Structuring decisions: Should electromagnetic hypersensitivity be acknowledged as a medical condition caused by electric fields?

I examine the question whether Electromagnetic hypersensitivity (EHS) should be acknowledged as a medical condition caused by electric fields (EFs) or not.

Figure 1 shows my attempt to structure my thinking about this topic using a decision tree technique. In rest of the post, I explain my thoughts about this decision tree. This type of exercise can help to clarifying thinking, to guide collection of facts and to help at communicating the decision. You can find conclusions at the end of this post.

Figure 1: Decision tree about whether EHS should be acknowledged as medical condition or not. Squares are decision nodes, in which a decision can be taken. Circles are chance nodes where one of the states will be realized. Triangles are consequence nodes.

 

Firstly, the level of detail included in a model should depend on the purpose of the model. Less detail can make it easier to get the overall picture whereas more detail can lead to more accurate reasoning. Here I do not go to very detailed analysis and do not seriously attempt to estimate the parameters in the model. Second, I remark that I am not an expert in this particular topic but (at least trying to be) an expert in decision making.

With these disclaimers, I can start explaining my thoughts :)
(If you are uncomfortable with symbols and equations, you can safely skip those parts and go to conclusions.)

I identify two possible scenarios, or, states of the world:
S1: The symptoms associated with EHS are caused by electromagnetic fields (EFs).
S2: The symptoms associated with EHS are not caused by electromagnetic fields.

I identify two time steps and related decision opportunities:
D1: Whether EHS is initially acknowledged as medical condition.
D2: Whether the decision is reversed after significant amount of new information becomes available. I assume that if the correct decision is made initially, then decision is not reversed. Otherwise, reversal is attempted.

These lead to four possible consequences: C1=C(Ackn,S1), C2=C(Acnk,S2), C3=C(Not, S1) and C4=C(Not, S2). Note that C2 and C3 additionally depend on whether the attempt to reverse the decision succeeds or not.

At least, the most important consequence should be taken into account. I consider these to include, at least, the following: Number of people with symptoms (N), level of symptoms (L) and costs (C). Number of people with symptoms and level of symptoms together characterize the amount of pain caused by EHS. Costs include money spent on treatment and prevention of EHS symptoms. 

At this point, the expected value framework can be used to assess whether 'acknowledging' or 'not acknowledging' is the better alternative. The expected value of the former is 
EV('acknowledging')=p*V(C1) + (1-p)*V(C2)) =V(C2) +p[V(C1) -V(C2)]
and of the latter is
EV('not acknowledging')=p*V(C3) + (1-p)*V(C4)= V(C4) +p[V(C3) -V(C4)],
where V is the value function and p the probability that EHS is caused by EFs.

Now 'not acknowledging' is the right decision if it has positive expected value, i.e.
EV('not acknowledging')-EV('acknowledging') is positive. This can be written as
V(C4)-V(C2)+p*[V(C3)+V(C2)-V(C1)-V(C4)].

This equation gives us the following insight: If our current belief about the probability of EHS being caused EFs is low (p in the previous equation), then the most important consequences are C4 and C2. What is more, if C2 is much worse than C4 (i.e. V(C2) is significantly lower than V(C4)), then EHS should not be initially acknowledged. 

Actually, I believe that C2 (EHS is acknowledged when it is not really caused by EFs) is the worst out of all possible consequences. My assumptions leading to this belief are:

• If EHS is acknowledged, the number of people who believe that they have symptoms due to EFs will be higher. 
• The symptoms will be experienced to be more severe.
• The costs to alleviating those symptoms will be high.
• Because of the 'default bias', if EHS is first acknowledged, it can be difficult to cancel the decision even if further evidence suggests it should. 

On the other hand, I believe C4 to be the best possible outcome: EHS not acknowledged to be caused by EFs when this indeed is the case. I believe that this would be the best since 1. EHS would not be caused EFs and 2. the number of people who experience symptoms will be much lower.

As conclusion, I think that for now, Electromagnetic hypersensitivity should not be acknowledged as a medical condition caused by electric fields.  At the moment the probability that EHS is caused by EFs seems low. Thus acknowledging EHS might unnecessarily increase the number of people who feel that they have symptoms and make the symptoms more severe. If EHS is acknowledged now, reversing the decision later might be difficult in the case that further research concludes that EHS is not caused by EFs.

It seems to make most sense, to wait for now, do more research and re-think the decision later.
However, I note that even if EHS is not recognized as a medical condition, it can make sense to take the current patients seriously and try to treat their symptoms.