Facebook YouTube Twitter Linkedin
Facebook YouTube Twitter Linkedin


18 September 2018

How can you measure behavior more accurately in your clinical trials?

Drug development typically relies upon clinical endpoints established within very controlled laboratory environments. Digital health technologies now provide the opportunity to transition data collection from the clinic into people’s personal lives, thereby providing more accurate conclusions about how a compound can influence their day-to-day life. 

Four reasons to use digital health in your clinical trials:

1. Improve the signal-to-noise ratio

2. Measure behaviour more accurately

3. Give meaning to daily activities

4. Reduce the treatment burden on patients


Smartphones and wearables are bringing laboratory-level precision to remote testing

Biological and physiological data collection is an invaluable tool for characterizing the underlying clinical deficits within a population, as well as elucidating the site of action for different drug compounds. In the case of neurological or psychiatric disorders, collecting neuroimaging and electrophysiology data within a lab is primarily aimed to isolate very specific cognitive processes. This is achieved by designing experimental manipulations that illicit particular cognitive experiences. However, the extended implementation of these procedures in longitudinal studies is incredibly demanding on the time and resources of both the testing site and the patients.

Many therapeutic areas, such as cancer research, require comprehensive onsite evaluations due to the specialist technology needed to assess the biological processes of interest. But, in the case of psychiatric disorders, many of the cognitive and social processes of interest occur outside of the laboratory.

Many people suffer from psychiatric symptoms triggered by their everyday environment, yet drugs are rarely evaluated for their objective improvement of these day-to-day activities. With the ubiquity of smartphones and wearables, psychological manipulations no longer have to be designed to simulate particular social or mental environmental situations as a proxy for a person’s experience within the real world, because cognitive and physiological signals can now be measured directly within patients’ day-to-day lives.


Why are personalized endpoints so important in psychiatry?

The primary concern for the majority of patients is seeing the benefits of an intervention in their everyday lives, improving their functional status, such as their ability to undertake day-to-day tasks, focus at work or remember a conversation.

Recent innovations in the field of Digital Health have created new methods of synthesizing physiological data collected from wearables and smartphones. This technology can passively detect meaningful measures of mental health and cognition that not only minimize the need for cumbersome physiological tests to be conducted in a laboratory, but also directly relate to the patient’s functioning within their home environment.

Multiple pilot and proof-of-concept studies have obtained promising results demonstrating how passive sensor data from wearables and smartphone devices can provide non-invasive measures, such as step-count or sleeping patterns (1, 2), that can be sensitive measures of a patient’s clinical symptoms in the real-world.

For example, a recent study of patients with schizophrenia demonstrated that a combination of passive and active data collected from a smartphone application could detect patterns of behavioural abnormalities (GPS mobility and social behaviours) that significantly predicted relapse, two weeks prior to occurrence (1,2). Whilst preliminary, this data demonstrates the utility of digital phenotyping methods for revealing clinically relevant changes in symptoms (2), and using these signals for predicting potentially dangerous behaviours outside of the laboratory.


Digital health will revolutionize drug development in psychiatry

Many psychiatric populations have a lack of introspection into their own behaviours and mood, thereby making questionnaire-based measures of functional status potentially troublesome. Recent studies have demonstrated that by passively collecting GPS data and activity patterns (e.g. location entropy) these measures can be used to objectively assess clinically important changes in patients’ ability to function within their own environments (3,4).

For example, GPS trajectories studying how frequently participants engaged in leisure activities, social situations or staying home, were related to the clinical severity of anxiety symptoms (3). New insights such as this, offer opportunities to create functional endpoints that directly map onto human behaviours and clinical symptoms that can provide important insights and meaning to both the patient and clinician (5).  

The transition from retrospective self-reporting of functional status to the analysis of actual human behaviour brings objectivity to an otherwise subjective tradition of psychiatric assessment. Furthermore, data collection occurring within a patient’s home environment reveals sensitive EMA relationships of one’s daily behaviour that could only be speculated upon within a laboratory setting.

Interested in how digital health can improve the objectivity of your clinical trials?

Watch our webinar


  1. https://www.nature.com/articles/s41386-018-0030-z
  2. https://www.nature.com/articles/s41746-018-0022-8
  3. https://www.ncbi.nlm.nih.gov/pubmed/29973337
  4. https://www.ncbi.nlm.nih.gov/pubmed/28258049
  5. https://www.nature.com/articles/s41746-017-0006-0

Written by Nathan Cashdollar, Matt Hobbs and Sally Jennings 

Tags : digital health | clinical trials | wearables | smartphones | precision psychiatry | personalised medicine | digital tools | high frequency | near-patient

Author portrait

Nathan Cashdollar PhD and Matt Hobbs MSc