Listen to our scientific advisor, Raeanne C. Moore, Ph.D., giving a presentation on Ecological Momentary Assessment at the November 2022 RCCN Workshop on mHealth and Digital Health Approaches to Research in Aging.
Transcript
Speaker 1:
Next speaker is going to join us virtually. It's Dr. Raeanne Moore.
Dr. Raeanne Moore:
Great. Thank you for having me virtually and thank you, Elizabeth, for accommodating. I'm sorry I can't be there with you all today, but happy to be participating from my home. And just a quick introduction as I get my PowerPoint set up here. My name is Raeanne Moore. I'm an associate professor in the department of Psychiatry at the University of California San Diego. And I do a lot of work with mobile health tools to improve assessment of mental health and cognition. I'm the co-founder of the Mental Health Technology Center, and I co-direct the Cognitive Dynamics lab here at UCSD, and I have my geriatric dog just at my feet at the moment snoring away. So hopefully she's not too loud. I feel like my talk should have gone before Marty's and kind of set him up for that. But I'm going to be talking about EMA and I'm going to be using an example from a study we've done with middle age and older adults with schizophrenia.
So I have a few disclosures, I've cofounded a couple companies as have funding as well from NIH. So disclosures. Okay, so what is EMA? I mean, I think everyone in the room already knows pretty much what it is, but EMA really is a way that can limit bias in self-report data because it's just asking people kind of as they go about their daily lives, about their thoughts and behavior. We're repeatedly collecting data in an individual's normal environment, really in as close in time as they carry out that behavior. So it can eliminate some of the biases with our traditional self-assessment tools that our biased and rely on retrospective recall, priming current psychological state by just asking people, "Right now how are you feeling? What are you doing? Who are you with?" A lot of these other interesting questions about daily life.
And to give you example of an EMA protocol, this is an example of a protocol from an observational study. So we do a lot of observational studies as well as using EMA and RCTs, but for an observational study protocol, for example, we give people text notifications three times a day for, say, 30 days, giving them EMA surveys that just ask about their mood, daily functioning and symptoms. And as Marty was saying, we design our EMA surveys so that regardless of how people respond and then the subsequent branching questions, we make sure no matter how they respond, they get the same number of questions so that we get responding as truthful and not just trying to quickly get through the survey. People can do this on their own phone or on a study phone.
And in our protocols we do offer compensation just to encourage adherence, and we typically have adherence rates higher than 80%. So we have really great adherence to our protocols. We also have staff do checking calls on the first day as well as the third day just to make sure that text notifications are being received and that people, if they have any questions with the technology. In addition, we implement a daily status tracking for participants so that our staff can monitor just very visually friendly, quickly in real time when people are completing their surveys and if there's patterns of missed surveys. And then if people missed three surveys or more in a row, the staff contact them to do some motivational interviewing or troubleshooting around what's going on.
So why do we use EMA for neuropsychiatric conditions? Patients with serious mental illness are reportedly lacking awareness of their symptoms. But is it really a lack of awareness or is it just that our traditional measures do have this bias, especially in terms of retrospective recall, current psychological state influencing reporting? And some things are really hard to remember. Traditional methods, they see their psychiatrist maybe once a year or every six months, ask them how their mood has been over the past period since their last visit. Symptoms really fluctuate and it can be really hard to average how you've been feeling over time.
So I'm going to give an example of a study that was published by one of my graduate students, Emma Parrish, where we looked at EMA to measure a functional disability in middle-aged and older adults with schizophrenia. Because the biggest problem we see in people with schizophrenia is not the positive symptoms or suicidal behaviors, but it's really this functional disability in the domains of social, vocational, and residential disability. And it's the number one complaint that people with schizophrenia have as they're presenting problem. So in this paper, we combined EMA with GPS as a way to see how people were functioning in their everyday life. If symptoms and mood are different when people are at home and alone versus home with others or away from others, kind of what happens when people are leaving their home. And we used the GPS as also a way to just validate the self-report measures.
So when you look at the movement patterns, the orange bars here are the people with schizophrenia. This was a sample of 105 participants with schizophrenia, 76 control participants. And we see that people with schizophrenia were more likely to be at home on a greater percentage of the surveys. They were also interestingly more likely to leave home, but they stayed out a much shorter period of time and we're more likely to return to their home more quickly. So thinking about maybe they're just leaving home to walk down the street to the 7-Eleven, grab a soda, come back home, but they're not staying out for very long. And what's happening when they're out? the orange and yellow lines are the people with schizophrenia. Orange represents anxiety and yellow sadness. And then the blue and green lines here we have the controls. Green is anxious and blue is sadness.
And overall the controls had significantly less anxiety and sadness over this 30-day EMA period. And we see interestingly that the participants with schizophrenia had the lowest negative moods when they were at home. And then there's an increase in anxiety and sadness when they leave the home and the time that they are out of the home. Opposite pattern for positive moods in that their positive moods dip when they're leaving the home and then when they're out of the home for people with schizophrenia. This was particularly salient for going to a clinical appointment. So we saw that going to a clinical appointment had a significant inquiries on sadness and anxiety in participants with schizophrenia. So thinking of the clinical implications for that and wanting to improve adherence to clinical treatment interventions, real-time mobile interventions could be done around that.
And so I know we were short on time because I was the last presenter before the break, so I went through these really quickly and am going to just end with some gaps in the field and opportunities for research, which have already been touched on by the speakers, but I'm part of the International Society of Clinical Trials Methodology working groups, and recently this society's working group determined that EMA is a mobile health technology that is determined to be ready for primetime and use in clinical trials, especially for measuring negative symptoms, whereas some of the other technologies aren't quite there yet, but there's still a need for improvements. And I think particularly around the need for streamlining data analysis, and this applies to all of the technologies that we're talking about today, but we get so much rich comprehensive data, and there's a lot of backend cleaning and pre-processing that needs to be done prior to analyses. And so automating some of these processes when we think about moving these technologies more into clinical trials and clinical practices is really important.
But there is such a need to develop tools and implement these tools to collect real-world data in clinical trials and conduct decentralized clinical trials. And most of the studies I do are not solely using EMA, but using EMA in combination with the mobile cognitive tests, as Jason was talking about, as well as GPS and some of the sensor-based data we can get from other mobile health tools and digital health tools. But there is a need, if we are moving to this, there's still really no best practice guidelines and particularly around data sharing. And so there's a need to develop best practice guidelines, especially for thinking about rigor and reproducibility in this work. And if different groups are going to be collecting things in different ways, how are we going to improve best practice guidelines around that?
So I'm going to end there, and even though I'm sad I can't be there, I'm grateful to be mentoring Laura who is in the room and you'll be hearing more about our work after the break. So stay tuned. That's all I have.
Speaker 1:
Thank you so much.
