Judith Kornfeld is the chief business and operations officer for ORCATECH, an R&D center formed in combination with the biomedical engineering and neurology departments at Oregon Health and Science University.
Her group has developed a system for clinical research based on remote sensing and pervasive computing whereby data is collected unobtrusively and continuously. ORCATECH installs sensors in patient homes, rather than outfitting patients with wearable devices, to monitor patient activity.
We chatted with Judith at the Partnerships In Clinical Trials conference in Hamburg, Germany, to discuss how technology is transforming clinical development.
Tell us about ORCATECH’s technology.
ORCATECH technology was developed to allow for the patient’s daily life be the source of data to measure health changes and meaningful outcomes in clinical research.
We have shown that, in order to detect meaningful change as an outcome of a therapy, you really need to look at patient data in a continuous manner, not in an episodic manner as it’s done today, and rely on objective data that reflect the patient’s real life. In essence, the approach to data collection in clinical trials has not changed since such trials were first conducted over 250 years ago, while medical technologies for treatment and drug development have advanced incredibly.
Moreover, this approach to clinical trials requires a smaller sample size per clinical trial by an order of magnitude, and the ability to detect change and response to therapy is much faster because of the density and the plurality of the data.
What are some of the types of data you’re collecting?
The main types of data we pick up are the movement around the house, leaving and entering the house, walking speed, sleep measurements, computer use, balance, weight, medication use, socialization and more. We also collect weekly reports via e-mail that allow us to obtain additional data that was not picked up through the automated system. Recently, we have started to look at driving parameters as well.
This is an open platform. We can deploy any sensor that collects data that is of interest to the questions asked in a clinical trial.
And this is dense data as you say?
Yes, because it’s continuous and mostly collected 24/7.
There’s a potential for too much data, right? What are the data considerations then?
This is a very important question. There is never really “too much data,” it’s all about making sense of this gold mine. In a way, that’s the secret sauce of ORCATECH. It’s the algorithms developed here that take all this data and interpret it into valuable clinical information.
These are digital biomarkers?
Exactly. These are digital biomarkers that can be obtained in a very accurate manner that reflect the patient’s parameters of real life. We measure what you want to improve. With medical treatments you want to improve patient’s lives. We are able to measure those parameters.
Digital biomarkers are a fairly new concept. What are some of the considerations for adopting these within the pharma industry?
You have to think about compliance. If you’re using a device, have you considered how it’s actually being used, how accurate it is and how it reflects what you’re trying to measure?
We emphasize multi-source data to get a real, true-to-life picture. We also make sure before we collect the data that we know exactly how patients will interact with these sensors. We conduct focus groups to see how patients will use it and how it relates to their real life.
The whole idea is to collect something that is ecologically valid, that the patient is not doing anything in a manner they wouldn’t do everyday. It is very important to not use a single device and follow one data stream that is isolated from reality. You need something that makes sense and gives a multidimensional picture.
We emphasize being less dependent on patient compliance, so that the data is whole and accurate. In our system, the only part we rely on subject compliance for is a weekly email questionnaire. This helps us get information that we could otherwise not get, and it puts the data we collect in context.
Even filling out this questionnaire – the way it’s done, the time it takes, the regularity – tells us a lot about the health and cognitive situation of a patient. It’s not only the questionnaire that’s the target, but how it’s filled out.
Where do you see drug development evolving with new technology like this?
We see that there is a bounty of compounds out there in drug pipelines, but very few get approved. The clinical trial process is very costly, long and not very efficient. This has to change.
There is a drive to reduce costs of healthcare, and therefore we must reduce the cost of R&D for treatments, both pharmaceuticals and medical devices. The cost of conducting clinical trials has to come down, and the way to bring it down and develop treatments more efficiently is to improve where we pick up data and how we pick up data to shorten the trial timelines and reduce the number of patients in trials.
It’s not just a matter of using a device in a trial in a sporadic manner without context. The whole design of the clinical trial has to change.
What will trials look like if everything goes well?
They will be conducted very differently than today. The emphasis of the protocols would be to collect continuous and objective data, less reliant on sporadic and often inaccurate patient reporting, or clinician’s sporadic observations sometimes uncorrelated to real-life testing. The whole process will be much more accurate and true to the source, and that’s where we hope things will go.
New technology can actually reduce costs?
New technologies may reduce costs in many ways, but this need to be measured by the cost of the technology, the cost of implementation and the increased efficiency of the clinical trial process.
With the ORCATECH technology we aim at all fronts. First, the sensors we use are off-the-shelf sensors that are very inexpensive. We do not develop any special devices, which have their own development costs. The fact that the data is collected in the home and not in an expensive clinical setting reduces costs.
But the real cost advantage happens in the long term, because smaller sample size and shortened clinical trials are where the costs savings will be most significant.
Time in a trial is very expensive, so if you can shorten the time to detect change or reach an endpoint, you reduce the cost enormously. There’s no aspect of our process that’s more costly than current clinical trials.
What are the concerns you hear about with this type of technology?
The number one concern we hear is about security and privacy, but none of the data we collect can identify a person. We do not use any cameras or facial recognition, and all the sensor data that is collected in the home is separated from the identity of the data creator (the subject) so when it leaves the home it’s already unidentified.
We use standards of data safety and security that are probably higher than what’s currently used in financial systems that we commonly use in a remote manner.
All our lives today are on the Internet, so the idea that data collected at home is more of a privacy breach than other methods is no longer valid. We comply with all HIPAA standards and we use IRBs in all our research.
This article originally appeared on our Forbes channel.