A large number of providers in a network does not make a network good.
Few providers and a handful of hospitals does not make a network bad.
A good network has a sufficient number of providers in a wide variety specialties so that members can quickly get appropriate and high quality care at locations that are reasonably close to the member and in a reasonably short time period. In my region, a good network could have 10 hospitals and 900 MD/DO providers in it. I would be happy buying that coverage for my family. A really bad network could have 25 hospitals in it and 3,500 MD/DOs and I would pay significant sums to avoid that product.
Most of the APTC hacking that I’ve been working on has been manipulations of the cost sharing elements of plan design. Those are the easiest elements to do as any idiot can run tweaks through the actuarial value calculator and come up with “unique” isomorphic clones to capture the low price Silver region. Manipulating networks takes a little more time and forethought but it could be done as another APTC hack.
There are a couple of different ways to create strategically bad networks.
Here is how it works.
A base network would be generated. It would be a narrow network that meets minimum state regulatory filing standards. A network ID would be assigned to this list of providers. The network would be attached to a plan design (HMO/EPO/PPO) and then a cost sharing structure. If one wishes to be a cynical bastard, the cost sharing structure would be one of the standard configurations suggested by CMS.
So far so good, this is the standard plan creation process.
The hack is to take the network that has been approved by the state and add one more doc. That doc might be three counties over from where the network will be sold. That does not matter. The network is now unique and it gets its own network ID. The new network would be attached to the same elements that the original network has. It would get a nifty focused group tested name to differentiate itself. The new clone will price out the same as one incremental provider that is out of the selling region will have no impact on utilization or net unit price.
Voila, another approach to “meaningful difference” that is neither meaningful nor different.
That is the simple way of hacking the exchanges via the network.
Feathers
Not the sort of hack you are talking about, but a good example of small networks not being bad. No idea if this is the current case, but when I worked at Harvard, I recommended to a woman who had just gotten hired that she take the Harvard Employees Health Plan. She said that she would never do that!!! She had talked with her mother and there were only two hospitals in the network!!! Who could be that foolish!!!
Those hospitals were Mass General and Beth Israel. Two of the top three in that year’s USNews best hospitals in the country polls. She didn’t last long.