While I am not a renal physician, I don't see any literature on this. Furthermore, the chances of the second patient having an infection on the same side as the first patient are 50%, so the results of your small case series are not really cause to start us searching for explanations. 

If six patients, sampled randomly, have infections on the same side, then this is statistically significant (binomial test). However, you need to discard the two cases you already know about, as they don't constitute a random sample. 

Thanks for a kind advice!

Now I understood that my question is not so important. 

It really helped to know that.

Nobuhiro

130

I would not say that the question is unimportant.

I will give you an example. We were at coffee one day and someone mentioned that a patient had died suddenly, just before discharge, though apparently making an uneventful recovery.

"That's odd", said a colleague, and he went on the say that the same thing had happened a few months earlier: same kind of infarct, same (rather trivial) heart block, sudden death.

We excluded those two patients and combed back through the hospital charts for the previous two years, and found a small group of patients with just that presentation, and discovered that this group ran a risk of sudden death that was far higher than you would expect based on their presentation.

It is noticing these little coincidences that can lead nowhere (which is reassuring) or lead to learning something new and important.

So never call curiosity unimportant!