Healthcare Innovations Roundup
A newsletter about building, evaluating and implementing digital innovations in healthcare (1-22-24)
Hi Everybody,
Welcome to the latest issue of our newsletter. Here what we have in store for you this week:
This issue at a glance:
Prediction of atrial fibrillation from at-home single-lead ECG signals
A website for GenAI tools for healthcare
Coalition for Health AI (CHAI), an organization that is leading AI assurance, technology, and policy initiatives for healthcare AI is backed by major healthcare systems
2023 year-end digital health funding
Estimated Savings From the Medicare Shared Savings Program (spoiler alert, it is not that much!)
Apple watch pulse oximetry saga (lots of drama and a bad outcome if you are Apple)
Enjoy,
Science Roundup
Academic Papers
Prediction of atrial fibrillation from at-home single-lead ECG signals without arrhythmias
Goal:
The goal was to develop a model for identifying atrial fibrillation (AF) within a two-week period, using data from non-AF single-lead ECG recordings that vary in length from 10 minutes to 24 hours.
How did they do this?
They used 459,889 single-lead ECG recordings of length up to two weeks and developed a model to estimate the risk of presenting AF in a two-week observation period by analyzing only non-AF ECG intervals of various length. In this context, AF risk is defined as the predicted probability of the occurrence of an AF event, as informed by the ECG data. The model integrated ECG morphology data with demographic and heart rhythm features using deep learning techniques. The performance of the model was evaluated using different input configurations and monitoring lengths
What did they find ?
The model with deep learning features produced the most accurate prediction. The model achieved high accuracy with an AUC of 0.80, outperforming other models that relied only on demographic and manual features. The accuracy of predictions was found to be influenced by the duration of the monitoring period. Longer monitoring periods led to more accurate predictions, suggesting that certain AF-related features may not be detectable in shorter intervals.
Why this matters:
This has the potential to improve detection of AF by identifying individuals with who would benefit from prolonged (>= 2 weeks) of monitoring with novel devices such as wrist worn wearable devices.
Blogs
Explore all the GenAI tools for healthcare
This is a great site. Board of innovation and Novartis have curate relevant tools that you can apply to your clinical practice today. I already use of the tools listed in my practice.
Backed by Mayo Clinic and Microsoft, a nonprofit forms to test AI tools used in health care
More than 1200 organizations across private and public sectors have joined the Coalition for Health AI (CHAI), an organization that is leading AI assurance, technology, and policy initiatives for healthcare AI
Industry Roundup
2023 year-end digital health funding: Break on through to the other side
📉 The year 2023 ended with a focus on survival, setting the stage for a pivotal 2024 in digital health funding and market dynamics.
📉 2023 saw a significant dip in digital health funding, with the lowest investment since 2019 at $10.7B across 492 deals.
💡 Despite tough market conditions, there was no surge in M&A activity or startup shutdowns.
💡Many companies adopted creative financing like extension rounds and silent deals to survive.
🚀 2024 may force hard choices for startups, potentially leading to a stronger, more consolidated sector.
Healthcare Economics and Policy Roundup
Estimated Savings From the Medicare Shared Savings Program
What is MSSP ?
The Medicare Shared Savings Program (MSSP) was launched in 2012 to improve efficiency and generate financial savings for the Centers for Medicare & Medicaid Services (CMS). Under the program, voluntarily constituted accountable care organizations (ACOs)—groups of clinicians, hospitals, and other institutional participants—face accountability for the total costs of care received by traditional Medicare beneficiaries. If medical spending is below a specific target (benchmark), ACOs are eligible for financial bonuses. For CMS to break even or achieve net savings for traditional Medicare beneficiaries in the MSSP, gross reductions in medical spending must equal or exceed the sum of bonus payments paid to ACOs. Reductions in medical spending may also spill over to Medicare Advantage (MA) because MA benchmarks are based on spending in traditional Medicare.
What are factors that influence MSSP on net savings for CMS?
Gross reductions in medical spending appear to be concentrated among physician-only ACOs and not ACOs affiliated with a hospital.
Savings to CMS depend on how MSSP benchmarks are set: benchmarks that are too easy to achieve will result in bonus payments that are too high relative to ACO performance.
For the MSSP to spill over to MA, the MSSP must only generate gross savings, not savings net of incentive payments. This is because benchmarks in MA depend on projected national per capita traditional Medicare spending (known as the US Per Capita Cost [USPCC]). Regardless of bonus payments paid to ACOs, if the MSSP reduces gross traditional Medicare spending, it will be reflected in the USPCC and reduce future MA payment
Question ?
Was the Medicare Shared Savings Program (MSSP) associated with net savings to the Centers for Medicare & Medicaid Services (CMS)?
What did they find?
Summing net losses from traditional Medicare and savings from MA, the MSSP was associated with savings of $4.339 billion based on derived estimates from McWilliams et al and $3.057 billion based on MedPAC-derived estimates. This represents approximately 0.075% of combined medical spending related to traditional Medicare and MA over the study period.
Regulatory Roundup
Apple to Pause Selling New Versions of Its Watch After Losing Patent Dispute but then an appeals court granted Apple’s emergency motion to pause the U.S. International Trade Commission’s ban on sales of its watches with a pulse oximetry feature. Apple introduced the blood oxygen sensor and app in 2020 in its Series 6 watches but at the end it looks like apple will remove pulse oximetry from Apple Watch Series 9 and Watch Ultra 2 to comply with the ruling by the International Trade Commission
Thank you again for reading my newsletter.
If you enjoy reading this newsletter, please share it with someone else who might enjoy reading it.
Talk soon,
Hamid