Q&A With Hyfe CEO Dr. Joe Brew

Dr. Joe Brew
CEO
Hyfe

Dr. Joe Brew, one of the co-founders and CEO of Hyfe, discusses his company’s work developing cough tracking software. Despite being one of the most common symptoms, doctors often rely anecdotal evidence in regard to coughing as opposed to quantifiable data.

Brew: HYFE is a software company that builds algorithms and products for quantifying what we think is the world’s most important, yet least quantified symptom: coughing. Think about your own experience with the medical system. You go to the doctor’s office and as I’m sure you’ve noticed; they quantify everything in medicine. You hop on a scale, they take your height, check your blood pressure, temperature, respiration rate, glucose, blood oxygen, pulse, every single thing is quantified because putting a number on a symptom, as opposed to asking subjective questions like do you feel hot, is useful in medicine.

For some reason, one of the most prevalent symptoms, coughing, is never really systematically quantified. Imagine if doctors just asked how your blood pressure was doing and then gave you some pills and tell you to come back in a week and tell them how it feels then. It’s actually very out of sync with modern medicine. We don’t think it’s because it’s not considered important. In fact, every single doctor we talk to says that it’s really important.

It's been hard to quantify. There haven’t been tools that are scalable, easy to use, and patient friendly.

(MDT:) How important is it to track coughing?

Brew: Cough is so important that it’s being used all the time right now in clinical decision making. This is despite the fact that it’s being used very sloppily. Based on your cough, doctors will treat, prescribe, and even admit you to the hospital. They listen to cough and ask about it because it matters. The problem is they don’t have anything firm and objective to grasp onto. They have patient recollection, subjective experience, and maybe five to ten minutes of observation during a consultation.

Almost every infectious respiratory disease has a cough as a characteristic. There’s also a lot of chronic illnesses associated with coughing. It’s well known in lung cancer academic research that cough is a common symptom among those diagnosed, and it’s one of the reasons why people go to the doctor to get screened for lung cancer in the first place. Yet, it’s not monitored.

Similarly, chronic cough is something that doesn’t get a lot of attention. It’s largely older folks and women. People can have hypersensitivity which causes them to cough a lot, which causes irritation in the larynx, which causes the hypersensitivity to get worse. It’s a vicious cycle. People cough so much that they break their ribs, vomit, and many can’t go out in public. It’s a really crippling disease.

While the infectious disease aspect is kind of obvious, there are many people that cough day in and day out for years without getting it checked. Think about trying to manage something like your blood pressure or calorie intake without quantification. For chronic coughers who are trying to reduce the coughing, many have never had a tool to quantify it. There’s a place for cough monitoring in both the infectious disease space but also in the long term, chronic space.

It's hard to get patients to do things that are obtrusive. For example, a patient isn’t going to weigh themselves every day, let alone every hour. Cough monitoring needs to be unobtrusive. It’s been shown that cough frequency along can be used to predict conditions like COPD up to 50% of the time, with a three-day lead time, and only get one false positive for every 100 days.

(MDT:) When it comes to monitoring coughing, what are the elements that you’re looking for?

Brew: In the cough space, it’s like moths flying to a lightbulb. There’s a tendency to try and do the coolest, hardest, and most science-fictiony stuff first. There are groups that are diagnosing diseases based on the sound of a cough using algorithms and AI. We’re doing similar things too, such as our algorithms that analyze the acoustic characteristics of a cough to screen for diseases like Covid-19, tuberculosis, and others.

This is cool and futuristic stuff, and I think in the future, your devices will analyze the acoustics of your voice and cough to diagnose. That’s many years out in terms of the science, and many more years out in terms of the regulatory approvals. We’re much more excited about the low hanging fruit, which is quantification. All of us have coughed, but none of us can quantify that. You have no idea whether you coughed ten times this morning, or six, 100, etc. It’s not something we pay attention to.

Unlike some of the AI, black box, science fiction stuff, there is absolutely no skepticism from the medical community around the importance and usability of cough frequency analysis. Everybody says they want to know how much their patients cough. Only some say they want an AI to tell them whether that cough is Covid or not.

We’re really focused on that base layer. It’s not as glamourous as using AI to diagnose, but we think it’s so applicable across so many conditions for so many people. It’s useful for anyone with lungs.

(MDT:) Does this involve remote monitoring?

Brew: We have a number of use cases, and remote patient monitoring is the one we think is going to have the most impact. Right now, you can go on a smart phone and download Cough Tracker, a wellness app meant to quantify your cough similar to how you might quantify your footsteps.

We have a lot of users that are monitoring their coughs and their bodies. There’s a ton of value there, but we think remote patient monitoring is the next frontier. That information might be interesting to you, but it’s probably more interesting to your doctor. Especially if you have conditions that are not severe enough to have you admitted under constant doctor supervision, but your doctor doesn’t want to send you into the dark for months at a time. These conditions can include asthma, COPD, congestive heart failure, chronic cough, etc. There are the kinds of things were the doctor might be extremely interested in keeping an eye on you. They send you home with some pills, a prescription, and a plan, but they want to doublecheck to make sure that the plan and the medicine is working. With remote monitoring, the doctor could tell if there’s been an improvement or not after a few days and make adjustments.

Remote patient monitoring can put this data stream in the hands of those who know best what to do with it. I don’t think that it’s an unreasonable assumption to think that when this information is put in the hands of medical professionals, they will be able to make an impact.