Plan sponsors, and the benefit advisors supporting them, have sought for years to find the means to best strategize benefits design and offerings for plan participants. To be successful, they have needed to think critically along with using traditional, actuarial, and underwriting tools, as well as basic algorithms to determine both value and knowable outcomes. These values and outcomes are categorized by administrative excellence, regulatory compliance, clinical outcomes, financial effectiveness, member and provider experience, and the addressing of underlying inequities in the health care system.
For years, success in the design, deployment, and execution of effective programs and solutions has been limited by quality of available data, the sophistication of the data warehouses in which the data is contained, as well as the ability of the owners of the data to discern, provide, and enact actionable outcomes. This has resulted in a limited ability to define and measure success of various programs and solutions; both plan sponsors and members have been ill-served by these limitations. This has also resulted in plan sponsors, and payers creating complicated ecosystems that are based upon the hope that an unproven series of point solutions will solve for a given set of supposed or apparent clinical and financial issues. This in turn has led to member and provider confusion and abrasion, resulting from a lack of integration and a movement away from "whole health."
Plan sponsors, and those advising them, have historically been fundamentally dependent on the payers for the quality and breadth of the data, as well as the accompanying analytics and reporting needed to determine necessary proof points. The payer community operates almost entirely from aged and/or antiquated administrative, adjudication, and clinical platforms. The resulting analytics are often based upon flawed, inaccurate data. While most payers have made significant investments around interoperability (the bi-directional flow of content and data between them and providers) and data warehousing, there remains significant room for improvement.
Several benefit advisors have begun the process of investing in their own data warehousing and analytics initiatives in order to better serve their clients' needs. As more advisors assume this functionality, they need to determine whether to build, buy, or rent the various pieces required to meet emerging market requirements. Success needs to be based on an understanding of the stakeholders whom the plan sponsor's benefits specialists serve, which typically include business owners, finance, labor relations, and plan participants. Creating reporting and analytics to meet the needs of these various constituents is a critical success factor. Benefit advisors need to provide informed, integrated, and proactive insights, as payers are only capable of showing their piece of the overall picture.
The emerging market
The changing marketplace requires the addition of data scientists and engineers, epidemiologists, masters/doctorates in public health, behavioral scientists, health care economists, as well as software coders and engineers. This increased expertise is essential to attaining successful alignment and accountability between plan sponsors and their vendor partners, as well as between plan sponsors and their internal constituents.
The magnitude of pursuing the emerging technological needs can only be adequately addressed by using a comprehensive, methodical, and transformative approach, with data science and analytics as a core pillar supporting meaningful growth in the industry. For instance, a need for a transformative approach is requisite for headwinds inclusive of a medical cost trend likely to land at nearly 7% in 2024 for individual and group markets, with pharmacy trend at around 10%. These increases are attributable to many factors, including staggering disruptions in the health care environment, which is still recovering from the COVID-19 Public Health Emergency (PHE). Other health care disruptions include staffing shortages and burnout, vertical and horizontal consolidation of health care delivery systems, increasing costs for both existing and new drugs including cell and gene therapies, as well as weight loss medications, to name a few.
Key data science considerations and definitions
Not all data science, analytics, and underlying platforms are the same. Some are more sophisticated and complex than others. All too rarely do many of these tools ingest, cleanse, integrate and enhance a broad range of structured and unstructured data sets. For instance, lab results down to the member level are rarely linked to other key data elements, thus limiting the ability to accurately discern morbidity, risk, and adherence. Not enough of the available platforms adequately incorporate data such as electronic medical records (EMRs) and non-traditional (consumer based) data sets that enable the development and inclusion of behavioral science modeling. Such models enable professionals to determine the optimal means of engaging both members/patients and providers to improve their experience and the resulting link to clinical outcomes.
How do we pursue more meaningful data science and analytics in the health benefits consulting realm? Start by encouraging benefit consultant decision makers to have their own database, constructed within a deep learning platform, inclusive of predictive modeling and sustainable artificial intelligence (AI) with cloud computing strategies.
AI, machine learning (ML), and natural language processing (NLP) are now being leveraged in multiple capacities in order to realize operational efficiencies, predict adverse health outcomes, train medical professionals, and even augment the pharmaceutical discovery and development process. Advanced technology can revolutionize the ways in which health care is delivered and received, as well as administered and managed. For example, ML algorithms can quickly process large amounts of clinical documentation, identify patterns, and make predictions about medical outcomes with precision and accuracy. Additionally, AI, ML, and NLP have begun to solve operational gaps in care delivery by standardizing processes such as patient intake, triage, and translating provider notes.
Optimal data warehouse and platform
Benefits advisors need a distinct set of algorithmic models and tools to bring value to their plan sponsors and stakeholders, even if this means they must rent the technology. The right human capital must be sought to manage, orchestrate, streamline, and interpret the data into actionable insights. These insights include, but are not limited to, a broad range of market dynamics, as well as how those dynamics drive and/or reflect each client's own experiences.
Population health approaches inclusive of risk adjustment and stratification are essential, so that clear patient cohorts are identified and the closure of gaps in care can occur. These capabilities enable the formation of personalized care. This includes identified gaps in care, transitions in care, practice, billing, referral, admitting, and prescribing patterns on the physician level, as well as complication rates at the patient level. Such work is accomplished through evaluation of lengthy longitudinal views of the physician and patient, providing deeper and more accurate insights around emerging individual risks, overall risk stratification, and accompanying opportunities for interception impact.
Predictive modeling promotes earlier and more thorough identification of provider behaviors and patient health events. There are numerous advantages to gaining these understandings, including promotion of a preventive care framework, assessment of a given provider's ability to provide quality care, and understanding of the patient's likely clinical journey with the opportunity to intercede at critical junctures. The use of a strong AI-based platform enables plan sponsors and their consultants to share needed insights with the various vendor partners serving the covered population. This should lead to improved collaboration, targeted and attainable performance standards, as well as continuous quality improvement in the areas of administration, compliance, finance, and clinical care management.
The ability to work together with a shared and accurate knowledge set enables the plan sponsor, consultant, and vendor partners to find and determine return on investment (ROI) for the various programs and solutions presently being utilized, as well as those being contemplated for implementation.
Conclusion
A robust generative AI platform with the needed underpinnings around data and human capital enables benefit consultants to provide new and deeper answers and insights into the challenges the industry has faced as we have endeavored to both offset and manage continuously high cost of care and annual trends. These platforms increasingly allow us to redefine and determine critical success factors/key decision points, including cost and quality of care, high performance providers, efficacy of clinical care and management, a 360-degree view of the member's administrative and clinical experience, appropriateness, adherence, and persistence to prescribed treatment plans, assessment, management, and abatement of administrative, compliance, clinical, and financial risks, as well as specific and overall vendor performance levels. Failure to include or adopt these tools and approaches will leave clients under-served and consultants on the outside looking in.
Stuart Piltch is managing director of Risk Strategies Consulting.
© Touchpoint Markets, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to TMSalesOperations@arc-network.com. For more information visit Asset & Logo Licensing.