Data-handling in a clinical trial is extensive and time-consuming work, often prone to human error. But this can be eased by deploying bots that can mirror a person. Robotic Process Automation (RPA) is a line of rule-based code which can complete repetitive and mundane tasks or business processes where a human doesn’t add much value.
Even though it entails the term “robotic”, RPA is not a humanoid robot, but rather a robotics software analogous to a digital worker that can be automated to execute end-to-end office-based functions. A recent GlobalData Thematic Intelligence report on RPA estimates that the global market will be worth $20 billion in 2030. GlobalData is the parent company of Clinical Trials Arena.
Automation has been in healthcare industry for quite some time and the 2019 Covid-19 pandemic has ramped up the adoption of RPA. In the UK, the National Health Service (NHS) required preparatory Covid-19 tests before surgeries, so automation was used to extract relevant data from the results without putting a strain on the healthcare workforce. The GlobalData report indicates that healthcare, pharma and medical device industries accounted for 14% of RPA spending in 2021. Experts tell Clinical Trials Arena that while there has been a growing and solid interest in RPA from the healthcare sector, the clinical trial industry is yet to step into the world of bots.
RPA can help streamline data transfer processes and ensure their quality by having an auditable trail. It can also improve patient experience by minimising the variability in clinical trial processes and allowing remote data reporting. This can result in costs benefits as participants are less likely to drop out of a clinical trial. It might sound like RPA is replacing people, but experts say RPA is here to bring back the meaning and fulfilment of working life. Yet RPA is not a silver bullet, as it needs quality maintenance to ensure business continuity, and human judgment cannot be replaced by a bot.
Speed and accuracy of RPA
One of the notable advantages of implementing RPA in a clinical trial process is speed. Where certain tasks might previously have taken a week to do, they are now completed in a matter of hours, says Ronan Fox, senior vice president of applied digital solutions and IT at ICON. For example, RPA cuts down the effort to manually transfer data from clinical trial sites to electronic trial master files (eTMF). It reduces delays and data loss by detecting anomalies more quickly and reliably than manual review, he explains.
Indeed, a human can inherently make a mistake and that can impact a clinical trial, says Leon Stafford, UK country manager at Digital Workforce. “It’s about doing the work more quickly, but also more accurately,” he adds. Tracing back the human error can be a challenge, whereas RPA has a clear auditable trail, allowing the opportunity to pinpoint the exact moment when the mistake was made, notes Juha Järvi, Finland country manager at Digital Workforce.
RPA can also be programmed to follow “if-then” rules. “It is the same if you gave instructions to a human who uses a specific computer program for the first time to carry out a task,” Järvi explains. “If-then” approach can be used when analysing possible patient outcome based on different laboratory results as they are either binary, positive or negative, or within reference range, between numbers A and B, he adds.
The stability of RPA can also improve patient experience. Often, the clinical trial variability, such as changing processes in reporting data to different people, can frustrate a participant, especially if they are at a difficult time of their life, says Greg Killian, vice president and business unit head for life sciences at EPAM Systems. The patient experience can be improved by bringing in the consistency in receiving and reporting information, potentially lowering the dropout rate.
A research study showed that RPA holds the potential to be used in remote patient monitoring with heart failure. Shwetha R, software engineer at Tydy and author of the paper, says that RPA can sort out the data and alert the clinician if certain anomalies are detected from wearable devices. This would allow the doctor to advise the patient remotely if a hospital visit is needed, resulting in less traveling for bedridden people. Even though this study was done in a healthcare setting, it could be done in a clinical trial too, as a patient can send out their data from wherever they are, she adds.
After all these improvements, the cost benefit will follow, Killian says. If less time is spent on fixing data quality and dealing with patients who are having poor experiences and dropping out of the trial, then efficiency should start coming in.
RPA is not here to replace humans
One might fear that automating certain tasks and letting a robot complete them would mean that people will start losing their jobs. While it might be the case in customer service or financial industries where retiring employees are replaced by robotic systems, the healthcare sector will remain a human dominated industry, Järvi says.
Automating mundane tasks, such as appointment scheduling or data entry, would allow clinicians to spend more time with patients. This could result in people feeling more fulfilled at their jobs and work closer to their licence. “They didn’t go to medical school to become a data monkey that is cleaning up data,” Killian explains.
With the ongoing trend of the “Great Resignation”, people have become an asset as many employees are quitting and changing their jobs, notes Karl Mielnicki, co-founder and chief technology officer at Flobotics. He explains that if RPA is used in a company, employees are more likely to stay as they can do other meaningful work. Subsequently, organisations hire more people after implementing robotics, as RPA improves and streamlines the business faster and results in more revenue, Mielnicki notes.
However, companies must be wise when introducing the notion of robotics. Stafford explains that some employees might fear that if they automate certain tasks, they will lose their job. “It’s the responsibility of the organisation to make the people feel secure,” he adds.
Not a silver bullet
Still, RPA is not the ultimate problem solver. As with any other technology, RPA requires knowledge and education to ensure realistic and achievable expectations, Fox says. “In designing bots, you need to be clear on the business needs and requirements stemming from that, and on the expectations from the business when deployed,” he adds. Indeed, as much as RPA is an IT solution, it is also a people in process strategy, Stafford notes.
Coming up with an RPA idea and developing it is easy, but for it to run sustainably 24/7 in a critical healthcare and clinical trial setting is challenging, Järvi says. Proper maintenance and support structures are needed to ensure RPA business continuity. “People think it’s an easy technology and an easy fix, but such thinking is wrong,” Mielnicki explains.
Also, what RPA lacks is human judgment. For example, if a patient doesn’t report certain events during a clinical trial, a clinician could spot those gaps and abnormalities. “You need the human judgment to hear what people are not saying and the machine has a hard time understanding that,” Killian says. Removing the human factor and judgment from the equation poses a risk to the trial and the patient.
With every technological advance comes the question of access and ability to operate devices. Shwetha R explains that older patients should also be considered in this process as they would need more help using technology if they cannot operate a phone correctly. Additionally, a good internet connection is needed to streamline the data from the patient through RPA to the clinician or the investigator. Clinical Trials Arena has previously reported on tech divide amongst decentralised clinical trial (DCT) participants.
Yet, RPA will play a huge role in the future of clinical trials. The number of trials in progress is continually growing and RPA will be needed for the industry to keep up with the growing trends, Fox says. There is a lot of untapped potential for clinical trials and the industry should learn from healthcare sector, where significant gaps between demand and delivery are filled by robots. “This kind of robot army and its capabilities are available for clinical trials to benefit from,” Stafford notes.
- Robotic Process Automation (RPA) can streamline some clinical trial processes, so they proceed more quickly and accurately. RPA can also improve the experience of study participants, favouring adherence.
- RPA won’t replace humans in the healthcare and clinical trial industries. On the contrary, it can help with employee retention, but organisations need to be cautious when introducing bots into the workflow.
- Realistic and achievable expectations need to be established before implementing RPA. Also, the lack of human judgment can pose potential risks to trial execution.