Friday, October 30, 2009

The Public Option: A Systems Analysis Approach

The Public Option, PO, has been proposed and there has been no analytical study of what may be the market dynamics as a result of its introduction. We have commenced a study of this effect. Specifically we have constructed a model of the market dynamics of a PO and what the participation could look like under varying conditions. We will be presenting the results in a White Paper shortly. However give the intensity of the discussions we felt that and an early discussion was warranted.

The analysis consists of the dynamics of patients and providers in a PO and in a private plan, PP. Specifically we look at the dynamics over time of a PO compared to a PP. We assumed that persons and providers were either in a PO or a PP. That they moved back and forth over time due to observable from the prior time interval. These driver from one to the other were either attractants or repellents.

For the person selecting a plan the attractant is the price difference between a PO and a PP. The greater the difference the more the person went to the PO. On the other hand the repellent was the quality as reflected in patients per provider. The person looked at the difference in patient per provider in a PO and a PP and the greater the difference the greater the repellent effect. Thus as more people went to a PO due to price the more patients per provider resulted which in turn became a repellent.

For the provider there are just repellents. First is the reimbursement, which is plan dependent only and not dependent upon patients. Second is the load of patients per provider.

We combined these into a dynamic model of the type:

N(k+1) = a N(k)+Attractant(k)-Repellent(k), where N is the number of persons say in a PO

and

P(k+1)= b P(k) + Attractant(k) - Repellent(k), where P is the number of providers.

Calculations for attractants, repellents are negative attractants, are shown below:



















These are of course subject to change but they demonstrate the viability of the approach. Now using these values we can determine the percent of the people who will participate in a PO plan and we show its dynamics below:



















Note that this shows an increase and then a decrease. This effect if first driven by cost, lower costs of the plan, and then driven by the lower quality where subscribers leave the plan.

We now depict the patient participation.



















The following is the provider participation. In all analyses we have found low provider participation. This is similar to what we see in Medicaid and what we are beginning to see in Medicare. There is a growing refusal by many, especially the most competent physicians, to participate. If one wants good care one must pay.



















The patient per provider ratio has the most impact in our analyses. The numbers are shown below for this example.



















The revenue per patient is also a factor but seems to be secondary. We depict that below:



















We believe that detailed studies of this type are essential. We also believe that systems type analysis are not what the economists do. They fall into two camps. The macro type who conjure up slopes of trends, which are meaningless and fail to account for the dynamic factors or the econometric types who used old data to project new trends, which have the seeds of their own destruction already sown.

The major concern we have is the providers ability to opt out of a PO. I have not yet studied the massive 1990 page Bill but that is a concern.