It’s a scary thought. Of the 18 million Americans estimated to suffer from sleep apnea -a condition that causes interrupted breathing whilst you’re asleep – 80% are thought to go undiagnosed.
This come a heavy cost. Obstructive sleep apnea (OSA) is associated with a range of comorbid diseases, including hypertension, heart disease, diabetes and depression to name a few.
OSA sufferers also experience increased daytime fatigue which leads to lost workplace productivity and a higher risk of road traffic and industrial accidents – and is thought to be implicated in at least 3 train crashes in New York in the last 3 years.
Why do so many people have undiagnosed sleep apnea?
Essentially it all comes down to cost and accessibility. Broadly there are two main ways to establish a sleep apnea diagnosis.
The ‘gold standard’ of sleep testing is called a polysomnograph (PSG), which involves an overnight stay in a sleep lab, doctors, technicians and heaps of wires, sensors and specialised medical equipment. Needless to say, a full PSG is expensive, sometimes costing many $1000s – not always an option for people with inadequate health insurance.
A cheaper alternative to PSG is home sleep apnea testing (HST). Home testing is more convenient and less expensive than a PSG and but reveals less data. And whilst the cost of HST runs into $100s rather than $1000s, it’s still too much for some.
We need better sleep apnea screening tools
For a long time PSG and HST were the only two options for reliably testing for sleep apnea. But in recent years, a wave of innovations in digital health, big data and biometric sensor technology and have paved the way for less expensive, reliable screening methods to detect cases of undiagnosed sleep apnea.
Since 2012, Washington DC based startup Zansors, a healthcare innovation company, has been researching and developing a range of affordable sensors, apps and algorithms designed to help patients take more control of their health issues.
Unlike many of the current swathe of consumer sleep monitoring devices, Zansors has made a point of focussing on rigorously tested evidence-based wearables, employing a multi-disciplinary team with expertise in many areas including medicine, bioengineering and data analytics.
Earlier this year, Zansors was awarded a $1.5 million grant by NIH to pursue Phase II testing of its latest innovation, a wearable at-home sensor for sleep apnea detection. Earlier Phase I testing at the University of Michigan provided positive results and showed that that the sensor performed well compared to gold-standard diagnostics for sleep apnea.
Speaking about the latest grant, Zansors cofounder and Chief Data Science Officer, Dr Abhijit Dasgupta said “We are using our technology to combine wearable sensors, algorithms and bioengineering into a first of its kind product”
We spoke to Dr Dasgupta via email to find our more about the future of sleep apnea screening and how Zansors’ sleep technology could benefit millions of needlessly undiagnosed OSA sufferers.
Jeff Mann (JM): Hi, thanks for taking the time to speak to Sleep Junkies and congratulations on your recent NIH funding. Can you give us a brief background to Zansors, how you got involved, and the ‘elevator pitch’ for your new wearable screening tool for sleep apnea?
Abhijit Dasgupta (AD): Zansors is a bioengineering startup based in Northern Virginia, specializing in sensor-mobile app combinations to help users “Know Themselves.”
We develop evidence-based sensor platforms and mobile apps for different aspects of a person’s health, with our first focus being the sleep experience.
These products are based on peer-reviewed research and validated in Institutional Review Board (IRB)-approved clinical studies.
The focus of the company is to produce high quality tools, as opposed to consumer toys, that can help users learn about themselves.
I guide the research and development of different products funded by NIH SBIR/STTR grants and develop algorithms based on domain knowledge and the best practices of data science. Zansors focuses on data as the means to provide evidence of effective products, and my background as a biostatistician advances that focus.
Our goal for the sleep apnea screening tool is to provide users with a small, comfortable, easy to wear, accurate tool based on sensors, algorithms and a mobile app that can help them self-screen for sleep apnea in their own bedrooms.
Sleep apnea is widely believed to be under-diagnosed, and we want our tool to be a way for individuals to know if they need to get help for apnea, which we know today is associated with a wide range of other health issues.
Our Phase I NIH-funded project tested our tool on 50 subjects at the University of Michigan Sleep Center against gold standard diagnostics, and in the newly funded Phase II project, we plan to test this product on over 100 subjects against both home and inpatient state-of-the-art clinical diagnostics for sleep apnea, enabling us to create a product a user can depend on.
JM: Home sleep apnea testing has been available for while, what are the main benefits and features of your sensor compared to existing products?
AD: Tests currently used for sleep apnea are uncomfortable and expensive, costing upwards of $4,000 per night. We are providing a tool that is affordable, accessible and easy to use, without creating disrupting everyday life.
Spending the night in a sleep lab with 20 wires on your body hooked up to clunky machines (or even in a home sleep test with a chest strap and several wires attached to you) cannot be conducive to normal sleep behavior.
Our product is almost invisible in terms of feeling it, so it doesn’t disturb or affect sleep patterns. This may enable us to get a more realistic reading on a person’s normal sleep patterns and habits.
Most importantly, we are going to empower our users to self-discover an aspect of their own health, specially one they cannot directly observe, in a reliable and objective manner. Because the technology is affordable, apnea screening will be available to underserved or poorer communities.
JM: The prevalence of sleep apnea is reported to be widely under-diagnosed. Many industrial accidents, including road and rail crashes have been attributed to sleep apnea. How do you consider your sensor can help improve this situation?
AD: It is reported that over 80% of the estimated 18 million or more cases of sleep apnea in America remain undiagnosed. What our sensor product can provide is a way for individuals to self-screen for apnea and get treated if necessary.
This changes the game because our tool will remove the barrier to reliable apnea screening, bring it from the clinic into the home, and circumvent the current need to go through our healthcare system and referrals to even get a preliminary finding.
It will also provide an affordable, comfortable way for someone to continue to monitor themselves without disrupting their lives.
You bring up the very important point of how poor sleep can affect occupational safety. We can envision our sensor being adopted by industry as a way to screen early for sleep issues among their critical personnel, thus providing the means to improve workplace safety.
We see this as particularly useful to transportation, trucking, and manufacturing/engineering industries where mistakes can be costly in terms of lives.
JM: One of the criticisms of modern sleep technology products (such as sleep trackers) are claims of accuracy and efficacy. When so-called ‘sleep scores’ are determined by proprietary algorithms, how can consumers be confident that their sleep wearable can actually yield any meaningful data?
AD: This is a very important point. Zansors believes in evidence-based products, and as such all our products have undergone peer-review at some stage, and are clinically proven in IRB-approved studies.
The apnea product has been thoroughly reviewed by the NIH through the grants process, and has successfully graduated from Phase I to Phase II through peer-review. Even if the algorithms are complex and proprietary, the results of our clinical studies stand as evidence of the effectiveness of the product. This is more than can be said for many sleep technologies on the market today.
In these studies, we are going head-to-head with current clinical technologies using standard clinical definitions and metrics, which should provide additional confidence that this tool can provide meaningful and reliable data.
“Sleep scores” are a somewhat murky area currently, where it is often not clear what inputs go into its computation or what aspects of sleep affect it. However, even with such scores, Zansors is moving in a data-driven, evidence-based manner to report on how you slept last night.
JM: There’s an interesting dilemma with the current wave of digital health products. For example I’m talking about the way that big data has allowed companies like FitBit or Resmed to gather millions of nights of sleep data from their users – but because the data is not shared in the wider scientific community, no-one is able to investigate, verify or falsify this potentially useful mine of data. Can we, or should we make the leap from proprietary knowledge (corporate big data), to a shared knowledge base, where science as a whole can benefit?
AD: First of all, your health data is private and deserves the highest protection. Zansors does not share the information users provide with any third parties without the users’ consent. This is both our company policy and is also governed by Federal regulations.
However, there can certainly be ways to share anonymized aggregate data that can be mined for further knowledge. This is actually done in many areas of biomedical research, especially in genomic research where anonymized genetic datasets are routinely released with publication of research papers; similarly anonymized clinical trial data is shared through ClinicalTrials.gov. There are paradigms already in place that can be used for sleep data as well.
There are two balancing acts here. One is the right of private companies to monetize their data, balanced against the need to share these rich datasets for common good. The other is to balance individual privacy with shared knowledge.
At Zansors, we feel that anyone who uses our product owns their own data, and so they are free to share it with whomever they wish.
There needs to be an appropriate balance between keeping your data private and benefiting scientific research by sharing the knowledge base. Our solutions can easily communicate with other stakeholders like healthcare providers through integration with EHR in the larger healthcare ecosystem if the patient decides to share their data.
JM: In terms of sleep sensing technology, most products currently focus on measuring and analyzing sleep data. In order to improveone’s sleep however, people need to change their habits and behavior. What in your view are the best ways to bridge the gap between passivesleep monitoring and active technologies that encourage people to actually improve their sleep quality?
AD: In short, they go hand-in-hand. Passive sleep monitoring provides an objective means to validate the results of active sleep improvement technologies – and creates a cycle of improvement. By allowing earlier detection of progress trajectories, we can build more dynamic regimes where the active technology can adapt to the needs of the user.
Technologies like mobile apps can not only deliver targeted information, content and programs to a user, but can also serve as a central hub to a smarter home environment where personal on-body sensors and environmental sensors and controls can coordinate to improve sleep quality.
One challenge of sleep is that we can’t directly observe our own sleep, only its aftereffects. Solid tools can both deliver active ways for sleep modification and provide validation of success or a way of adaptive modification in case desired objectives aren’t achieved.
This system also allows involved third parties like doctors and therapists to monitor progress and be pinged in case preset rules are violated or particular metrics met.
JM: What are your predictions for future developments in sleep and technology?
AD: The promise of a sensor-rich environment is here today, but is underleveraged. Omnipresent cheap sensors, both wearable and environmental, will transform our understanding of sleep in the wild, outside of artificial clinical settings.
The data collected by these sensors will allow us to develop less expensive, more personalized means to (a) monitor sleep comprehensively in daily life, (b) identify issues early so that individuals can make much better informed choices to seek help, and (c) provide dynamic individualized programs to improve sleep quality.
Zansors believes that with the help of technology, everyone has the right to experience good sleep, which then cascades into many aspects of good health. The symbiosis of technology, knowledge and self-understanding alongside professionally trained personnel will be transformative for individuals to get on the path of healthy sleep.