There are only a finite number of patients available in the world to recruit for a clinical trial. How do you go about getting them into your trial and, once you do, how do you make sure the trial is run efficiently and using the best new technology? Andrew Putwain speaks to Subrata Bose, head of feasibility operations and recruitment strategy at Bayer, about this issue.
“Failure to fill a trial on time is costing the industry millions every year,” Subrata Bose, head of feasibility operations and recruitment strategy at Bayer, says to start off our conversation. It’s a bold proclamation to make, but Bose believes that failures in the recruitment process lead to more costly delays down the line. And the evidence agrees with him. The need to save money in these processes is one of the key themes at every conference on clinical trials, and the trouble with filling patient quotas and making trials run in a timely fashion are key issues that are highly publicised.
Clinical study recruitment is a large and complex area. In today’s changing markets, attempting to ascertain what will or won’t work to attract patients to sign up for a particular study and keep them engaged for the trial’s run, while making sure that all rules are complied with, and that the trial is run on time and on budget, is an immense task. Many trials are coming unstuck at the first hurdle – filling up that sheet for patients to even get involved. In other words, one of the many commonplace trends currently dogging the trial environment is its failure to meet patient recruitment numbers.
“I think failure to include patients on time [is a problem],” Bose says. “Unplanned time is another key cause for delays in the clinical study industry, and I think it is the biggest challenge to the success of the trial.”
There is also the issue of resources. “Competition for clinical trial sites and participants has increased enormously,” he explains. “The number of trials has nearly tripled during 2000–2016.
“In addition, protocols are getting more complex and regulatory requirements have expanded as well,” he continues.
The numbers Bose shares are from clinicaltrials.gov, which is a registry and results database of publicly and privately supported clinical studies of human participants conducted around the world. They show the huge rise in the number of trials being undertaken and the patients involved in them. “Everyone is competing for a finite number of patients and sites,” he says. This means many trials miss out, or face severe and expensive delays. “The cost and time efforts needed to recruit the right patients at the right time result in an extensive amount of money poured into the study team,” he remarks.
Plan for success
Better planning and use of technology could increase the productivity of the trials. These details cover different angles including early planning, conduct of operational feasibility and proper research by companies investing in technology to enable data-driven decision-making, as well as improved computational forecasting with robust estimations on sites – as long as they’re all properly timelined.
“There is also the option of improving engagement with stakeholders,” says Bose. “For example, through using multiple platforms such as apps and mobile messaging, we can see that the industry is willing to empower sites to take the initiative with steps to improve recruitment.”
“We also see a trend of investing in organisational development, establishing dedicated functions to improve feasibility, forecasting, planning and monitoring,” he continues.
There is a network of companies implementing facilitated protocol review processes so that they can pick up design aspects of the protocol through multiple sources. An example of this is organising surveys and focus group discussions, and using technology to analyse protocols.
There is plenty more of that technology to come; the industry is now using advanced algorithms, informatics and data to select the best sites, estimate true patient potentials and other operational planning aspects. Hospital sites with patient databases have been preferred for inclusion in trials.
“The use of technology is growing. You can apply technology in operational planning through mathematical models, mining historical and published data to optimise operational aspects, and looking at the true potential of a site number for clinical trials. You can analyse historical data to estimate the proliferation of the study, look at scenarios produced by mathematical models to estimate the rate of recruitment, and so on,” explains Bose.
The technology that’s now available can be used to understand the standard of care, conduct surveys, ascertain the opinion of the investigators on the particular study design and quickly analyse huge amounts of data to assess the feasibility aspect of the protocol.
“Today, you can apply natural language-processing algorithms on a vast amount of past protocols to understand how they are all designed and how your protocol elements differ from similar ones,” he says.
Electronic medical records
The electronic medical record (EMR) has created a buzz in the industry. Many hope the system will make sorting through anonymised patient records easier, and aid the selection and recruitment process. Bose agrees, yet, he still sees the concept as a work in progress. His view is that EMR is going to be a great asset further down the line because it doesn’t have the necessary algorithms, technology or background service-centre elements to be that useful yet.
“EMR could potentially lead to an uptake in recruitment in the future,” he says, “but currently, it doesn’t offer that much help in the recruitment process. It’s evolving and the technology behind EMR is going to provide benefits.”
In the healthcare setting in hospitals, for example, anonymised health records of the individual could be used for recruitment of clinical trials by alerting healthcare professionals whose patients might be eligible for a study. “Data systems have the potential to aid the trial recruiter through the referral process, allowing more people who are available for the trial by diagnosis to be seen,” he explains. The recruiter would be able to see their clinical profile very quickly, which could enable a redesign and a massive outpatient approach.
But the problems are manifest; many of them come back to insufficient data rather than any persistent issues in the system.
“Unfortunately, we do not have a readily available set-up on granular digital healthcare data covering the entire healthcare spectrum,” Bose says. “It’s impossible to find centralised healthcare records or systems at a regional level, let alone a national level, which could aid with more specific types of patient recruitment. Integration of disparate systems, standardisation of records and triggering the plug-in within the existing healthcare system could potentially change the practice in the future, but it needs massive investment.”
Another application of anonymised EMRs is checking the feasibility of the protocol, such as the inclusion and exclusion criteria. This involves simulating the criteria and checking the available number of potential patients.
“It would help to add the inclusion/ exclusion criteria before you finalise a protocol. Say, you’re looking for a patient population with a particular condition. EMR analysis can indicate whether by increasing the age limit slightly, you could include more patients. This can be done by extrapolating from a small amount of completely anonymised data.”
While this may sound like a simple tweak, it is far from straightforward. This would call for a platform with conversion of exclusion/inclusion criteria into searchable codes built into it that is readily available. Also, it needs good-quality granular medical records with health data such as lab reports, physicians’ notes, a good algorithm that can process unstructured text and a physician who can ensure correct interpretation.
“Different countries and regions use their own structures and platforms. We need more data for handling [that],” Bose remarks.
The problems that he sees with recruitment are commonplace and, while technology is a great help, it requires investment and ideas across the board to provide the necessary infrastructure for the trials.
“There is a simple solution: better planning and technology,” he continues. Problems on the ground need to be worked out before a trial begins. “A more cohesive strategy is needed. The patient is a key stakeholder, and better use of technology to understand patients’ insights and standard of care, reduce their burden and bring their views to the study design earlier would add a lot of value.”
It’s also integral to look at more use of data and analytics for decisionmaking, especially as EMR develops, and implement better outreach to patients and other stakeholders. Personalised outreach is the way to go. And these changes appear to be on their way, but whether or not they will arrive in time to help fill those new trials is another matter.
Subrata Bose
Subrata Bose is head of feasibility operations and recruitment strategy within pharmaceuticals at Bayer. Before joining Bayer in 2015, he gained more than 11 years’ experience in strategy consulting and clinical research, leading large-scale R&D strategic projects with major pharmaceutical companies.