How Beddr can help solve the sleep data problem
The science of sleep has evolved considerably over the past several decades, but for all that we know, there is even more that we don’t. Part of what holds the industry back is a lack of high quality data from which to infer the true relationship between our actions, choices, sleep quality, and our overall health.
Sleep has a data problem
A tremendous amount of sleep research is conducted every year. Typically, the sample sizes are very small and it is difficult to track the participant over a longer period of time. As a result, there is no complete, curated and clinically valid set of sleep data. Much of this is due to the fragmented nature of the sleep industry.
On the one side, there is the gold standard of sleep testing, polysomnography (PSG). A PSG test captures a tremendous amount of data, but is restricted to only a single night due to its cost, complexity and inconvenience. That makes it very thorough at making an initial diagnosis, but impractical at tracking how our sleep evolves and responds to various changes or specific therapies. As a result, it misses out on the impact that various therapies could have on delivering real, helpful results to someone who is struggling with their sleep health.
On the other side, there are consumer sleep trackers. While these are easy to use and readily accessible, they can’t capture the necessary clinical data required to truly understand sleep problems. The data they are able to capture lacks the accuracy required for a formal diagnosis of a sleeping problem.
A Beddr Approach
From the beginning, we believed that data science would power the next phase of sleep medicine. Our first task was to build a way to easily and affordably gather high-quality, clinically-valid data over a longer period of time.
Clinically Valid Data
There are any number of sleep gadgets that gather sleep data. Some will claim that they have recorded millions of nights of sleep data. But most of that data is not particularly useful in the eyes of a trained sleep physician. This is why when a patient brings their sleep data from a fitness tracker, most physicians politely nod their heads, but are unable to do much with the data.
The Beddr SleepTuner is an FDA-Class II registered device that gathers highly accurate information including blood oxygen saturation, heart rate, heart rate variability, sleep position, sleep duration, and night-over-night variation. The information gathered represents the key data that sleep physicians have relied upon for decades to more fully understand a patient's sleep problems and ultimately arrive at a formal diagnosis.
Gathering Data Over Multiple Nights
Our sleep changes night-to-night and is highly dependent upon changes in our lives.
Historically, it has been very difficult to gather clinically valid sleep data over multiple nights. Most of this is due to cost and complexity, but also to the limitations of insurance reimbursement.
A single overnight test at a sleep lab fails to capture the full picture of a person’s sleep health and is often dismissed by the participant because they do not believe the test accurately reflects their “real” sleep in the comfort of their own bedroom. More importantly, this kind of experience is unable to measure the impact of changes we make in an effort to improve our sleep.
Beddr is designed for the user to gather clinically valid data at different points in time. Hence it not only establishes a baseline analysis of your sleep, but it also measures the impact of simple choices like changing sleep position or reducing alcohol consumption before bedtime.
Applying Complex Statistical Models
Beddr’s data science team applies complex statistical models to the data that is gathered. As we gather more data, we anonymize it and feed it into our proprietary statistical models, enabling the system to generate insights, make recommendations and predict a range of outcomes. As more data is fed into the system, our system “learns” and grows more personalized and precise.
To build our data models we have worked with some of the leading experts in the field of sleep medicine.
How Data Science and Predictive Analytics Will Benefit Users and Clinicians
There are a number of ways that the data gathered by Beddr will be used to help people with sleep problems as well as sleep physicians. These include:
- Revealing new insights into the cause and effect of changes to diet, weight, exercise, alcohol consumption and sleep position on sleep quality.
- Enabling cohort analysis to predict the impact that changes or therapies could have on an individual with a similar set of characteristics.
- Providing a clinical decision support tool to diagnosing physicians that augments their clinical experience and intuition with access to a dataset that will reveal new insights.
- Assisting clinicians in engaging, monitoring, and following up with patients who struggle with sleep issues or are attempting to adjust to a new treatment routine.
The Future of Sleep Medicine
Data science has tremendous potential to help patients and clinicians better understand and solve the escalating sleep crisis. Predictive analytics provides the unique opportunity to harness the “wisdom of the crowd”. Leveraging the success of others to motivate and instill confidence in each individual and their ability to improve their sleep quality and overall health. It will also provide clinicians answers to some of the most vexing questions in sleep medicine and provide them with a toolset to better support their patients to the best outcome possible.
Beddr has taken the first step towards making this possible by creating the SleepTuner, the first accurate, affordable sensor that captures clinically valid data over multiple nights from the comfort of a person’s home.
Mike Kisch likes to make the complex simple, engaging, and accessible to more people. He is passionate about applying this philosophy to healthcare. Previously, Mike was the founding CEO of Soundhawk, a wearable hearing enhancement company that developed the first connected hearing device. He led the company from concept to commercialization and multi-million dollars in revenue. He holds an MBA from Washington University in St. Louis and a BA from University of Wisconsin-Madison.