Connected devices: a tsunami of clinical trial data

I

nternet of Things (IoT) is the catchall term covering devices that collect and exchange data with other devices over the internet. And, according to experts involved in their rollout in clinical trials, the sector is at a tipping point.

In the past three to four years, the sector has experimented with new forms of data capture, says Medidata digital biomarkers solution expert Ben Schlatka. But now, it is at a phase of frequent and high adoption with structured research questions, Schlatka notes.

In clinical trials, the most common IoT devices are electronic Clinical Outcome Assessments (eCOA) or Patient Report Assessments (ePRO), sensors, and wearable devices. The areas with the most clinical trial IoT uptake are cardiovascular diseases, central nervous system disorders, chronic diseases, dermatology, diabetes, and pain management, notes Gabriele Brambilla, CEO of healthcare technology company Alira Health.

While IoT integration in trials is being accelerated, several issues remain. With devices collecting an unprecedented volume of data, are sponsors or contract research organisations (CROs) equipped to use that data to bolster a trial’s success prospects? Despite eCOA and ePROs being integrated into clinical trials for years, decentralisation rollout issues exist. As for sensors and wearable devices, with many options available in the market, how should sponsors choose the right approach?

There are two types of IoT data collected in clinical trials: the first is data gathered as per the protocol, while the second involves information that can be used to help design other trials in the future, Brambilla says. For example, the data can help target the right patients for a future trial’s success, he explains.

From a different angle, Schlatka says the first type of IoT data is that which one would previously collect at a trial site. Now, as a decentralised form of data collection for the same data points, data gathering happens more frequently. The second type is new data informative of the participant’s unstructured environment, which puts treatment data in a new context, he adds.

“The clinical trials industry is still getting to grips with the data we have,” says Gillian Livock, Medable senior vice president and general manager, connected sensors and digital measures. It is imperative for the sector to use the data thoughtfully because participants put in the effort to report the data, she adds.

IoT technology helps collect data that can answer primary endpoint questions, says Matt Bonam, AstraZeneca head of digital health, biopharmaceuticals research and development. IoT data can also provide exploratory data that can be used retrospectively to find biomarkers to develop a pipeline, Livock adds.

But a fishing expedition should be avoided. It is not optimal to collect all types of data and then find a meaningful trend, Bonam cautions. The scientific process should be protected, by starting with a clear hypothesis and data collected in a structured way, he adds.

There are exceptions. For example, in using data to predict the onset of heart failure, the critical parameters in cardiac function are known, but which of these would be more or less important is unknown, Bonam explains.

Patients should be informed why data is being collected and, if there is an exploratory element, it must be explained thoroughly, Bonam says. When establishing what data to collect, it is important to find measures that matter to the patient, Livock adds.

One of the more mainstream IoT approaches in clinical trials is the use of eCOA or ePROs, says Bryan Lubel, executive vice president connected health at IoT provider KORE Wireless. According to Clinical Trial Arena’s exclusive decentralised clinical trial (DCT) adoption tracker, eCOA and ePRO use sharply increased from 2020 to 2021, largely driven by more widespread use among CROs. In clinical trials, smartphone or tablet devices are locked down to be a single-purpose managed device, Lubel notes.

While ePRO and eCOA have been used for years, what’s currently different is how they are used. Previously, these approaches, while electronic, were utilised at clinical trial sites and therefore not decentralising the study, Brambilla explains. Now, these approaches are being used in a patient’s home, he adds.

There are several issues sponsors need to manage. One is the cost and time it takes to invest in such devices that won’t impact the drug’s development timeline, Lubel says. There are also questions regarding the logistics of delivering the device to the patient, and how quickly collected data can be validated, he adds.

To integrate eCOAs or ePROs in clinical trials, companies either purchase other companies that already have this technology, or develops them internally from scratch, Livock says. The latter’s advantage is that the technology is synergistic across any of the decentralised trial modules already being used by the company, she adds. Users won’t have to switch applications if the trial have different decentralisation modules.

Patient compliance is another issue; how often patients enter the required data, and if it is the patient doing the data entry, Bonam notes. There are simple identification methods like fingerprint and face recognition technologies that help alleviate this concern, he adds. Patients need to be informed why and how the data is used to ensure high engagement.

Meanwhile, sensors and wearables are a hot button topic in the clinical trials sector because, until now, they were not taken seriously, Lubel says. “It was considered a science project.”

There is a new appreciation that sensors provide a much broader and deeper insight into an investigational approach’s efficacy, Lubel adds. According to our DCT Adoption Tracker, the use of remote monitoring using sensors and devices increased from an average of 60 trials per year between 2010 and 2016, jumping to more than 250 in 2021.

The challenge faced by sponsors is the wealth of device options available. And all these devices have their own roadmap, architecture, and Bluetooth configurations, among others, Livock notes.

Bonam says AstraZeneca uses only gold standard devices, which means the approach already has regulatory backing. And, if a device is to be used outside its intended use, there is a validation process that compares it to a gold standard, he notes. AstraZeneca has a device catalogue featuring tested and approved devices, as well as those that do not meet standards, so it’s easier to find the right device when designing a trial.

What’s fascinating about IoT devices is that the proof-of-concept on their utility is typically collected in the real world, before they are integrated into clinical trials, Brambilla explains. And so, the more these devices are being used in practice, the more likely they will also be used in studies, he adds.

The rollout of sensors and wearables in trials can be geographically dependent. For instance, there can be two pulse oximeters with identical features, but one is approved in the US and another in the EU, Lubel adds. A trial can have data from multiple types of oximeters, and each software can only capture or display data in a specific way, thus needing integration work, he notes. Data privacy also varies according to geography, he says, noting that, for example, data collected in certain countries must stay there.

To bolster participant engagement, the sensor or device needs to fit discreetly into daily life, Livock says. For example, expensive devices may not be ideal in developing countries as they can draw attention to the user, and thus alternative solutions are needed. She adds, ultimately, while choosing the sensor or wearable choice, sponsors should ask this fundamental question: “Why are we collecting the data and why does it matter?”

When it comes to IoT, does the size of the clinical trial sponsor make a difference? Large pharma and MedTech firms might be more open to exploring new ways of collecting data via a remote approach, Brambilla says. Smaller companies, meanwhile, have one shot at trial success and so are more risk-averse, he adds.

Bonam differs: while large pharma companies may have an edge in being able to push for innovation, all sponsors have the fundamental goal of reaching a greater proportion of society. This is to everyone, particularly the patient’s, benefit, as participants who are highly engaged in trials do better overall, he adds. “Technology is there to make trials more accessible.”