Mudit Dandwate, CEO and Co-Founder, Dozee tells Viveka Roychowdhury about the evolution of his start up, and how they pivoted to track vital signs of COVID-19 patients in wards, spanning urban hot spots as well as in tier 3 and rural towns.
Dozee, a contactless sensor, placed under a patient’s mattress, captures real-time body vitals without using any external wires or touching the user’s body, and timely detection of a heart condition/attack with Dozee has proven life-saving in 70+ cases, according to Dandwate.
He believes remote patient care with innovations like Dozee can provide affordable, accessible, predictive healthcare in people’s homes, with hospitals used only for surgeries and higher order care, while homes become the hub of primary care. This is how we will ensure healthcare is accessible to a billion population.
Tell us how Dozee’s business model/rationale evolved from its launch in 2015, to being productised after five years.
Our product in its current version has come into place after four years of R&D. We have worked extensively with doctors, hospitals and healthcare experts to understand the needs of the patients and doctors and iterated every feature to best suit the needs of the users.
Healthcare which is a basic necessity in its existing state was very complex, we saw an opportunity of making it simple, to make healthcare more accessible. Currently in India, only 1 lakh beds are being continuously monitored and there are about 19 lakh beds which are dependent on human intervention. We hope to bridge this gap in the coming years.
What is the total quantum of funds raised so far, after the recent INR 12.5 crore raised this September? Who are the funders and what’s the path ahead?
Dozee recently raised INR 12.5 crore from Prime Venture Partners, YourNest Venture Capital and 3one4 Capital. We are already seeing a strong adoption for remote patient care, both in hospitals and at home. We plan to continue working on our AI so that it becomes even more efficient in decoding the health data and identifying illnesses and diseases faster. We also plan to extend our suite of services in the future to provide consumers with a holistic platform for all their healthcare needs. We are especially expanding the scope of our data science to ensure Dozee can accurately predict specific disorders, esp. in neurology, pulmonology and cardiology space.
On the hospital front, we are seeing good traction this year and we have partnered with over 30 hospitals already. We are planning to scale across the country and have our devices at many more wards.
We are also targeting Tier II cities as the doctor to patient ratio is extremely skewered here and having our device in such regions will ensure that hospitals run efficiently by automating the vitals monitoring process. We have already tied up with Indira Gandhi Government Medical College & Hospital and Government Medical College in Nagpur recently where 250 of our devices have been deployed. We hope to partner with more institutions in smaller cities.
We’re also in the middle of our regulatory approvals (FDA, ISO 13485, etc.) for access to international markets like the US. Our plan is to establish ourselves throughout India in the next 12 months and then launch it in the international market.
What is the technology behind Dozee? What is the degree of accuracy?
Dozee works on a technology called ballistocardiography (BCG), a non-invasive method that tracks vibrations from every heartbeat, respiration and smallest of body movements. Dozee has built a proprietary Advanced Health Intelligence System that extracts biomarkers and vital parameters from the noisy vibration data captured by the sensors. The AI system helps in identifying health deterioration through abnormalities across all vital parameters Dozee measures. In many cases, the AI system has successfully predicted early health deterioration in conditions ranging from fever to heart failure, even before the patient could see the symptoms.
Dozee tracks key vitals of the human body such as heart rate, respiration rate, oxygen saturation, sleep stages, stress-recovery and more with a medical-grade 98.4 per cent accuracy. The contactless sensor, placed under the mattress captures real-time body vitals without using any external wires or touching the user’s body. The sensor captures micro-vibrations produced by the body every time the heart pumps blood, during inhalation, exhalation, muscle twitches, tremors and body movements. An AI-powered, Early Warning System then converts these signals into biomarkers and uses the data to present an analysis of the patient’s health through a smartphone app, for retail users.
What have been the learnings and tangible outcomes since we now know that the SARS-CoV-2 virus, which results in COVID-19, impacts many more organs than the respiratory system, especially the heart
COVID-19 primarily affects the respiratory tract. Dozee was able to detect an increase in breathing and tachycardia among many COVID patients. The device has also closely tracked reduced heart rate variability, increased sleep apnea and breathlessness. We are currently in the process of researching an indicative of heart damage caused by the novel COVID virus.
Can you share some use-case examples of patients/people whose heart attacks were averted by the use of Dozee.
Sudden cardiac death is largely preventable if patients with a heart condition/attack get treatment in time. Early detection, diagnosis and management of these cases can help in bringing down the burden of heart attacks. Continued and consistent monitoring of vitals leads to early detection of future conditions. Dozee placed under the mattress continuously collects data while the patient is in bed and the system alerts healthcare professionals if rates change, indicating patient deterioration. Patients with remote monitoring devices have a probability of survival that is nearly two and a half times greater than those without it.
Timely detection with Dozee has proven life-saving in 70+ cases. This includes an elderly woman living alone in Bengaluru. Her grandson, who had put Dozee under her mattress, got an alert of a falling heartbeat. On consultation with the doctor, it was found that she needed a cardiac pacemaker to set things right. Further delay would have proven catastrophic.
Dozee has similar success stories in hospital settings, where patients have been shifted to ICU timely due to alerts raised.
These devices are mostly an urban phenomenon or are there cases where it has been used in quarantine centers in rural areas during COVID-19? If so, what is the use case? Lack of doctors in rural areas can be supplemented by such devices?
Yes, Dozee is already being used in tier-3 and rural towns for monitoring COVID-19 patients in wards. Eg:, Thiruvallur in Tamil Nadu etc
We are also working with governmental organisations to screen patients for cardiac failure in villages of Rajasthan, Assam etc – this is bringing quality care to villages and ensuring patients at-risk can be identified early on.
How many devices in use before and after the pandemic? What has been the rise in business this year?
The company has seen 5X growth in sales of its remote health monitoring device during the pandemic in hospitals. Currently, we have installed 1400+ devices in hospital beds. We have also seen a 50 per cent growth in consumer settings as well for home monitoring and we are currently monitoring 5k+ users at home.
All medical devices are work in progress, as there are further revisions based on users’ feedback. So which features are the future versions of this device going to add on?
There are feature upgrades planned in various verticals, right from communication protocols, sensing technology to even adding more vital signs. We will be adding more reports helpful in various disorders like cardiac arrest, neurological conditions etc. We also plan to upgrade the hardware, making it more efficient and accessible especially in rural settings using a SIM card.
What is the scope of AI in early prediction and diagnosis of disease conditions like CVDs, etc?
AI is at the core of our offering. We’ve developed AI/ML algorithms that extract biomarkers and vital parameters from the noisy vibration data capture by the sensors. These are unsupervised pattern recognition algorithms that calibrate to each new user automatically. The result is detection of heart rate with 98.4 per cent accuracy when compared to an ECG for over 5000 hours of validation data.
We also use deep learning algorithms to identify anomalies in the vital signals, for eg, sleep apnea episodes in the respiratory signal captured by Dozee. Another set of deep learning algorithms extract patterns within the cardiac cycle to measure individual valve openings and closing timings. This capability helps us screen for heart failure. The most important use of AI however is in identification of health deterioration through abnormalities across the vital parameters Dozee measures. Our engine for identifying these events is called Advanced Health Intelligence (AHI).
What is the scope of contactless remote health monitoring devices, from a cost benefit analysis PoV? Moving from the individual patient, to the country level?
There is immense potential here for a product like Dozee. Since our product is made for hospitals as well as individuals, we hope to gain more market share in the vitals monitoring segment. We are also aiming at Tier II cities right now as the doctor to patient ratio is extremely skewered here and having our device in such regions will ensure that hospitals run efficiently by automating the vitals monitoring process. At country level, with increased device usage, we aim to reduce burden on healthcare infra by upgrading home beds into hospital-like beds.
Since the device captures body vitals real time and transmits them to doctors, what are the data protection and privacy norms? With whom is the data shared? And where is it stored?
We follow HIPPA and GDPR privacy norms. Data is shared with the third party only if the user wants to provide access to their data. Users also have the power to revoke access anytime.
Inside the system, User’s Personal Identifiable Information and health data is stored in different databases. So, it is not possible to identify the health data of a particular user, until and unless the user authenticates.
Is Dozee using and sharing this patient data for any predictive population level studies etc?
Our AI engine works on health data to continuously improve itself and provide more predictive outcomes to users. This is how accuracy keeps improving and other health conditions open up.
How can startups like Dozee create a better healthcare ecosystem for the post-COVID generations? And at what cost?
New innovations like Dozee can provide affordable, accessible, predictive healthcare in people’s homes. Moving forward, hospitals will be used only for surgeries and higher order care, while homes become the hub of primary care. This is how we will ensure healthcare is accessible to a billion population.