Health payers have the hard job of financially protecting the insured for the duration of the plan. This in turn requires payers to be astute at predicting the future, a challenging job in the best of times. Of late, payer business conditions have been especially volatile, with under-appreciated risks and greater than expected medical and pharmacy costs, putting the payer industry through some of its worst years in recent memory.
But if conditions of greater uncertainty and volatility persist in healthcare, what are payers to do? Is there a next generation of forecasting capabilities that could support the payer industry in the coming years? We begin exploring these questions in this blog across three main categories: revenues, medical costs, and administrative costs.
Revenues – Resilience to market volatility (aka “where will the money come from”)
Payers have lately been buffeted by event-driven shifts in insurance risk pools. Events such as the expiration of the enhanced advanced premium tax credits for exchange plans and exits of some players from the Medicare Advantage (MA) have moved pools of risk around the market, significantly affecting some payers with unanticipated high-risk pools. Even with a (still) relatively generous risk adjustment mechanism in MA, these effects can be hard for payers to manage.
But what if we could be more anticipatory about future risk pools? Could we develop analytically driven insights into drivers of member mix and their risk profiles, including accounting for how changes in a payer’s product portfolio and competitors’ product portfolios interact in impacting the incoming risk pool? Payers would need to develop and continuously hone these models based on a combination of quantitative market research techniques like conjoint analyses, real-world data, and synthetic data from wargaming type exercises, given the game theoretical nature of competitive product portfolio strategies. Developing scenarios would help prepare payers to appreciate what the incoming risk pool could look like probabilistically. Over time, artificial intelligence (AI) based continuous learning systems could drive this process, with an interdisciplinary team of strategists, actuaries, product, and research specialists.
Another area where foresight would be beneficial is the trajectory of the employer/group health insurance market. Given the gradual but persistent decline in the small group market, and the introduction of radically new and potentially discontinuous coverage arrangements like ICHRA, payers need to develop forecasting muscle to navigate the evolution of the employer market.
Medical costs – Embracing new data sources and looking farther out
This is the big one. There are so many potential applications and nuances here. One example on the unit cost front is embracing provider and payer transparency data. While not perfect, analyzing transparency data (potentially in conjunction with traditional claims-based data) is increasingly becoming table stakes for both payers and providers. The data offer more granular insights on unit prices by entity and geography, which can better prepare parties for rate negotiations, product pricing, and responsiveness to employer fiduciary role inquiries.
Another example is embracing demand and financial data from providers and incorporating that into payer strategies. Sg2’s Impact of Change can provide detailed multi-year forecasts on demand in local geographic areas by type of procedure/service. The Sg2 forecast captures innovations in care delivery (e.g., GLP-1s) which will impact demand across services in varying ways – to the positive and negative. In an environment where novel therapeutics are emerging more quickly, capturing demand impact on payers is critical for more accurate forecasting. Further, analyzing health system financial statements in detail can provide new insights that allow payer executives to better understand the health system’s financial situation and anticipate their likely strategies, all of which can better prepare payers to strategize for their relationships and negotiations with health systems.
Furthermore, earliest detection of utilization trends is critical to managing the plan to target profitability. This requires payers to develop better early detection analytic techniques as well as access to new, richer data sources that do not have the natural time lag of claims data. For example, access to EMR data from partnering providers and/or real-time patient communications have the potential to offer earlier insights into notable shifts in medical utilization trends. Earlier detection of fraud would also be valuable to payers, and could be a collaborative effort with other payers and/or risk-bearing provider partners, given the shared challenge of fraud detection.
Further, broader economic conditions by geography could impact the consumer share of wallet directed towards health spend, an important consideration given the ongoing shift to consumer/patient paid services. How can payers leverage economic indicators to better predict healthcare spend?
Overall, the goal of payers should be to develop a deeper understanding of historical trends as well as both in-year and multi-year implications and forecasts. The latter would be especially helpful if CMS introduces multi-year health plans, as it has alluded to recently.
Administrative costs – Flexibly adjusting to scale and mix shifts
The greater degree of volatility in the payer business has increased the urgency to develop a more flexible administrative cost structure – one that can flex up and down with both operational and care delivery volume, quickly. That way, when the next big shift occurs (e.g., potential small group conversion to ICHRA, new therapies/diagnostics), payers are better prepared to respond with a more flexible administrative platform.
Activity-based costing (ABC) is critical here and payers need to build a deeper understanding of workload drivers and fixed versus variable costs by key functional areas. While ABC has historically been notoriously hard to maintain, advanced analytics/AI may be the enabler to maintaining these tools, including helping focus payer leaders’ attention on what really matters amid a lot of data generated through these platforms.
This will be a journey and a “product roadmap” could be established to drive progress here. This kind of insight will enable payers to start developing flex-up and flex-down scenarios for operations and resulting admin costs in response to shifts in market demand. Early detection of demand trends linked with deep insights into admin cost drivers should enable payers to drive operations more in sync with demand, a growing benefit in an increasingly margin-constrained business environment.
And of course, culture eats everything for breakfast
The bigger aspiration that weaves together these ideas is building payer organizations that become significantly better at sensing, forecasting, and adapting day-to-day operations accordingly. This is a transformative shift in how many payers operate.
As in most transformations, cultural aspects will be key to success. For example, interdisciplinary teams will become mission-critical and will need to transcend old functional loyalties – the mission will need to supersede the function. Also, payers are understandably highly risk-aware organizations that take pains to understand and manage risk. This is essential. However, sometimes this careful approach can inadvertently bleed into other areas, leading to greater precision and consensus behaviors than necessary. Speed is sacrificed in the process, robbing payers of the opportunity to respond with alacrity to changing business conditions.
While many challenges hover on the horizon for payers, they have opportunities to transform their forecasting to be more proactive, more interdisciplinary, and more farsighted. In a volatile world, prescience is precious.
The author would like to thank John Poziemski, Erik Swanson, and Brian Esser for their insights and comments on this blog.
