The U.S. healthcare system is more reactive than proactive in promoting good health to its patients. Essential care is only delivered for chronic diseases after a critical illness strikes an individual and not before. The shift from a fee-based healthcare system as it currently exists today to a “value-based” or “outcome-based” care model is slowly happening. Teddy Cha and Hai Po Sun recognize this gradual change occurring and are working to accelerate proactive treatment of chronic diseases with their startup, pulseData. pulseData aggregates patient medical data, uses machine learning to predict who is most likely to experience chronic kidney disease (CKD) and proactively matches these high-risk patients with the necessary renal care needed. The healthcare startup raised $16.5M in their Series A from Bain Capital and Two Bear Capital leading the round.
"The future of healthcare will depend on leveraging advanced technologies that keep patients healthier and costs lower. pulseData is focused on having exactly this impact on renal disease, a chronic condition that affects millions of patients and costs the healthcare system over $100 billion a year," said Mike Goguen, Founder and Managing Partner of Two Bear Capital. "We're excited to continue our partnership with such a brilliant and mission-driven team of entrepreneurs."
Frederick Daso: What are the underlying incentives in America’s multi-trillion dollar, fee-based healthcare system that facilitate essential medical care only being delivered after a critical medical event?
Teddy Cha: Those that pay for healthcare - Insurers, employers, the government - pay for procedures and prescriptions. It does not matter whether the procedure makes a person better. The worse a person gets, the more catastrophic and costly the procedure, the higher the price. Kidney disease is a prime example - the industry reimburses $100k to $200k per year for a dialysis patient. Still, proactive care that would keep a person healthier and avoid disease progression is not paid for.
Expensive acute hospital events are lucrative, whereas chronic disease management and proactive primary care get comparatively paltry reimbursements.
Daso: What are the core challenges behind proactively identifying who needs medical care before serious illness occurs? I can imagine that medical privacy laws like HIPAA get in the way of accessing relevant patient information that could be the starting point for predictive analytics.
>The Fascinating History And Evolution Of Extended Reality (XR) – Covering AR, VR And MR
>Samsung CXL DRAM, Kioxia Enterprise/Data Center SSDs And Infineon Space Memory
>Facebook To Be Summonsed Before South Africa’s Parliament Over ‘Harmful Misinformation’
Cha: Healthcare data (Electronic Medical Records, insurance claims) are great at recording what gets paid for. It’s essentially a billing and receipts system. So the first challenge is building an inference map from billing receipts to actual clinical state or chronic disease progression. This requires human doctors and experienced care experts, teaching our machine learning engines what patterns to look for.
To keep patient data safe, we’ve built a proprietary solution, allowing our clients’ software platform to be deployed within their own secure, HIPAA-compliant environment. So patient data never has to leave the building. Our algorithms learn and predict without ever exposing this data.
Daso: Are you expecting an increase in tax expenditures for Medicare/Medicaid costs in treating CKD?
Cha: This is a blueprint for how to decrease Medicare/Medicaid costs substantially. The Centers for Medicare and Medicaid Services (CMS) 2021 budget is $1.2T, and about 20% of this is spent on someone with CKD or ESRD. Getting this spend under control is one of the key priorities for CMS. CMS is rolling out comprehensive reform in 2021, changing the payment model for kidney disease to incentivize better outcomes and lower cost. This is a blueprint for shifting expenditures for every cardiorenal chronic disease: paying for proactive and preventative care, improving lives while eliminating large avoidable outcomes and costs.
Daso: What factors do you weigh the most in your risk models to quantify who is most likely to need early medical intervention?
Cha: Kidney function is pretty easily measured with some commonly available blood and urine lab tests. Unfortunately, these tests are not part of the standard panel that most primary care physicians order, so a large majority of CKD sufferers are systematically undiagnosed or under-diagnosed.
One of the strengths of new technology is that we can use dozens or hundreds of factors to estimate risk precisely. When one factor doesn’t exist, we can compare the pattern or ‘fingerprint’ of this person’s medical history to thousands of other people that show a similar pattern. Often it’s not just one factor that’s the most predictive, it’s the velocity or trend of that factor, or the ratio of that factor against another, or this factor existing when another does not.
And the models will also get much better over time - many of the most important factors aren’t captured in the current medical record. A person’s social or economic context, the barriers to getting care, the ability to afford nutrition or access reliable health services, the availability of family and community support - these are critical predictors for disease progression but are mostly absent from any available health data.
Daso: How does your solution reshape the prevailing incentives in play to shift U.S. healthcare systems to be more proactive instead of reactive to an aging, chronically ill population? What common factor have you found among early adopters who believe in your view of “value-based” care?
Cha: The payment models and incentives are shifting dramatically already. CKD/ESRD payment models are the start, and already hundreds of billions in healthcare spend has now moved into value-based or outcome-based structures. So providers and payers all need a better counting machine - a way to accurately identify which people have what chronic diseases, and then match the right level of care to precisely the right population.
The common factor among the most successful and elegant adopters is a willingness to partner. Traditional healthcare providers need to be open to new solutions from data scientists and engineers. At the same time, technology innovators need to learn from the hands, feet and hearts that provide patient care.
Daso: What are some of the costs associated with acquiring the data to feed your models? How do these costs evolve at scale?
Cha: The good news is that health systems and payers have already paid to have their data captured and stored. They need to permit us to use it.
Over time the system learns to gather more data very efficiently - the riskier a person becomes, the more data we should be gathering, and the more ROI is available for gathering this data. We spend more to gather more data on riskier patients because we can save more on preventing otherwise expensive events.
I think this kind of system can net-save $1T from the U.S. healthcare spend, even after the costs of acquiring new data and of providing proactive, precise care.
Daso: How did you and your cofounder, Hai Po Sun, build out your initial team, given the need for specific areas of domain expertise? What would you say to those looking to make an impact in the space pulseData is disrupting?
Cha: We’ve been fortunate. Our team consists of outstanding technologists that we’ve worked with before, rare medical professionals who also speak python or can build data models, or domain experts who chafed at the status quo. Common across the team is a sense of mission - we can take what we’ve learned from other industries and make healthcare better.
There is a massive need in healthcare right now for great engineers, technologists, data scientists, statisticians, finance and economics experts. So your skills and experiences will make a huge impact! But only if you have enough humility to partner with and learn from those that are providing care. Healthcare happens when one human delivers an empathetic touch to another - the best solutions amplify this rather than replace it. Technology can help this happen more frequently and more precisely, connecting the right people at the right time.
Correction: Hai Po Sun’s last name is now correctly stated.
Source : https://www.forbes.com/sites/frederickdaso/2021/04/19/healthcare-startup-pulsedata-raises-165m-to-help-lower-costs-to-treat-kidney-disease/