Caspia Health

Lifestyle Medicine (LM) is a newer field of medicine that prioritizes evidence-based, therapeutic lifestyle changes over pharmaceuticals and surgeries to prevent, treat, and even reverse chronic disease. There are six pillars, including diet and nutrition, physical and mental activity, sleep, emotional well-being, avoiding toxic substances and environments, and social connectedness and purpose. The practice of LM is onerous from the perspective of the professional. There is significantly more information to collect and process about an individual's lifestyle. There is also a health coaching component that takes additional training, time, and efforts. Lastly, there is the tracking component of patient's progress on the action plans. The workflow of current medical care does not allow for easy integration of LM care into any medical field, including primary care. Workarounds, like shared medical appointments, separate LM clinics, and short-cutting the LM care are all techniques employed currently to distribute LM care. Technologic solutions can be very useful, but there are no dedicated platforms for distributing LM care. 

We have developed an integrated Lifestyle Medicine (LM) platform with AI generated, evidence-based lifestyle recommendations using inputs from various sources, like patient wearables, screening tools, questionnaires, medical history, family history, physical exam, and lab results. 

A link to the physician-facing platform is here: https://propatienttech.com

A link to the consumer/patient-facing platform is here: https://caspiahealth.com 

Data Collection

There are no existing pathways to collect the multitudes of patient data required for an LM visit asynchronously. Using our platform, Caspia Health, patients can link their MyChart account(s), Apple Watch, FitBit, MyFitnessPal, and other integrations in order to get the most accurate, time-stamped data directly from the source. Our platform processes this data and returns it to the physician in a view that is ideal for them to be able to make LM recommendations. For other kinds of data, such as health beliefs and family history, we have developed a user-friendly and smart platform that sends patients gentle reminders with embedded teaser questions to fill in this information. Lastly, for validated screening tools that help diagnose depression, anxiety, sleep apnea, or alcohol abuse, we have sought and obtained licensure to distribute these through our platform to patients as well. 

Data Analysis

Pro-Patient Tech assimilates all the data collected into a view that is familiar to the physician without the physician having to manually calculate anything. We are using healthcare standards for the data is already a part of medical care. Lifestyles can also be quantified and described, and that is also collected as part of LM care. Currently, there is no proposed framework for viewing a patient's lifestyle information. Pro-Patient Tech is generating a series of white papers to propose our concepts. 

From this data, we extrapolate an overall Lifestyle Score and analyze the biggest gaps between a patient's current health and their health goals. Using the current lifestyle state and explainable AI, Pro-Patient Tech proposed lifestyle change recommendations starting with the ones that offer the most benefit and organized into each of the LM pillars. The physician will then have a list of ideas to start discussing with the patient to see which they feel most motivated and confident to achieve. 

Action Planning

The final action plan is sent to the patient through the Caspia Health platform. Each recommendation comes with at least three separate resources to help the patient achieve their goals, e.g. beginner recipes, information about mindfulness, and ideas for stretching exercises. Also, there is a built in tracking mechanism for certain goals to aide the patient while they are transitioning. Certain action items are also linked to wearables so patients can track those goals passively. 

Current Target Market

We are currently working with Lifestyle Medicine (LM) physicians who are actively trying to integrate LM into their care and/or are setting up LM clinics. These physicians all have the same pain points: administering questionnaires, synthesizing the data, and action planning with the patient. Currently, there is no way to scale their care since all these tasks are done by synchronous communication with the patient and takes valuable physician time and resources. It is possible to hire someone to collect this information, however that is costly and they often don't collect the information in a way that the physician can use. Part of the reason for this is that there is no standard format for lifestyle information. Using Pro-Patient Tech, these physicians should be able to successfully care for their LM patient with extreme time-saving efficiencies and also scale the care they provide. 

Beyond Current Target Market

Once we make progress with the platform among current LM physicians, we will work with US healthcare systems working in underserved communities and learn how we can adapt the workflow to better fit their needs. Finally, we plan to work with global health providers to understand how we can modify our rubrics to meet their multicultural viewpoints about lifestyle. 

Nupur Garg is an MIT graduate with a double major in Chemical Engineering and Biology. She attended Yale Medical School and then went to Mount Sinai in residency in Emergency Medicine. She worked as an EM physician for over 10 years including through the pandemic before learning about Lifestyle Medicine. She fell in love with LM, wrote a book in it, contributed a chapter in another book, has been on podcasts about it, and teaches a course in it to other physicians. She also started her own clinic: Lifestyle and Family Medicine. 

Dr. Garg also has extensive experience in medical research. She has published peer-reviewed papers in informatics and has been invited to write editorials in journals as well. She has done several social entrepreneurship projects in the past, including global health telemdicine project while she was an MIT student. 

Misha Manulis is a full-stack programmer with over 20 years of development solution. He has built and sold an IoT solution for monitoring biotech lab spaces and equipment. He has also built a B2C device for (beer) homebrewing, BrewBit, raising $90,000 on Kickstarter. He is also the author of an IoT monthly newsletter, making IoT accessible to everyone. 

We have hired a team of contractors to do various tasks including managing our Lifestyle Medicine newsletter, AI programmer, front-end programmer, app developer, and other designers and programmers. 

Dr. Garg is a LM physician, actively using Propatient Tech in her practice. She is continuously helping to iterate the product to get it ready for product-market fit. 

We are applying to MIT Solve because we are interested in serving a large scope of people with this technology. There are certain areas that we have no expertise in delivering this workflow solution to the most underserved areas of the world. We are thinking of a full consumer solution that uses AI to help guide people completely autonomously or with some oversight. There are many other potential ways to extrapolate this technology to various other communities. We are applying to MIT Solve to seek that guidance. 

Innovation 1: Building in AI taxonomies for all data points from the beginning

Instead of just building another Electronic Health Record (EHRs), we are incorporating the latest technologies to design a platform that assists physicians and patients achieve their care goals with no limits on potential add-ons. One example is applying artificial intelligence (AI) models to the data we are collecting. Currently, EHRs are not built with the right framework to deploy AI models, which inhibits progress in patient care. For example, there are over 40 validated AI models and only one deployed in major health systems. By building this platform using AI taxonomies from the start, we are poised to incorporate AI models in the future as well.

Innovation 2: Integrations

We are incorporating as much wearable and bio-data as possible to deliver the most accurate information to the physician. The advice (output) is only as good as the input (patient data). If wearables or other integratible health data doesn't exist, then this platform is still fully functional. 

Innovation 3: LM data processing

There is no existing standard way to track and display information pertaining to LM visits, e.g. servings of various meal types, quantified sedentary vs. non-sedentary behavior, perspectives on life, etc. We are developing these concepts in conjunction with our platform.  

The impact of Lifestyle Medicine (LM) on an individual and community setting has been proven over and over through national and international studies, e.g. cancer incidence research, longevity research, nutrition research, and many other areas of research. There are several potential outcomes of applying our platform and widely distributing LM care, including increased awareness of lifestyle impacts on health, increased health tracking, improved health outcomes, and improved economic welfare. We are excited about this eventuality but cannot get there without the assistance of other community partners. 

We plan on conducting decentralized trials in each community to ensure the effectiveness of the language used for action planning. Furthermore, we recognize that each country can have more than one community. We will be cautious about overgeneralizing. AI will also be used to assist in this area. Through careful consideration and resourceful use of the latest technology, we can endeavor towards health equity for everyone. 

Improve the health of people across the world through basic lifestyle changes through AI generated health, culturally relevant recommendations. 

Measurable outcomes include any major predictor of health of a community. 

Every single lifestyle medicine recommendation is good for all aspects of one's health, and not just for the reason it was recommended. Some indicators can be: Alcohol Attributable Years of Life Lost, Maternal Mortality Ratio, Oral Health: Per capita availability of sugar, and more. 

We use extensive technology to develop this platform to allow for the workflow solutions that this platform provides. The most innovation technology that is used is called explainable artificial intelligence (XAI). XAI uses machine learning (ML) models that can be explained by experts in the field to make predictions. In this case, we are predicting the best possible lifestyle change recommendations among over 200 evidence-based recommendations to prevent, treat, and reverse chronic illness.

Full time: 0

Part-time: 2

Contractors: 5

We have been working on this solution for 1.33 years part-time. 

Early on in our company, we have set expectations to be a safe space for all employees. We are non-discriminatory to not only culture but also differently abled people. We are a safe space for autism spectrum disorder, especially.

We plan to sell B2B2C initially before launching our B2C that could directly benefit from Lifestyle Medicine (LM) care. 

There are over 5000 physicians who are certified in LM at this time that are struggling with the various workflow aspects of the care. We have been in touch with over 50 physicians so far, and they are all interested in our platform. Initially, we have a pricing model of 10/month for the first 100 patients and 5/month per each next 100 patients. As the features and interest grows, we can increase the price. 

We will expand to direct to consumer market after understanding the engagements from this initial patient base. From there, we will monetize various services. 

Through the B2B model, we will be financially sustainable. We will also have various ways to monetize the patient engagement on the B2C app. 

For the base product, most users are used to paying significantly more for what we are offering. They would be more likely to switch to us at a lower price point and thereby provide LM care in their practice. We will also be able to maintain flexibility in going up in pricing as needed since we are an order of magnitude less than typical EHR pricing.