Is Your Vision For Innovation In Health Technology A Reality, or Pixie Dust?

• I think we are mostly chasing “innovation” without full consideration of its impact and usefulness
• Impact analysis has to consider organization culture and readiness, as well as impact to others like  partners and the community (not just patients)
• Analysis of impact and effectiveness needs rigor and depth
• Lack of proper planning and analysis can disrupt and destroy value and quality, versus supporting and augmenting

Previously I started a discussion around four main areas of emerging healthcare related investment and influence. The four areas I think make sense are community health data collection and engagement, patient based care delivery and clinical services, operational support and data management services and in the financing and insurance aspect of the ecosystem. Note I am not discussing pharmaceutical or bio-medical investments. As a health actuary with nearly 30 years’ experience, I have a strong appreciation for how all this data being collected can be used, as consumer engagement and social determinant based information informs care delivery as much as product design and physician reimbursement / network contracting.  Fundamentally there has to be alignment and collaboration between doctor pay, health insurance product design and the funding of all this innovation and digital health based transformation. 

Of particular interest is the technology being developed that supports the operations of care delivery, as well as the technology that supports the financial analytics and insurance/ payer. This is a natural “hand-shake” and data sharing opportunity. Of course I also see a significant amount of very technical but incredibly important risks. 

Operational Support and Data Management

There are emerging tools target population health based, and genomic based, case management and transition management, and proactively identifying gaps in care. The excitement here comes from the juxtaposition of these type of tools with the advent of cloud-based technology to create data platforms that integrate data across EHR platforms. We can include the social determinant and community health based information collected from the consumer interaction technologies discussed above with real-time clinical data to create more powerful analytic support. Tools that could take full advantage of this structured data environment include predictive analytics and episodic grouper-like methodologies for case management and resource management across the care continuum, even attempting to personalize the pathways and care delivery with cognitive computing that adapts to patient preferences and socio-demographic circumstances.  

It is exciting to consider the analytics, predictive modeling and related infrastructure supporting proactive care management and shifting care delivery to more appropriate settings, ostensibly less intensive settings. With the excitement comes risk; this is perhaps the area with the greatest mathematical and actuarial complexity, requiring rigor and discipline in evaluating the embedded analytics and approaches of the emergent technology at a very granular level. There are already examples of failed multi-million dollar decision support system implementations that resulted in worse quality. A recent article on AI ontology (“Ontology to Oncology” by Siddharth Pai) is a great read around the deeper science required in “safe engineering” of these critical cognitive applications and care management tools.     

Financing and Insurance

Insurance company operations, and product design, are also benefiting from the rush to innovate. New companies and new products are being funded that encourage utilization of community based tools, providing rewards for engagement and care-delivery compliance. This potential disruption of the traditional insurance marketplace is also cause for some excitement. There is an emerging recognition that encounter or claims data is not enough to address the clinical transformation inherent in population health minded approaches, and is not necessarily timely or accurate/complete. For example, Medicaid patients that transition in and out of the program generate incomplete longitudinal records, and high-deductible commercial product designs may discourage full reporting of all health related activity. 

The insurer can play the role as both a financier and potentially a facilitator in the data collection and data sharing, especially socially determinant data. Insurers have a natural infrastructure and capacity for this development and investment. Their resources can be used to encourage innovation or investment in needed technologies/tools, as well as to assist in the identification of clinical inefficiencies through physician contracting, network design and configuration. The goal is practical - have access to the right data at the right time and in the right amount (avoiding overload and misdirection). This supports physicians in their goals of advancing outcomes-based improvements, assists with achieving rewards and incentives that track performance metrics, and advances the dialog to a more collaborative perspective regarding innovation with purpose and results. Physicians are already overloaded with various EHR-related data headaches, and two aspirin will not make that headache go away. A collaborative partnership would benefit all.

Innovative analytic tools and capabilities are also providing better support to care management and disease management programs, including proactive identification of target populations, even specific clinical co-morbidities. At the same time, there are Medicare / CMS changes in the data reporting standards to encounter-based data (EDS, which requires a more comprehensive data set than currently reported). Coordinating these efforts across the enterprise will be more efficient and optimize the sophistication of analytic support provided. This is another bridge between the insurer and physician, with the potential to further reinforce the shared responsibilities and alignment of incentives centered on patient care and population health.

The concerns in this space involve the diligence of financial and statistical analysis that supports the development of these complex outcomes based contracts and new product designs. Actuarial modeling and predictive analytics are needed that adjust the baseline data set to be on a consistent population base and time period with what is going to contracted for. Other adjustments are needed to reflect the full amount of costs to actually be incurred, and include adjustments for poly-chronic risks and other shifts in expected utilization patterns given benefit plan changes and contracting incentives. For stability and solvency purposes, rigorous analysis into clinical pathway deviations and “break points” that trigger reinsurance provisions or special case management considerations need to be modeled and understood by all parties as part of the collaborative relationship. Having the financial tools and comprehensive data today is a critical difference from the efforts at risk-based contracting in the past.  

Conclusion

Clinical and social data collection and analysis is interesting and can support numerous research studies. There is no doubt we have much to learn about what drives people’s behavior, the impact to their health, and what their care-based experience can be. From a practical point, though, the investment in the technology to achieve these insights is a use of financial resources that are required system wide. For example, the stakeholders impacted by an investment in new decision support tools include not just the physicians and their care delivery process, but the patient and the insurer / payer as well. 

Consideration of investments need to consider this ripple effect, and developers of the emerging technology should reflect their value propositions with this in mind. The impact and value analysis should reflect the care continuum, as well as the financial continuum, and the human continuum. We need to remain vigilant with the objectives of improved population health and community health, improved efficiency, and improved outcomes.   We need to look beyond, not on the near-term and obvious, but to better understand what perhaps has not been considered.  This will lead to better investment decisions which are also more impactful. 

I am looking to drive a change in the way we invest in and leverage emerging (and potentially transformative) digital health and population health technology. These emerging platforms should be leveraged from consumer engagement all the way through teaming and pathway decision making and workflow automation / clinical services delivery. And they need to be supported by designs and incentives in insurance products and public health policy. This change cannot wait.

This is a second in my series. The first can be reviewed here: "Digital Health is NOT Dead, maybe" https://www.linkedin.com/pulse/navigating-health-tech-risks-mark-jamilkowski

See aligned thinking https://lnkd.in/eQkWsi8 and https://lnkd.in/eAFenf2 by Ashish Atreja, MD, MPH, Mt Sinai @atreja

Also https://www.geekwire.com/sponsor-post/innovate-healthcare-fund-solution-problem-not-vendor/ by Aaron Martin, at PSJH

And Why Patient-Reported Outcomes Data is Key to Healthcare Quality https://shar.es/1VXsab @hitanalytics