Something interesting happened this month as we examined the 2018 experience of two different plan sponsors in two separate geographies with different demographics. The only similarity between the plans is they both have 80,000 or more individual claimants on their drug plan annually.
At the same time last year, we completed a series of predictive analytics on each of these plans. In the first case, we predicted it would see a one per cent decrease in plan spending. That’s not common for a group that’s been growing in terms of plan member population year over year, so it was likely met with some skepticism. As it turns out, plan costs decreased by 3.1 per cent year over year. In the second case, we predicted a 3.7 per cent increase in spending. The figure came in at 4.2 per cent.
As we begin 2019, it’s surprising how many administrative-services only plans are still forecasting plan spending using historical numbers and/or well-established actuarial formulas that don’t leverage all the information current plan experience is telling us.
In health benefits, like drug plans, the past is a poor indicator of the future. A group’s demographic and disease state severity profile changes, but more importantly, the cost drivers change significantly. This must be accounted for, as does the geographical distribution of the plan — given the significant coordination of benefits differences existing region to region — and the plan’s current design.
There’s a good reason why any industry stakeholder tasked with predicting plan or block spending in the next year missed accounting for hepatitis C claims — it’s hard to see what’s ahead if you’re only looking in the rearview mirror. In 2013, hepatitis therapies accounted for 0.6 per cent of plan spending for active plans and 0.2 per cent for retiree plans. By 2015, those figures were 3.3 per cent and 1.9 per cent respectively.
A plan may be looking to predict their claims’ experience for the end of 2019, 2020 and/or 2021 for a number of reasons, such as assisting in strategic planning around plan design, trying to value its post-employment liabilities or looking at its current standing relative to its future obligations.
Here are some elements to consider when examining the existing experience:
- How many new chronic specialty drug claimants will emerge in the next 12, 24 and 36 months and in which disease-state areas? That can be assessed by examining the group’s current disease state severity profile.
- What will be the impact of the near-term drug pipeline on the plan? One simple example that’s relevant today is the new specialty therapy for migraine prevention. A plan can assess how many individuals would clinically qualify based on the current profile of migraine claimants.
- How many new diabetic patients will emerge in each of the next three years and how will that impact treatment costs within the plan?
- How saturated is the plan with other individual age-related chronic conditions based on it’s existing disease state and demographic profile? How will under-saturation in certain areas manifest in the next 12 months?
- What will be the impact of increasing or decreasing the coordination of benefits with public plans given the geography of members, existing therapy, age and formulary status? How will the expected movement of claimants into and out of the plan impact that number?
- What will be the impact in 2019 of the new generic drug prices introduced in April 2018?
- In Ontario, what will be the impact of adding OHIP+ claims back in that would have disappeared from the experience in 2018 but are expected to resurface in plans before the end of the first quarter of 2019?
As this isn’t an exhaustive list, there are certainly other variables that can be added into the predictive factors that need assessing for a given group. But it does highlight how rich plan data is and what kind of asset it can be from a planning perspective.