a scientific experiment involving people. The clinical data generated from the subjects is key to determining the effectiveness and safety of the treatment. If this is the core data, then this means that all other data, and information, is ancillary.
Undoubtedly, this other data is important for many reasons: ensuring the study is well planned, managed, ethical, compliant, that the clinical data can be trusted, etc. Nonetheless, it does not determine efficacy or safety and so it can be considered ancillary.
If we accept this principle, then what are the implications? Have we structured organizations and developed and applied technologies that fail to appreciate that the clinical data is the core output of a clinical trial? Have boundaries been created in the wrong places and silos inadvertently created? We have many organizational groups managing different aspects of the clinical trial data. Clinical data is also spread across different systems and databases; it has become fragmented. When we try to integrate these often disparate systems, some never intended or designed to be integrated, the result is a patchwork of systems with inadequate interoperability and data all over the place. This is the so-called “Frankenstein” of siloed systems, so rightly highlighted earlier this year.1 Despite the laudable efforts of many to move to industry data standards, we still end up mapping and reformatting data and meta data. The result is not only wasted time and a high maintenance cost, but a serious impediment to our ability to have critical clinical data aggregated and monitored in real time. Instead, it can be days and often weeks before it is available together for review and interpretation.
Have we built too much on legacy? We add new technologies on to old, and often just tweak underlying processes and SOPs. (We even tried to shoehorn the paper world of good clinical practice (GCP) regulations and guidelines into the mobile, digital world of the 21st century). If a great technology addresses one problem or provides new capabilities, the unintended consequence can be rather than make things better, overall they get worse.
Over the last decade or so, management of clinical data has been driven down the commodity route, with off-shoring encouraged to save money. In some cases, job roles have been made narrower to allow for a more task-based approach with rapid training of less experienced resources. Nothing wrong with reducing costs, but with clinical trials becoming more sophisticated, the number of data sources increasing, and the types of data more complex, maybe this strategy needs to change. Surely we should be applying greater expertise and sophistication to derive valuable information from the data, and sooner? Clinical trial data is not an ancillary byproduct. It is the output of the clinical trial and arguably the whole purpose why the trial was conducted in the first place.
Sponsors take large risks, each spending millions, if not billions of dollars, on researching and developing new and better treatments or addressing unmet medical needs. Could those same sponsors spend or risk a little more on applying more innovation in the conduct and management of clinical trials? Do we as an industry prefer old, low-risk methods and systems that are tried and tested and are we too accepting of their limitations? The accusation is often made that the industry is too conservative. So what can we do to change this? By not investing more and driving change, are sponsors missing out on a tremendous opportunity to not only improve the efficiency of their operations but also on enhancing the value of their portfolio of R&D prospects? Should CROs be doing more, or are they indirectly held back trying to meet the requirements of their clients?
Consider that with many treatments becoming increasingly sophisticated, personalized, and “biological” (e.g. cell and gene therapy) could we be inadvertently failing or delaying in the application of modern technologies, such as live analytics and the use of artificial intelligence, that could help determine and more quickly prove the effectiveness and safety of new treatments? Is the boundary between controlled clinical trials and real-world evidence studies another unnecessary silo we have created, exacerbated by limitations of our technologies?
We still tend to focus on technology and systems separate from processes. Though there are indications in the market this is now changing, traditionally, clinical service vendors are also separate from clinical trial software vendors. So, looking back over the last two or three decades, how well has this worked? How modern, fit for the real world, and meeting the needs of its users are these systems and tools? Who believes we are at the forefront of all industries in applying modern technology and smart processes in clinical trials? Shouldn’t we be at the forefront? After all, clinical trials may involve us genetically modifying live cells inside people. The reality is today we often don’t know, from one week to the next, how these patients are responding to clinical trial treatments and assessments. Imagine if we didn’t have sensors on passenger jet aircraft monitoring 1,001 aspects of the plane and providing information to the pilots and to maintenance on the ground. Instead, every few weeks we asked the pilot how the plane was flying, or asked did anything seem not right?2
There are many consequences for patients and their families (i.e., ourselves too), doctors, nurses, study coordinators, etc. at the investigator sites and for our industry.
The “insane cost of developing new drugs” when considering the high and late failure rate is well known,3 so too the consequences, including potentially good or life-saving treatments remaining undeveloped. A third of all new marketed products are found to have serious safety issues not recognized before in the clinical trial data.4 Surveys continue to report significant barriers in participating as a principal investigator (PI), while half of first-time investigators say never again.5,6 Another consequence of the high cost of development is the very high price of many new treatments. Admittedly, clinical trials are only one component of the R&D cost, but, nevertheless, this cost is sizeable and a late-stage clinical trial failure or a wrong go/no-go decision can be disastrous. The high cost of new treatments leaves governments and healthcare payers reeling, struggling, or refusing to pay, while large pharma share prices have tended to stagnate with the response being many pharma companies have spent more on share buybacks and dividends than on R&D.7 It isn’t good business for the sponsor companies either.
Has the time come for the life sciences industry to truly learn from other industries? We are seeing mega large, global technology companies move into selected areas of clinical research. This should be positive, but aren’t the solutions to some of our challenges more fundamental?
In general terms, we can stop doing the same thing, and stop repeating the same mistakes. We should stand well back, look up, and ask ourselves: if we were to develop new methods and processes for designing, conducting, and managing clinical trials, and ensuring optimal patient safety, what would they look like? What are the outcomes we want to achieve? Even with blue sky thinking, the chances are most of it can be done today, or the right pieces put in place now to enable many advances, with the rest slotting in when ready/available.
We will also need a variety of software systems and tools to conduct and manage clinical trials. One system for everything is probably not realistic nor optimal. We obviously need speciality disciplines, experts, and organizational structures to support them. However, design boundaries in the right places and do not perpetuate legacy. Accept that clinical trial data is the core output of a study and place a boundary around this, not carve it up into silos within.
Why not have one purpose-designed system (or systems) with one database that can manage all the data for your clinical trial, rather than have separate systems and tools for eDC, ePRO, esource, lab, safety, econsent, reporting, analytics, etc. Avoid the Frankenstein of patchworked systems, all with their own databases, management, and support systems, and all the costs, inefficiencies, and delays they cause-even when “integrated.” Using one system, together with embedded workflow and communication tools commonly in use elsewhere, can unite rather than fragment clinical trial teams, allowing them to operate more easily as one team, including between sponsors, CROs, and sites.
Look carefully at the content and boundaries for the “ancillary” data. Is it any wonder that one of the favorite systems for clinical operations staff to complain about is the clinical trial management system (CTMS)? What does it actually do, or is it a legacy concept? Now we have electronic trial master file (eTMF) systems, trial supply logistics tools, investigator grant management tools, etc. Where should these boundaries be optimally drawn elsewhere? How do we want our teams to communicate and work?
Web portals designed to bring information from diverse sources and display, depending on the user needs and role, are a useful approach to overcoming the shortcomings of many separate systems and tools. However, they still require integration of many backend databases and can be complex and costly to set up and maintain. Why not avoid having so many separate systems and tools in the first place?
A modern smartphone can be a phone, a music player, a video player, a camera, a calculator, a diary, a computer, etc. It is not all those separate devices integrated. Why not have apps with all the functionality needed, communicating with a single platform around a single data repository, rather than all these separate systems?
A complete rework of our approach, technologies, and processes may seem scary, costly, and time-consuming. Indeed, possibly a distraction to developing new treatments. It needn’t be, though. It needs a simplified, back-to-basics and sound principles approach. There are solutions and technologies that already exist that can sweep away the past and present, the Frankenstein’s monsters and ghosts, and bring in a much brighter future. For those with courage, who ask the right questions, there is the opportunity to leave the rest way behind.
David Connelly, PhD, is Founder and CEO, Cmed Group Ltd.