University at Buffalo, 2022 New Investigators Research Grant
Effects of Speaker and Listener Characteristics on the Minimally Detectable Change of Speech Intelligibility: Dysarthria Etiology, Speech Severity, and Listener Experience
Tell us about this study.
Broadly, I’m interested in patients with dysarthria, a motor speech disorder that affects a person’s ability to produce speech that is intelligible or understandable to a listener. I’m interested in how intelligibility changes over time with disease progression, as well as how speech rehabilitation can improve intelligibility.
This study included people with amyotrophic lateral sclerosis (ALS), Parkinson’s disease, multiple sclerosis, and Huntington’s disease. These neurodegenerative diseases affect speech in different ways, but for many individuals, speech gets worse over time. Measuring intelligibility is an important way we can examine those changes because intelligibility is a functional measure. We speak to be understood, and when someone can’t be understood, it can affect how they feel about themselves and their quality of life.
This project looks at something called the minimally detectable change, or MDC. The MDC tells us how much a measure of intelligibility needs to change for us to be confident that it represents a real change rather than normal variability or measurement error.
For example, if a patient is 90% intelligible during an evaluation and 85% intelligible six weeks later, is that difference simply variability in the measure, or is it a real change that we should care about? The MDC helps us answer that question.
We looked at whether the MDC differs based on the cause and severity of dysarthria. We also examined whether characteristics of the listener matter—for example, whether the listener is a trained speech-language pathologist or someone with no experience with dysarthria.
What sparked your interest in this research?
This line of research started when I was a PhD student in rehabilitation sciences. I learned alongside physical therapists and occupational therapists, and that experience really influenced the types of questions I ask and the way I answer them.
As I was studying dysarthria and intelligibility, my PhD advisor and I kept coming back to a question: How much does someone need to change for us to care about that change? How much change in intelligibility is meaningful to a patient?
My physical therapy and occupational therapy colleagues were using measures such as the MDC and the minimal clinically important difference. These thresholds were well established for many of the outcome measures used in those fields, but when I looked through the speech-language pathology literature, there was very little work systematically evaluating them.
My first PhD project looked at these measures in people with ALS. That led to more questions: Do these measures look different for different types of dysarthria or different patients? I continued the work through my postdoc and realized there is so much more to explore.
It’s really neat to take well-established methods from other rehabilitation fields and apply them to ours.
What impact do you hope this research will have?
The first place I see this helping is research. I want us to be able to better interpret whether changes in speech actually matter—to researchers, clinicians, and the patients we’re trying to help.
Our field often looks at whether a change is statistically significant. But statistical significance doesn’t necessarily mean that a change is meaningful to a patient or makes a difference in everyday life. For example, research might show that a therapy increases intelligibility by 1% on average and that the change is statistically significant. But a 1% change could still fall within normal measurement variability.
My hope is that this work will help us interpret research in a way that tells us whether we’re seeing a real or meaningful change.
I also hope it will help speech-language pathologists make decisions about when therapies should start or change. For people with ALS, for example, there has long been a question about when to begin recommending augmentative and alternative communication, or AAC. If we better understand what changes are meaningful to patients, we may be able to intervene at the right time and make better decisions about communication support.
Ultimately, the goal for me is to improve quality of life in patients with dysarthria. I hope this research can help us understand the changes that matter to patients and, in turn, improve how patients feel about themselves and how they communicate with the people in their lives.
How is your approach innovative?
One thing that’s exciting about this work is bringing measurement methods that are well established in other rehabilitation fields into speech-language pathology.
We measure intelligibility by recording speakers saying sentences and then asking listeners to write, word for word, what they think the person said. From those responses, we calculate a percentage of intelligibility.
For this ASHFoundation-funded study, I had recordings from 100 speakers, including healthy control speakers and speakers with ALS, multiple sclerosis, Parkinson’s disease, and Huntington’s disease. About 60 listeners (30 people with no experience with dysarthria and 30 speech-language pathologists) listened to subsets of those recordings.
We then looked at how reliable those listeners were. If someone listens to the same speech at different times, do they hear and write the same thing? That helps us understand how much variability or error exists in the measure.
We found that for speakers with mild impairment, the MDC was about 5% to 6%. In other words, we need a change in intelligibility of about 6% before we can say it is outside of measurement error.
We also found that the MDC is strongly influenced by the severity of the speaker’s dysarthria. For someone with very severe dysarthria, the change may need to be much larger, sometimes between 20% and 30%, because there is more variability in what listeners understand and how they transcribe the speech.
Interestingly, the cause of the dysarthria did not have the same impact. It didn’t really matter whether someone had ALS or Parkinson’s disease; severity was the main driver of differences in the MDC. That suggests we may be able to apply similar MDCs across patient populations.
Now that we have a better idea of these percentages, I’m interested in clinically important change: What is actually meaningful to patients, care partners, and clinicians? The MDC tells us how much change we need to get outside of measurement variability. The next step is understanding how much change is meaningful.
What challenges did you face, and how did ASHFoundation support you?
When I applied for the grant, I was a postdoc at the University at Buffalo and knew I was about to start a faculty position there. Starting your first faculty position is a lot—there are so many things to learn and so many things going on.
This was the first grant I had the opportunity to apply for to support my own research. Even having the opportunity to apply and a deadline to work toward was helpful. We all need deadlines to push things forward.
Having funding specifically set aside for this project was also incredibly valuable. Starting that first project when you’re totally independent can be intimidating, and the ASHFoundation grant was the push and support I needed to get it going. It was so nice to have the ASHFoundation say, in a way, “We believe that you can do this, and here’s some money to do it.” That support really means a lot when you’re starting out.
View More Recipient Spotlights