OR WAIT null SECS
Pandemic-forced move to remote monitoring calls for industry-wide collaboration.
A simple question: Will remote site monitoring (and remote patient monitoring), adopted out of need over the last year, become permanent assets in clinical trial protocols?
The simple answer is yes. The success achieved in the creation of the coronavirus vaccines, from speedy development to speedier authorization, leaves little doubt that remote monitoring has value. COVID-19, said Samuel Volchenboum, MD, PhD, MS, University of Chicago Medicine, forced the issue of enrolling patients and running trials remotely.
But define value: how and to what extent? Is the value sought in a piecemeal fashion, as in wanting cost savings from fewer site visits? Or by focusing on patients, making enrollment and adherence more attainable by allowing use of apps or wearable devices so they can participate from wherever they live? Or just looking for those anomalies, findable in real time, which therefore allow protocol tweaks if need be?
No, said those interviewed for this story, the sole way to get total value out of remote monitoring is to design the entire trial with checks and balances, recording data in real time so there is control and transparency, ensuring all sources are traceable, trackable and analyzable. In other words, trying remote monitoring without imbuing the protocol with risk-based quality monitoring from stem to stern will save neither money or time.
Which brings us to discussion of the complex answer. Before COVID-19, these checks and balances were available, but few sponsors wanted any part of them. Using source data verification and electronic data capture was familiar and comfortable. And most companies invested in their own configured systems developed in the 2000s. Those systems didn’t include the necessary and facile software as a service that can pinpoint the across-data-type anomalies that we have today, said Lisa Moneymaker, vice president, Clinical Operations Technologies, Medidata. “Tech [inside industry] couldn’t
Further, Penelope Manasco, MD, MDMS, founder, MANA RBM, added that having new tech doesn’t change mindset. “I think traditional methods are deeply ingrained. Whenever we talk to a [client] they say, ‘but this is how I learned to do clinical trial oversight.’ It is a mind switch.”
In addition, Moneymaker points to an oft-discussed industry characteristic: “Industry is slow to pivot; it’s one thing to say be risk-based, but in many ways, we are extremely risk averse,” said Moneymaker. However, she noted that if industry doesn’t start making changes such as incorporating transparency trial-wide and monitoring data in real-time to ensure data sources are verifiable, industry will find it difficult to reach its desired ROIs.
Which leaves advancing remote-based quality monitoring where?
It’s not as if any stakeholder has to go far to see what remote-based monitoring can do. A relatively ordinary event, set into motion 21 years ago, precipitated the following: a diagnosis of pediatric acute lymphoblastic leukemia (ALL) today is no longer an absolute death sentence as it was 50 years ago.
In 2000, two pediatric oncology groups merged to become the Children’s Oncology Group. Soon after, the human genome was mapped, allowing researchers to delve into the genetic reasons why a child can develop ALL,1 giving pediatric oncologists better ways to diagnose and treat. In 1975, the death rate was 1.1 out of 100,000 children. By 2013, the rate was 0.2.2 Today, most children diagnosed with ALL make the five-year survival threshold.3
But these aren’t the only reasons, said Volchenboum, also director of the university’s Pediatric Cancer Data Commons. The trials’ uniform design, standardized data language, use of central monitoring, and the heterogeneous patient population have all contributed to that survival rate.
By running these trials consistently across sites, with all data trackable, traceable, and compared to baseline, with protocol constantly followed, COG achieved “[five-year] cure rates for acute leukemia to go from 0 in the 1950s, to over 90% for acute leukemia,” he said.
Elsewhere in the world, Beijing specifically, researchers looked at how they could implement strategies that would reduce the impact of COVID-19 on staff and patients involved in 572 oncology trials. The authors established a remote management model, making changes to trial processing, enrollee management and quality control. Measures used to achieve compliance rate included differences in rate of protocol compliance, patients lost in follow-up or trial withdrawal and deaths. The study published last month, said the researchers obtained a compliance rate of 85.24% for 3,718 patients, during both COVID-19 outbreaks. No one was diagnosed with COVID-19, the authors said.4
A few numbers, a few comments.
Patil said if the trial design allows review of patient data remotely and holistically, sponsors will learn what analytics can do to safeguard endpoint information, such as finding if there are unreported endpoints and pinpointing incorrectly enrolled patients. Pharma also learned that patients stay in trials if they can report remotely. “They don’t have to drive to the site,” he said.
Agreed, said Oracle’s Jim Streeter, global vice president, Life Sciences Product Strategy. “They stay in the trial if it is remote. That is one thing [the industry] learned.”
Designing a remote-based quality monitoring trial is work, said Patil. Upfront, it requires more rigor around planning and design to ensure it is quality-focused. With a more seamless upfront design, and more patient adherence to on-going site compliance, data can be locked for analysis faster. “We can close 30% to 40% quicker than in a traditionally run trial,” he said.
While some specialties may work better with remote site monitoring, like ophthalmology, many specialties work well with both remote site and remote patient monitoring. Studies on the latter often focus on oncology. One oncology study that examined remote patient monitoring noted that decentralization could improve the trial’s adaptability, as these trials have sick patients who aren’t close to a test site, and other complexities.7
Matossian conducts trials primarily on dry eye disease. The five trials her group recently started or will start soon all have remote site monitoring. Before COVID, Matossian would conduct patient observations and fill out various paperwork. Monitors would come on-site to review the papers, ensure consents were properly filled out, etc. Now all data are sent via portal via scan and upload. For her, remote site monitoring means cost savings for sponsors because the information is collected via portal, therefore, not requiring costs related to travel. COVID-19, she said, has “created a new way of looking at things. Moving forward, we have to be nimble and adapt.”
Regulators in the US and Europe were prescient: both FDA (in 2013) and EMA (in 2010) told sponsors that they embraced remote monitoring. Said FDA: The agency “specifically encourages greater use of centralized monitoring methods where appropriate…”8,9
And both agencies asked sponsors to rethink its definition of what constitutes an important error. Capturing a wrong age or comorbidity shouldn’t affect outcomes as long as those errors are distributed throughout the trial sites.
But, FDA wrote, missing an adverse event could alter an endpoint. The EMA noted that resources should be used to address “the most important issues and priorities, namely patient safety and the quality of trial data.”8,9
Manasco applauded the regulators. “They came out as visionaries.”
What is super important, said Patil, is to know what the critical data points are and to make sure they have integrity and quality. “Every single database has errors; they are never 100% error free.” All endpoint data—the lab work, safety checkpoints, patient eligibility—must be congruent. They must be analysis ready.
Manasco wants everyone to meet the Pancake Man. On the worksheet, a trial participant was three feet tall and weighed 200 pounds. He is a real person, she said, so most likely taller than a hobbit. Investigating the worksheet, she heard from a trial assistant that those figures were on the worksheet, so it is correct.
“We hear this all the time, they say data is right and we say it is not compatible with life,”
Medidata has uncovered possible data fabrication. In one situation, hypertension was included in the intake medical history, but no hypertensive readings were recorded throughout the study, said Moneymaker. In another case, a look at adverse events across the sites showed at least three events on average, but a site with data anomalies did not have any.
At Oracle, data management has lined up data against data for years. “Data management can create a set of queries so the outliers don’t happen,” said Streeter. An oncology study can have 800 to 900 edit checks that can be unimportant or not even necessary to check, he said. One of the more interesting anomalies that Oracle has found are male participants with pregnancy tests. There are the checks and balances, he said, whether caught by the data manager, the CRA review, or the statistical review.
Some inaccuracies are inherent in trials because people will be people, especially if they are children. Volchenboum said the problem with diaries and surveys is that people do not fill them in as instructed. “I used to see children in the waiting room with their diaries. They were supposed to take notes every night, instead of scribbling vigorously in my office.” But with a wearable, “we can see if they are wearing it. If not, nurses can call. Data will be much more accurate.”
Manasco agreed: “If you don’t police your patients, you won’t know how good the quality of your trials are.” And that may become more of a factor since studies show that participants like remote; they like wearables; and they will even use their own phone.10
COVID-19 might have led industry down the path to the remote-monitoring waters, but will it drink?
Pharma is more sipping than slurping. Sponsors have had a difficult time getting their heads around that it is more important to focus on executing the protocol correctly and ensuring that the right metrics are delivered than results, said Manasco.
One Oracle client has two out of its 18 trials using remote technology. Why? “They are still trying to figure out how to work it. This company saw the advantages of it and are trying it out,” Streeter said. There are a lot of questions from sponsors on how to add the technologies in.
But Volchenboum has some answers; five years ago he co-founded a company called Litmus Health, which is a platform for data collection for wearables in clinical trials. Litmus develops metrics for wearables. The algorithms need to be transparent, he said, otherwise, the sponsor couldn’t report the data because it wouldn’t know how the findings were calculated. The wearable has to be used in the trial for what it was designed for, per FDA.
Right now, said Moneymaker, industry is at a unique tipping point, because pharma trials that survived the COVID-19 clinical trial freeze-out are in hybrid transitional-remote monitoring mode. “Those that could pivot, already had pieces in place.”
Pharma’s trepidation is easing, said Volchenboum, because “It is too attractive a way to collect data.” FDA thinks so; The rules and regulations for using what it calls drug development tools were published in the federal register in November.11
A lot of conversations need to take place among players in the clinical trial space to achieve standardization across the board, Moneymaker said. Industry, vendors, purchasing, and companies like Oracle, Medidata, IQVIA need to mind-meld because the tsunami changes that a digital transformation requires demands nothing less.
Doing a side-by-side comparison of different oversight methods with remote quality assurance is a project MANA RBM’s plans for the near future. “We need to test how we do clinical trials better. We haven’t done that as an industry. We need to be data driven in oversight.”
Christine Bahls is a freelance writer and researcher