THE EVOLVING LANDSCAPE OF DIGITAL TOOLS TO ASSESS PATIENTS WITH SCHIZOPHRENIA
In this section
The idea of using digital technologies in helping to assess and monitor patients with schizophrenia more naturalistically has been an active area of research for many years. During the ‘Managing early onset schizophrenia in the digital age’ symposium held at the 34th European Congress of Neuropsychopharmacology, a snapshot of the evolving landscape of digital tools being used or investigated was provided.
During his presentation, Martien Kas, Professor of Behavioral Neuroscience of the Groningen Institute for Evolutionary Life Sciences at the University of Groningen in the Netherlands, discussed how digital tools such as smartphones, smartwatches, and other wearable healthcare devices are capable of not only measuring patients’ physical movements, but also capturing data on their mental health and general well-being. Digital phenotyping is a rapidly emerging field, he observed, which is evidenced by the sheer number of papers published on the topic in the past 5 years alone.
Pros and potential of digital monitoring
Keeping track of patients was a real challenge during the COVID-19 pandemic, Prof. Kas noted. The use of digital technology is therefore perhaps even more topical now than it has ever been, he suggested, as it has the potential to provide regular, longitudinal measurements of patients’ activities in a real-world setting.
As an example, Prof Kas showed data collected continuously over a 42-day period from a single patient’s smartphone. This showed the daily rhythm of their phone use, with periods of higher, lower and no usage. This is a very basic way to look at a patient’s behaviour, just by how often they are opening and interacting with their phone, he said.
Digital technology has much more potential than that of course, Prof. Kas added. Could there be a way of using the data collected as an early warning system that the patient is about to relapse, or perhaps monitor the effects of an early intervention?
There is some preliminary evidence that the passive collection of data via a smartphone could be used to identify individuals who may be about to relapse. Indeed, one study by Henson et al,1 found anomalies in a patient’s usual sleep duration and mobility data shortly before a clinical relapse was confirmed. This is just one example illustrating where the technology could potentially go, Prof. Kas said.
Digital phenotyping is still a very new field and there are still many challenges ahead, Prof. Kas said. The first challenge is determining what is being measured, when it should be measured, and if it is even clinically relevant? Second, there will be a need of technical validation and then third, regulatory approval.
In reference to what exactly is being measured, Prof. Kas noted that schizophrenia is thought to follow a neurodevelopmental path with early brain changes occurring long before overt psychosis occurs. If that is the case, then perhaps those early brain changes may be linked to changes in behaviour that could be picked up earlier using digital monitoring.
Social withdrawal, for example, is something that could potentially be measured using digital technology and is a potential proxy for later psychosis. It’s not unique to patients with schizophrenia, Prof. Kas said, but increasing social withdrawal could be a clear sign of deteriorating mental health. Indeed, data have shown that social withdrawal may precede the onset of psychosis by almost a decade.2
How can digital tools help?
Current means of measuring social functioning or withdrawal include the WHO Disability Assessment Schedule (WHODAS) 2.0. This is a generic tool for measuring a person’s health and disability and ask people to rate how they feel they have got along with others or participated in social activities in the past 30 days. That can be quite a task for someone with schizophrenia, Prof. Kas suggested. Not only that, but patients’ responses often do not match up to researchers’ ratings of them, with patients overestimating their social functioning.3 This highlights the need for more objective assessment, which digital tools may be able to provide, said Prof. Kas.
Prof. Kas gave a few examples of how digital monitoring had been used to assess patients with schizophrenia. One of these used a smartphone to actively ask the patient questions such as ‘where are you now?’ and ‘what are you doing?’
Another example involved the passive collection of data from a smartphone app that had been developed by Prof. Kas’ research group.4 Once downloaded the app monitors the smartphone users’ call history, SMS messaging history, Wi-Fi and location data and app usage. Of course, the actual content or context of calls or messages is not monitored to ensure the users’ privacy is not invaded and everything complies with general data protection regulation (GDPR)5. This provides a lot of data without the patient having to be active in the process, Prof. Kas said.
Once all these data have been collected, what can you do with it? One of the ways it can be used is to see many unique places an individual visits or how much time they spend at home.4 They you can compare this to control groups of people who do not have schizophrenia.
Paving the way to regulatory approval
One of the big challenges ahead will be technical validation; there are many types of smartphones and how people use smartphones differs by age. Accuracy studies are needed for the tools that have currently been developed and Prof Kas noted that his team had done that for their app. Their work has been submitted for publication and the data looked “pretty good”.
Regulatory approval is then the next step and a recent paper6 has set out three criteria to help digital health technologies to pass EU regulatory approval: 1) verify the accuracy, precision, and reliability of the tool, 2) validate that it does what you think it does, and then 3) assess it in a clinical trial.
Concluding, Prof. Kas observed that digital phenotyping can help monitor patients over time in a real-world setting and that, potentially, it could be used in clinical trials to monitor patient outcomes and help implement patient-centred treatment approaches. There is still lots of work to be done, but there is a lot of promise ahead.
- Henson P, D’Mello R, Vaidyam A, et al. Anomaly detection to predict relapse risk in schizophrenia. Transl Psychiatry. 2021;11(1):28.
- Cullen K, Guimaraes A, Wozniak J, et al. Trajectories of social withdrawal and cognitive decline in the schizophrenia prodrome. Clin Schizophre Relat Psychoses. 2018;4(4):229–38.
- Jongs N, Penninx B, Arango C, et al. Effect of disease related biases on the subjective assessment of social functioning in Alzheimer’s disease and schizophrenia patients. J Psychiatr Res. 2020;S0022-3956(20);31072–4.
- Jagesar RR, Vorstman JA, Kas MJ. Requirements and operational guidelines for secure and sustainable digital phenotyping: design and development study. J Med Internet Res. 2021;23(4):e20996.
- Mulder T, Jagesar RR, Klingenberg AM, et al. New European privacy regulation: Assessing the impact for digital medicine innovations. Eur Psychiatry. 2018:54:57¬–8.
- Mantua V, Arango C, Balabanov P, Butlen-Ducuing F. Digital health technologies in clinical trials for central nervous system drugs: an EU regulatory perspective. Nat Rev Drug Discov. 2021;20(2):83–4