The skinniest Exchange narrow networks from 2014 won’t be as skinny for the 2015 open enrollment period. There are three major drivers.
The first is a federal regulatory driver. The 2014 regulation required the minimally acceptable network to contract with 20% of the Essential Community Providers in a service area. An ECP is a center that serves underserved populations. Federal primary care clinics, family planning, AIDS care clinics are the most common ECPs. The 2015 regulation is 30% of the ECPS in a service area must be in network. For most networks, this is a fairly minor tweak, but the skinniest networks will need to add facilities and docs to meet requirements.
The second is that providers are starting to see that the Exchanges are here to stay and that it is better to get Medicare +8% from an Exchange product than who the hell knows what from the spare cash flow of an uninsured individual. Mayhew Insurance has seen a steady stream of inquiries of providers who want into the Exchange networks. They are coming in either to help one or two current patients who switched their insurance, or they are seeing that the Exchanges will improve their bottom lines. The narrowest and usually the lowest paying networks will probably see the highest percentage increase in provider participation in 2015 as the base level of participation for the super-skinnies would be very low regionally.
The final reason why networks are getting larger is amazingly technical and geeky. The speed assumption in one of the major geo-access/GIS software packages changed.
Now what is this and why is it important?
This is important because of how state regulators measure accessibility adequacy is based on some massive assumptions. My state requires each network and product combination to offer adequate access in twenty four categories of providers. Urban county residents are expected to be able to see at least two in-network primary care physicians within 20 45 driving minutes. Rural county residents should see two PCPS within forty five driving minutes. People should see one neurological surgeon within 2o or 45 driving minutes based on county type. Different states have a different number of provider categories and definitions of “adequate.”
None of the software packages perfom a point to point, turn by turn actual roadway comparison between every single person who lives in a service area and every single provider. That would be an amazingly nasty optimization problem that requires much more powerful computers than most insurers want to buy. Instead, the software packages make a series of assumptions about the location of each individual within a zip code, the location of each provider by category within a zip code and then imputes a straight line distance calcuation with an average speed.
There are several major flaws in this approach. Trips have highly variable drive times on a very predictable cycles. The regional Level III trauma center is 6 minutes from my house at 6:15AM on Saturday mornings, and forty five minutes from my house at 7:22AM on Wednesday mornings. The software assumes it is 12 minutes for accessibility purposes in all cases. Not all rural, suburban or urban zip codes are created equally. One Manhattan mile is treated the same as one Gary, Indiana road mile.
The change between the 2013 and 2014 speeds is significant.
Urban assumed speed went from 35mph to 25mph, suburban went from 45 mph to 40 mph and rural speeds dropped from 65 mph to 55 mph. I think these assumptions are more in line with reality, but they arebig changes. In an urban county, the coverage radii of a single provider dropped from 11.5 miles to 8.25 miles, for a service area reduction of almost 50%. In rural areas, the radius dropped from 48 miles to 41 miles for a 30% drop in service area.
This is a major change for the skinniest networks that are not concentrated in urban counties. Urban counties tend to have provider density where the minimal number of providers in the network is significantly above any filing guideline. However, skinny networks in suburban and rural areas will need to either recruit new specialists to provider coverage to outlying/fringe counties, or drop hard to cover counties or write extensive narratives explaining to their state regulators that under the old speed assumptions, the current network is sufficient, recruitment is ongoing, but the current speed assumptions have created new gaps without any changes in the approved provider listing.
Companies will do all three.