Sleep Sensors for Data-Driven Asthma Management: Interview with Tueo Health’s Dr. Bronwyn Harris


About 10% of children between 5-17 years of age, or more than 7 million kids in the US alone, suffer from asthma. In half of these cases, the child’s asthma is not well controlled. This is typically due to a variety of reasons, such as inconsistent use of inhalers or the inability of parents or caregivers to successfully identify signs of poor asthma control. Given the difficulty that families face in keeping track of their child’s symptoms, solutions that enable better measurement of asthma control are urgently needed. That’s why Tueo Health (pronounced as “too-ee-oh”) , a startup out of Stanford University, is currently recruiting participants in a trial to further test their holistic solution that monitors children with asthma, and provides actionable insights and guidance to caregivers. Medgadget recently had a chance to speak with Tueo’s co-founder and CEO, Dr. Bronwyn Harris, about the company’s innovative approach.

Mohammad Saleh, Medgadget: Tell us about your background and how you came to be where you are now.

Dr. Bronwyn Harris, Tueo Health: I started out as a biomedical engineer. I did my undergraduate degree at Brown University and then considered medical school – but I couldn’t think of a strong reason to do it over being an engineer. So I went into a PhD program for Biomedical Engineering where I very quickly realized that I liked engineering and liked research, but in short spurts and for bigger picture thinking. I really enjoyed the time that I spent with the physicians and learning how healthcare was actually delivered, including the important challenges. I also realized how big of a gap there was between ongoing projects and what needs actually existed in the hospital. I felt that I was well-suited to be able to communicate with both worlds, but in order to do that I needed clinical experience. So I changed course, went to medical and then came out to Stanford for my pediatrics training. During my cardiology fellowship, I was also able to go through the Stanford Biodesign fellowship. It’s a needs-based innovation program where multi-disciplinary teams were put through a rigorous training to identify and understand clinical needs before you think about inventing. I realized that this is the kind of thing I want to do.  During the fellowship I met two of my co-founders, Todd Murphy and Michael Carchia. Todd and I were on a team together and we looked at a lot of different needs, but one that rose to the top was asthma. We felt that it wasn’t being addressed fully yet, and that there was a lot of opportunity out there. A digital health solution was really the answer. We can have a huge impact when digital health is used appropriately – with a clinical slant to make sure we’re actually moving the needle in healthcare towards better outcomes.


Medgadget: What is Tueo and what does it do?

Harris: Over the past few years, there has been a growth in very comprehensive sleep monitors that can tell you a lot about physiologic parameters. We’re using an off-the-shelf sensor that gets attached to a mattress and can monitor the child every night – all they have to do is sleep in their bed. Our solution takes the data from these monitors to create an individualized baseline for each child so we can know when they deviate from their normal. That’s really powerful, because we can then engage the family – we know when to reach out, when a problem might be occurring, when to send automated alerts and targeted educational material that follows NIH guidance. The ultimate result is allowing people to improve adherence and control because they’re able to see the impact of their actions and remain engaged. It’s hard to stick to a medication like inhalers while worrying about any potential side-effects. It’s even harder if you can’t see the positive effects of these medications, which usually takes weeks. So, having objective data and the ability to track their symptoms is really powerful.


Medgadget: So a parent places a sensor on their child’s bed… and then what happens? How do they interact with Tueo?

Harris: The sensor is placed on the mattress and then connected to Wi-Fi. It calibrates to the child and the bed. Once that’s in place the data is automatically sent to us – nothing else needs to be done. The family interacts with us through the app. A lot of it is automated – they get to view a summary of the data we collect, there are education modules for them, and if there is a significant change they receive automated alerts. When there’s a change, the automated alert flow asks questions because we need to know what else is going on – what symptoms the child may be experiencing, any medications they are taking, and so on. There’s also always an option to talk to an asthma educator or coach who can help further the conversation. This is not to replace the medical team – it’s meant as an augment for education and guidance. Within the app, there’s also the ability to request a summary report that can be shared with a physician. It’s a powerful way to summarize everything that has been recorded, and it’s put in a format that physicians are used to seeing. It can answer a lot of the detail-oriented questions they might have during a clinic visit but are generally hard for parents to remember from weeks or months ago.

Medgadget: You mentioned some physiological parameters that these sensors are picking up on. What kind of cues factor into the data that you collect and analyze?

Screenshots of the symptom tracking screens on the Tueo AppHarris: We collect a variety of different factors. It’s not just one sensor that we can work with, but the one we used in our initial study is an accelerometer-based sensor that measures small vibrations. From that, the way you can actually get physiologic parameters is actually based on an old scientific principle called ballistocardiography. Every time your heart beats there’s a characteristic vibration from the blood being ejected into the vessels. When a child is lying still, you can feel that vibration and you can deduce beat-to-beat time. So, you can get not just average heart-rate over time, but also heart-rate variability and effects on respiratory rate. There are other, more detailed parameters that can be measured using this technique that we are also collecting.

It makes a lot of intuitive sense – asthma is an inflammatory disease, and when your body is in an inflamed state, it’s not surprising that there are significant changes in your physiology but we have been the first to prove this concept. In patients in our observational study, these changes can occur weeks before they’re having an exacerbation – before the family can actually tell that the child is having symptoms. It’s a really powerful tool, but it needs to be incorporated as a full solution. The data alone doesn’t automatically lead to improved outcomes. The key is to take powerful data and figure out a way to incorporate it into patients’ lives.


Medgadget: Do you plan on looking into wearable devices?

Harris: There are a few problems with wearables… It’s not completely passive, and data shows that while there’s a subset of people who will adhere and wear it, you wouldn’t reach the full patient population. A lot more people sleep in their beds. Especially for children, it’s challenging – a wearable is one more thing to have to do. Additionally, the data from a wearable is not as clean. While you can get heart rate, most of them don’t give you heart-rate variability and most aren’t as accurate. Using our approach, we usually see five to six hours of quality physiologic data throughout the night corresponding to times when they’re lying still.

We’re also typically asked why we’re only collecting data at night only – a wearable would collect data throughout the day. But night-time is actually the best time to measure this data. In addition to being a much more passive approach, asthma symptoms typically get worse at night due to the mechanics of lying down. You also have a natural low in your endogenous steroids, which results in increased inflammation and worse asthma control. And obviously monitoring at night decreases the confounding variables. We’ve found that healthy children and asthma patients who are well controlled both have very consistent parameters on a nightly-basis when they’re sleeping. During the day, it’s much different – you might have a stressful day, have midterms, you’re super-active or dehydrated. A lot of those confounding variables are removed when you’re asleep, making it an ideal time to monitor.

Medgadget: Does Tueo currently serve as an intervention? Do you notice a change in the physiology and then prompt the parents to change the way they’re dealing with the disease?

Harris: We’re not officially diagnosing or changing treatment plans. However, there are lots of things that one can do to improve asthma control outcomes. For example, a simple thing is medication adherence. If we’re seeing a change, we can check in and ask if they’re properly taking their medicine. If they are, but we’re still seeing a change, the guidelines say that you should review inhaler technique – it’s particularly challenging for children to properly receive this medication. There are also a lot of environmental changes – even as simple as washing the bedsheets more often. Dust mites are a trigger for many children. Sheets should ideally be washed once a week in hot water, or the dust mites will not be killed. So, there are lots of little educational components that have been shown to make a big difference but that are hard to get the word out there about and engage families if you don’t have something like this insightful data to really draw them in.

One other really unique thing about asthma is that already the way the medical system is set up is that there are asthma action plans. That’s really the gold-standard right now. When a physician sends a patient home, they’re given a plan that they’re supposed to be using at home. It’s three or four zones – typically the red, yellow, and green zones. A physician usually indicates a change of medication depending on the zone. Green zone is doing well with no asthma symptoms, so stick to the prescribed medication. The yellow zone is a little ambiguous for mild or moderate symptoms such as some coughing or trouble with activities, and some physicians can prescribe taking double the controller medication when patients are in this zone or add another medicine. The red zone is more of an emergency situation. This plan is prescribed by the physician, and we can help families better use that tool. That action plan can be put on our app by taking a simple picture so they don’t worry about losing that piece of paper. We then refer them to that plan when there’s a noticeable change or they’re reporting symptoms. So, we’re not prescribing or changing the medication, but we can help have the information readily available from their physician so that parents would know what to do when they’re in that situation.


Medgadget: Could you touch on the science underlying the algorithms that analyze the collected data?

Harris: We have a great team who put this all together, led by co-founder and CTO, Michael Carchia, who has a unique background in medicine, engineering and data science. We ran an observational study at Stanford with Dr. David Cornfield, director of the pediatric pulmonary department, and Dr. Michelle Huffaker, an allergist, allowing us to collect a data set from children who have asthma with various levels of control. Utilizing that data, we were able to create an algorithm to specify individualized expectations for each patient. The more data we have about our particular patient, including external datasets such as weather and air quality, the more individualized and personalized the algorithm is for that child.


Medgadget: Based on your past studies, how accurate are these algorithms?

Harris: We’re going to be publishing a paper soon so I can’t give specific statistics yet, but from the initial observational study we did in Stanford, we’ve performed statistical analysis to look at the prediction of asthma symptoms compared to non-asthma symptoms based on these parameters. The accuracy of these parameters for detecting asthma symptoms is impressive. It’s important to note for our solution, though, that we are not making the claim that we are assessing asthma control. We’re looking at these physiologic parameters and an adjusted overall score, and we’re looking for deviations to be used as an engagement tool. I believe that down the road this can completely change the way that we care for asthma, but we’re not there yet. We don’t have a sliding scale like a glucometer where it’s the case of “hey, your score changed so you should automatically change your medication.” We are working to get larger data sets, so we can get there.


Medgadget: What does the current market look like for digital health solutions to asthma control?

Harris: We’re not the only company to realize that asthma is a big problem. It’s getting more press and people are paying more attention. Most of the other solutions out there right now are focused on either medication tracking – like a Bluetooth device on the inhaler that keeps track of medication adherence. Those are also absolutely useful, but it’s not directly tied to the patient – it’s still based on subjective measures. There are two types of medications; the controller medication needs to be taken regularly as prescribed, so you can give patients a reminder but you don’t have other data to really encourage them like we do with the physiologic data. The other type of medication are rescue inhalers, so tracking its use can give you an idea about their asthma control. However, it’s still based on the patient feeling symptoms. Some patients are poor perceivers and tend to not realize when their asthma is poorly controlled. Other cases may involve something unrelated to asthma, but the patient thinks it might be so they try to use their inhaler giving the trackers a false indication that their asthma is not well-controlled. I have a five-year-old who has asthma and sometimes my two-year-old would press the inhaler, completely unrelated to my daughter having any symptoms. This type of tracking is not directly tied to the physiology of the patient, it is tied to actions around activating the inhaler.

Screenshot of the tools available for caregivers through Tueo

The other class of solutions that exist are connected peak-flow meters, sometimes used in combination with the trackers. The problem with this tracking is that it’s pretty late. They measure different changes in the airway, and so to have any significant change, you need to already have some airway narrowing or obstruction. We’re not talking about the kind of physiological changes that we can see weeks in advance. And there’s of course the question of active vs passive tracking. Using the peak-flow meter is a very active process. To be predictive with these meters, you would have to use it regularly once or twice a day to really know what the baseline is and what qualifies as a deviation. Consistency with this is very challenging with children, and the American Academy of Pediatrics doesn’t recommend their use in children because they don’t feel there’s adequate data showing that it improves care.

As of now, the other solutions also don’t have a complete solution with respect to family engagement and educator support built into it. There are parallels in diabetes where many companies are building comprehensive solutions, but in asthma there hasn’t been a good type of monitoring to enable that. We are the first company to put this all together.


Medgadget: Tell us about the clinical trial that Tueo is actively recruiting for.

Harris: We don’t want to put this solution out there and just convince parents to buy it without data. We want this prove that this is a product that actually enables better control of children’s asthma, so it’s really important for us to run a randomized clinical trial to show improved outcomes. We’re running a virtual study, now open for recruitment across the 48 contiguous states. We’re recruiting 260 children between the ages of 6-17 years for a four-month study where half of the patients are in the control arm (and won’t receive a monitor) while the other half do.

If anyone reading this wants to learn more about the study or see if their child may be eligible, they can consent online and begin the study. [Find more information here.]


Medgadget: What’s your vision for Tueo over the coming five years?

Harris: Being a pediatrician and understanding asthma as both a physician and a parent, I was passionate about starting with pediatric asthma. Beginning in pediatrics also made sense because children usually don’t have other co-morbidities or chronic diseases, and parents are often passionate about caring for their child and more willing to adopt a new solution. But adults can also benefit from this, so we will be quickly expanding into the adult patient population.

There are also other disease states that can benefit. COPD [Chronic Obstructive Pulmonary Disease] will be a natural next step. We’re also looking into other potential diseases where there are physiologic changes that can benefit from this type of monitoring. This is a really exciting area of utilizing data at home to get a better picture of patients with chronic disease. We don’t just want to have data for data’s sake, but to use that data, follow guidelines, help medical personnel do a better job, and help patients and their families better manage their care.

More about Tueo on the company’s website: Tueo Health…

Source: Medgadget


Please enter your comment!
Please enter your name here