18 September 2018
Does using digital health in clinical trials generate real-world outcomes?
High-frequency digital health assessments can not only characterize clinically-relevant data as it occurs in real-time, but capture this information within the patient’s home environment. Therefore, digital health platforms provide an ecologically valid data source which directly maps on to an individual’s daily functioning and real-world outcomes.
3. Give meaning to daily activities
What is the applicability of laboratory assessments to patients’ ability to function in the real-world?
Traditionally, in order to assess disease progression or evaluate the efficacy of a novel treatment, psychiatrists will implement a series of tests during each clinic visit (i.e. low frequency testing) which are often pen-and-paper based and require a skilled Rater to score the testing session. Furthermore, the majority of these tests rely on the patient’s ability to subjectively recall certain symptoms and behaviours since the previous visit e.g. How many times has “X” happened in the last month?. Unfortunately, many psychiatric and neurological conditions are characterised by having poor insight into one’s own behaviours, as well as having profound memory impairments, and therefore these patients will not possess the ability to retrospectively describe their own behaviours with a high degree of accuracy (1,2). These particular difficulties in psychiatric assessment have been a long-standing diagnostic concern for clinicians (1) and this problem can be significantly exacerbated when applying a similar reliance upon subjective assessments in a drug trial setting, where there are multiple testing sites, multiple testing time points, and multiple Raters involved.
What can real-world assessments offer clinical trials?
The ideal method to reduce potentially confounding influences upon a critical medical assessment is to conduct experiments within a controlled laboratory environment designed to minimize external environmental stressors and extraneous human social interactions. This is the optimal approach to isolate very specific neural and cognitive metrics, however the core component of many psychopathologies and psychiatric conditions is having abnormal reactions or coping mechanisms to their environmental stressors and human social interactions. Therefore, a controlled laboratory setting might not be the ideal method to assess how a drug can help a patient in the real world. With this notion in mind, having the ability to capture ecologically valid information from psychiatric populations within their day-to-day lives is paramount to accurately characterizing their behaviour and as such, the ideal method to implement a drug trial within these populations. Instead of speculating upon the relevance of a drug’s effect within a laboratory to one’s activities of daily living, this disparity can be reduced by literally testing psychiatric populations within their own environment to assess how it actually effects their daily lives while it is occurring in real-time, with ecological momentary assessments.
What are Ecological Momentary Assessments?
Ecological momentary assessments (EMAs) are a data collection technique designed to capture in-the-moment information from patients within their real-world environments. EMA’s have a long history of use within social science research, but advances in data storage, the ubiquity of devices such as smartphones and smartwatches, as well as the breadth of biological sensors embedded within these devices, has revolutionised the ease of implementing this technique within clinical trials.
What are the benefits of introducing Ecological Momentary Assessments into clinical trials?
By deploying cognitive tasks, functional status assessments and clinical scales on devices participants carry with them throughout their day, researchers can increase the frequency of data sampling and obtain a more granular insight into the patients’ health status both intra-day and longitudinally.
Moreover, prompting participants to complete assessments as the behaviours and symptoms occur within their everyday lives minimizes the temporal bias of retrospective self-reports. Patients can log occurrences as they happen, creating a more accurate data source and thereby providing a more sensitive measure of a drug compound’s effectiveness. Through collecting data in real-world environments, researchers can be more confident that effects observed directly map onto tangible aspects of the patient’s disorder and how it affects their activities of daily living, which could only be speculated upon in a controlled laboratory setting.
Lifestyle interactions can also be ascertained through the additional sensors many of these consumer-facing devices possess, such as heartrate, activity and sleep quality monitoring. These in turn, can become valid clinical endpoints for safety signals, as well as co-variates to explain abnormal performance, such as poor cognitive performance due to disrupted sleep patterns versus drug effects.
The EMA approach to behavioural testing also provides unprecedented access to identifying adherence related efficacy signals. By capturing accurate indications of drug adherence by, researchers can better identify responders versus non-responders and their relationship with drug consumption in a shorter time period than conventional trial designs. Dynamic prompting also offers the possibility of asking individual participants why they may be non-adherent, providing invaluable information relating to side-effects or the ease-of-use for the current drug delivery format in real world daily routines.
Until recently, clinical trials have primarily relied on collecting data from patients during infrequent visits within a clinic. Since the ubiquity of smartphone and wearable devices, remote testing can be used to establish ecologically valid datasets within clinical populations, and reduce the subjective reliance on a patients’ personal insight and memory performance.
Many drug trials are beginning to adopt EMA methodology into their study designs, enabling patients to participate in clinical trials remotely, thus engaging with hard-to-reach and under-represented patient populations. This novel approach also reveals the richness of naturalistic and ecologically valid data collection during patients’ daily routines to accurately assess how therapies affect their lives.
Interested in discovering how you can improve the validity of your clinical trials?
Written by Nathan Cashdollar, Matt Hobbs and Sally Jennings
Nathan Cashdollar PhD and Matt Hobbs MSc