Podcast: Play in new window | Download
A revolution is imminent in American healthcare, and “the revolution will not be televised” for passive observation. Value-based care transformation, like any other important movement, requires the active participation of all leaders on the frontlines. However, for these leaders to make the right decisions, they need to embrace innovation in order to realize the fullest potential of generative AI and predictive analytics. Through the reengineering of care delivery, we can achieve a more personalized, proactive, and efficient outcomes-based model that can ultimately transform population health.
As we navigate this transformative journey, data will play a pivotal role in reshaping the landscape of care delivery. And no one knows this better than Nassib Chamoun, Founder President & CEO of Health Data Analytics Institute (HDAI), our guest this week on Race to Value. In this episode, you will hear from a leader and primary inventor of a broad-based population health data analytics platform, enabling healthcare providers to make informed decisions based on real-time information. Tune in to an informative conversation covering such topics as data aggregation, predictive analytics, digital twinning, network management, generative AI in clinical care, and future advancements in technology-enabled value-based care.
Episode Bookmarks:
01:30 The Imminent “Big Data” Revolution in Value-Based Care
02:00 Introduction to Nassib Chamoun of Health Data Analytics Institute
03:00 As a teenager living in Beirut, Nassib experienced the horror of a civil war.
04:00 The inventor of Bispectral Index monitoring – a technology standard in operating rooms around the world.
05:00 Nassib discusses the pivotal moments in his life that shaped a passion for data analytics in healthcare.
07:00 80% of health information in EHRs is unstructured and entirely unusable unless converted to discrete data.
07:45 CMS provided HDAI a highly coveted Innovator’s License that allows the company access to data on 100 million Medicare beneficiaries.
09:00 How Big Data drives powerful AI algorithms and predictive models in healthcare.
10:00 “If you can’t measure something, you can’t improve it.”
11:00 Understanding the intersection between cost, outcomes, and utilization.
11:30 Making data actionable in order to effectuate change in care delivery.
11:45 Data overload can actually lead to clinical inefficiencies if it isn’t curated appropriately.
12:30 The artful curation of data to drive operational improvements at point-of-care.
14:00 The limitations of claims data in making timely clinical decisions and treatment interventions.
15:00 Interpretation of unstructured EHR data to extract potential new conditions and HCC coding opportunities.
16:00 The importance of clinical judgement in augmenting AI-based recommendations in value-based care.
17:00 Combining behavioral, psychosocial, and biometric data with the existing sciences of epidemiology and clinical medicine.
18:00 Generalized clinical use cases of AI at the point-of-care to improve costs, outcomes, and utilization.
19:00 “To be successful in value-based care, you must operationalize two separate goals: Prevention and Avoidance of Complications.”
20:30 “The goal of AI is to very simply do what a clinician does, but do it repeatedly and do it continuously for every patient in their cohort.”
21:00 How staffing limitations and an aging populations necessitates a more optimal use of technology in VBC.
22:00 In 2032, U.S. healthcare spending will reach $8 trillion (ahead of the economy of Japan) making it the third largest economy in the world!
22:45 Leveraging predictive models to drive more effective care coordination and interdisciplinary team-based care.
24:30 Patient engagement as one of the more challenging aspects of value-based care.
26:30 The integration of predictive analytics and digital twinning for individualized patient care.
28:45 Using multiple predictors to serve every component of the care team.
29:30 How the use of RAF scores to normalize utilization creates noise within a dataset.
30:30 Using digital twin matching for improved predictive modeling of patient outcomes.
32:30 Aggregation from the patient level as an opportunity to reduce predictive variance.
34:00 Benchmarking against clinical exemplars in population health management.
35:00 Empowering the care team through AI and predictive analytics to drive clinical interventions.
36:30 How Houston Methodist Coordinated Care leveraged the HDAI analytics platform to enable care teams.
38:30 Using AI as a core capability in a health system (for both FFS and value-based care).
40:00 Leveraging AI to make appropriate care management resource allocation decisions at the network level.
41:00 Using analytics to better understand post-discharge outcomes and mortality.
42:30 Early interventions driven by predictive analytics have reduced both hospital mortalities and readmissions.
44:00 AI insights to drive network management decisions in building a clinically integrated network.
45:00 With the explosion of ChatGPT, we are seeing Generative AI emerge as a technology offering transformative opportunities across society.
47:00 “A lot of the big tech companies have jumped into healthcare without the appreciation of what’s involved here.”
48:00 “There are three components of trustworthy AI in healthcare: Transparency, Explainability, and Reduction of Bias.”
49:00 Transparency
50:00 Explainability
51:30 Bias
52:30 Overcoming implicit bias in algorithms to reduce disparities in care.
53:30 The complexity of Generative AI algorithms.
54:00 HDAI’s approach to generative AI in the creation of transparent, explainable, and bias-free models.
56:30 Connecting the source code in generative AI to the recommendations made to clinicians.
58:30 Will we eventually see AI models enriched with crime data, geospatial analytics, food availability data, climate change impacts, consumer purchasing data, and biometrics?
60:00 The potential harm of applying negative use cases of AI to restrict care.
61:30 Arthur C. Clarke: “Any sufficiently advanced technology is indistinguishable from magic.”
61:45 The 4th Industrial Revolution of technology in healthcare.
63:00 EHR data has been largely underutilized until the recent tidal wave of AI.
64:00 Creating synergies within care teams on the basis of synthesized information and AI recommendations.
64:45 The application of AI in the future of health genomics.
65:30 Parting thoughts of optimism for the future of AI-enabled value-based care transformation.