I had an interesting conversation with a plan advisor last week. We were discussing the drug plan experience of two different plan sponsors. The timing of our discussion was interesting because both of his clients in question had just begun a new benefits year on January 1. From what I understood, the renewal process for these clients had gone relatively well, and it was unlikely there would be much activity on these accounts until later in the year when the 2013 renewal process would begin to heat up. The main purpose of our conversation wasn’t to look ahead but for me to answer a few technical questions related to specific drug therapies that had surfaced.
Before wrapping up the conversation, I asked, “Looking ahead to the 2013 renewal for these clients, what does the claims experience from 2010 and 2011 tell you about what you are going to see over the next 24 months?” I’m still awaiting an answer, and I don’t think one is forthcoming. That’s unfortunate because this is a shining example of the wasted opportunity many plan sponsors and advisors incur year after year—they are not making use of the predictive capability of the claims data that make up a plan’s experience.
By not using these next few months to proactively evaluate their experiences—current and future—to ensure no surprises for the 2013 renewal and beyond, these plans run a high risk of not acting until their backs are against the wall, with any emerging issues already becoming part of the experience. Looking ahead, if there are no immediate concerns within the plan experience, that is great news—it will make the renewal work later in the year that much more straightforward. However, if there are key indicators in plan utilization, population health or plan costs that cause concern, there would be six to nine months of time to assess plan options to deal with these issues that sit just below the surface.
The most recent 24 months of data is one of the most important pieces of information that plan sponsors and advisors have at their disposal to manage employee health benefits. This data can help them to understand not only what happened in the past but also what is going to happen in the next 24 months. The problem is that the default for most plans attempting to make decisions on a go-forward basis is to consider only high-level aggregate data each year that is retrospective in nature.
Here are some of the limitations in using only summary-level retrospective claims data as opposed to more granular transactional-level data.
Trying to assess specifics
Summary data (i.e., Top 50 or Top 100 drugs reports by amount paid or number of paid claims, top therapeutic classifications and/or claims frequency reports) do not help to explain how the current demographic profile correlates to its disease state profile, and how that profile has trended over the last 24 to 36 months, because the claims are rolled together.
Failing to determine saturation rates
Simply having a list of top general disease states does not provide the saturation rate of age-related chronic conditions in a plan experience—a key cost driver for many groups. So what if “cardiovascular” is your biggest category? The question that needs to be answered is whether or not the plan experience within that area is already fully saturated, or whether we can expect significant growth in the coming 24 months.
Not accounting for speciality drugs
A top drugs report will tell you whether any high-cost specialty drugs exist within the plan experience, but it won’t tell you the current saturation rate of specialty drugs within key areas such as rheumatoid arthritis, multiple sclerosis and Crohn’s disease. We seem to quickly lose sight that more than 80% of plan spending for most groups takes place within traditional therapies, and that by better managing traditional claims, we can free up significant resources for funding existing and future specialty claims. Plans that are under-represented in key specialty areas based on their demographic profile need to prioritize managing non-specialty plan spending.
Trying to make trend predictions
Looking at high-level spending and utilization trends is of little value in understanding what is to come over the next 24 months when we haven’t accounted for changes in demographic and geographic variables, including the number of unique claimants, the impact of public/private sector co-ordination of benefits, the ratio of acute to intermittent to chronic therapies and the average age of the plan member at the time of the claim.
Not assessing impact on experience
If there is no understanding of what impact the current plan design is having on the plan experience, how can we assess if it’s viable for the next 24 to 36 months? The impact of the current plan design does not manifest itself in high-level utilization figures. Transactional data, on the other hand, allow a plan sponsor to answer this question and can also be used to calculate which alternate plan design will have the greatest impact on managing plan costs without adversely impacting the health of employees.
Fearing pay-direct
There continues to be some incredible mythology in the group benefits space suggesting that a move to a pay-direct drug plan will result in significantly higher plan costs. If that fear of a sudden, significant and sustained increase exists, why not simply measure the “shoeboxing” effect and carry-forward of claims and compare them with the savings to be realized with real-time claims adjudication? Would that not make more sense than retelling the group benefits version of the Loch Ness monster tale? This is of considerable importance for Ontario-based plan sponsors that are about to enter the Twilight Zone as generic prices fall to the floor on April 1, 2012, for both the public and private sectors.
Failing to look back on the short and long terms
If we haven’t looked at the short-term and/or long-term disability experiences and reflected on the drug plan experience, why not? Why would we not use disability data and granular drug claims data to assess patterns in the drug therapies related to the biggest health concerns in the employee population? If there are leading indicators in the drug claims experience (which is far more prevalent and robust than disability data)—such as changes in the incidence of a given disease state, changes to levels of adherence to therapy and/or changes in the co-morbidity profile of disability claimants—a plan sponsor can begin to measure the impact the drug plan is having on spending elsewhere and how that impact can be enhanced.
It is encouraging to see that better reporting and better use of data is becoming more of a focal point. I understand that the CPBI Ontario conference next month in Toronto is devoting a session to the topic. That’s a positive step because it’s unlikely that we would have been anywhere near the agenda even a couple of years ago. We’re starting to see some progress, but plan sponsors across the country would benefit immediately if using data for predictive purposes became the norm, not simply the exception.