Digital Technologies to Measure Function in Older Adults with Multimorbidities

Join Dr. Sarah Berry and Junhong Zhou, PhD, from Harvard Medical School, as they review the characteristics of novel digital tools to measure functions in older adults with multimorbidity, while underscoring both the advantages and challenges of using these digital technologies in measurement.

To view a transcript click here then select the transcript tab. 

Transcript

Sarah Berry, MD, MPH: Hi, my name is Sarah Berry. I'm Associate Professor in Medicine at Harvard Medical School and the Marcus Institute at Hebrew Senior Life. And I'm here today to talk about the AGS/ AGING LEARNING Collaborative. Specifically, we're gonna talk about Data Measures and Measurements: Digital Technologies. I'm delighted today to be joined by Dr. Junhong Zhou, and I'm gonna refer to him as Jay-Z. That's how I know him. He's an instructor in medicine at Harvard Medical School and a research scientist too here at the Marcus Institute. Great to have you here today, Jay-Z. 

Junhong Zhou, PhD: Hi, how are you Sarah? 

Sarah Berry, MD, MPH: Yeah, I'm doing well. Thanks. So I brought you here today to talk about digital technologies and, and first I was hoping you could just tell our audience like what are some of the new digital tools available to measure function in in older adults?

Junhong Zhou, PhD: Actually, [00:01:00] currently, there are three main categories for the digital tools that are available to the older. Like I will introduce or one by one. Now the first is about the digital interface using like smartphone or tablet, like the iPad or the Android tablet. Cause we know traditionally, like many clinical tests are completed using the paper or pen, like the questionnaires or some like the very simple tests.

And we always ask patients or other adults to write down their subjective answers or responses to questions, and then there may be some potential subjective bias or even errors happening, right, when transcribing the results into like the clinical use or into the research. 

Sarah Berry, MD, MPH: Yeah, I've had that problem before.

Junhong Zhou, PhD: Yes. But now there are many like applications on smart phones or tablets, right, allowing the users to answer those questions digitally. So we can like just write down [00:02:00] the answers in the, on our smartphone or tablet, and the application will collect the data automatically and accurately. Right there, there will be no error.

What... is written down will be the actual response from the users. So this can save, first, save a lot of the human resources. Cause maybe we can have one study personnel, for example, to administrate the procedure. We don't need to have [an] additional person to do the test, actually, right. The users, the other adults, can do the test by themselves and also, of course improve the validity of the research. 

So the second type will be the active wearable sensor. For example, like in the assessment of mobility or even walking, traditionally we use the stopwatch to count the time that the participant needs to complete the test, right? It's really rough. And now we have some like small -sized [00:03:00] wearable sensor systems that [consist of] the accelerometer or gyroscope. Then can like a very [specifically] measure the motion pattern. And to obtain the outcomes of mobility, like the gait performance, for example, the gait variability, get strike time, those very detailed outcomes instead of the just average walking speed.

So this can like boost our research to look more into the insights of the, like the functionalities of the participants and their performance of the test. 

And the third type will be the passive digital systems. So we call it passive, is due to [the way] they can continuously track or monitor the functions without significantly interfering with the activities of other end.

One example is an intelligent home platform, like integrating all the different kinds of the sensors, like the [00:04:00] sensors to measure the temperature, measure the night, measure the physical activities of the residents when they were, they are at home. So these systems can require the other end of information passively and they don't need the users or the residents [to] actively use it. [It] can continuously measure their functions and other things like the, even the environmental characteristics, right? Cause we know maybe those kinds of characteristics also influence our health. So this kind of system allows the recording of the personal activities, which is also very important to be measured in other end. But by in traditionally we cannot do that because we always ask participants or patients to, to actively do the tests, right? So it's not a continuous monitoring or track, but this kind of passive systems can do this kind [00:05:00] of continuous reports. 

Sarah Berry, MD, MPH: Wow, there's a lot of new technologies out there. Thanks for, for summarizing those. 

What sort of outcomes can be measured using these, these digital tools? You started to talk about some. Are, are these patient centered outcomes or what, what sort of outcomes are we talking about?

Junhong Zhou, PhD: That's a good question. So actually, compared to those traditionally measured like patient centered outcomes, this kind of new digital tools can provide accesses, high frequency and continuous measurement of the functions and all the measurements that allow all, like most of the measurements can be completed at the same time.

The outcomes of the functions are more detailed cause we can really look into many subtle aspects of one function that may be lowered in traditional patient-centered outcomes. Still, like for example, for the working on mobility, like we displayed or showed in our slides, we can [00:06:00] easily obtain much more robots than the or beyond the traditionally measures.

Like patient centered outcomes like the turning time, right? And step length and the minimum time than the foot needs to leave the ground which is really a, a sensitive marker for patients with Parkinson's disease and many other things. Because traditionally we always measure some- something like, it's like an average across entire trial, entire test. But for now, we can even separate, for example, if we do a walking test, but we have the straight walking period and also due to the environment, right? Sometimes the participants will do turns and then do other motions. Then for now, we can separate those different spaces or periods, and then to graph the information from those periods separately, and for our purpose of research or for our purpose of the clinical practice.

[00:07:00] So this kind of assessment, yeah. Can provide more insights into the, for example, the aging process for our research or the, even the pathology of the age related conditions. And also in on other hand for research, right? So maybe we can have more like freedoms or flexibility to design our study by selecting the outcomes that will really lead to me, right? Because sometimes for example, for a clinical test or questionnaire, we all, we have to do all the things, right? To get the outcomes maybe will cost like half an hour or even one hour or more like time to do. So half an hour. With this digital tools, we can select what kind of information we, we wanna extract. And then to me it's very objectively and very even passively, right? Without interfering with the patient. 

Sarah Berry, MD, MPH: So you've spoken about some of the advantages with these digital [00:08:00] technologies, right? That we can avoid bias, we can get a lot more detailed information, and that we can cut down on sort of the users or the people that have to measure these outcomes. It's more efficient. But what are some of the challenges in using these, these digital technologies? I'm guessing with older adults, there may be some barriers. 

Junhong Zhou, PhD: Yes. This is a good question. Cause yeah, we talk a lot about the advantages or like, The use of this kind of thing, but actually it's a new field. It's a brand new field with the technology of the computer science, right? So there are a lot of, like, still a lot of challenges or like outlooks in this kind of technologies. First is how we can understand or interpret the results, right? We can have numerous data or numerous outcomes to be measured, right. Tons of data, like a, even like a big data science. But, but the, for our clinical research or [00:09:00] clinical practice, the most important thing is what the use or what's the significance or importance of these outcomes can be, right? Which, so the analysis and the interpretation of the data, all of the outcomes are really important.

We need to tease out the irrelevant signals and tease out the signals with poor data quality. And then we need to think about how we can integrate the data from different kinds of sensors. Right? We, we should think about more about the- for example, the interaction between the older adult's behavior or, and their environment, right? That, that will raise a lot of new questions for the interpretation of the results. 

And also sometimes, because we wanna use it in different scenarios and, or even in, like, for example, order own home or in, instead of waiting in the line, [00:10:00] right? Or waiting the clinics. So sometimes the data will be lack of validation. For example, we record the information from a senior living home and the data will be automatically uploaded to the cloud-based server and then we can only have the access to the data. The first question is how we ensure the validity of the data. That will be a question. So that will require, or that- that leads like the more real designed proctor for the data recording and data transportation and then the data processing. So that will be another question. 

And then the next one I feel is also important one is the privacy of the security of the data, right? Because if we still, if we record activities in the senior living home or senior living house, right? We will of course record [00:11:00] all kinds of the activities from the rooms.

So sometimes there will be issues with privacy. So we need to be very careful about this to have a very healthy, well designed and also have a data safety monitoring team to manage or to administrate the process to ensure there will be no leak of the privacy or leak of the personal information, sensitive information, to other things that's also important.

And also sometimes, and especially for now, meaning like this kind of technology are really expensive, like the sensors, if we want to [use the] whole platform, sometimes we need to, it costs too much. So the users may not choose this kind of method due to their financial conditions. 

And also they need some like, knowledge to use, right? You need to, how to interact with for example, [00:12:00] the application or how to use Zoom on a smartphone. Which I don't think that's a problem for now, but maybe for some new technology like the sensors, they will, they will need the users to be very familiar with it. Then they can successfully use it.

And also maybe some old fashioned like living style people, they are not interested in the new technology, so they may not prefer to using this kind of thing. 

Sarah Berry, MD, MPH: I can imagine. I appreciate you outlining those. It's always a trade off between the, the advantages and the disadvantages when we're selecting how to measure our outcomes, but there really does seem to be some promise in these new technologies.

I wondered if you would just comment briefly on, you know, are any of these technologies, are there sort of ways to measure function across multiple systems at the same time? That- that seems to be [00:13:00] efficient or maybe a benefit of these devices. 

Junhong Zhou, PhD: Yeah, that, for this question, so yeah, this is also an important feature for the new digital tools. As we know traditionally, we oftentimes measure folks, one or only a few aspects of the functions of one individual. However, we know that even one type of functional decline may result from the dysfunction from multiple underlying biophysiological systems, right? 

For example, as we showed in our slides. So the walking speed, which is I think one of the most common symptoms in other end of it, may be resulted from, for example, cardio impairment, poor muscle strengths, or even altered blood pressure. Right? So actually there are multiple conditions [that] can contribute to one symptom or one functional decline. 

So in traditional fashion of the measurements, it's challenging to [00:14:00] measure all these things at the same time, right? We can only do it one by one, but sometimes we don't know what are happening at the same time. So we may not [be] able to catch the point. Maybe that's the, the issue is coming from the optimalities of the interactions between the system instead of the problems within one or two systems, right? But with the help of the digital tools, right, we can be much easier to capture those functions simultaneously.

So this one allows us to obtain them more like a full picture of the functions or the impairments of one individual. And especially if the person has multiple comorbidities, and then the information we obtain, like the simultaneously measured characteristics across different systems or across the person, he or she self or even the environment will [00:15:00] help the health management and back of the diseases and conditions or it may help the prescription of them.

So that's the necessity or the importance of the simultaneous measurement of the functionalities across multiple systems using the digital tools. 

Sarah Berry, MD, MPH: That's exciting. Jay-Z thanks for sharing. I, I certainly took away, there are some challenges and I would need some help in using these digital technologies in research, but the ability to cut down on bias, to measure things efficiently and to measure, as you saying multiple things simultaneously really shows great promise in the future.

So thank you for your time today and for talking about this with our listeners. 

Junhong Zhou, PhD: Thanks. It's really exciting and I'm more than happy to do this.