Don't be fooled into dialyzing patients that don't need to be dialyzed! You have to beware of the pseudohyperkalemia.
The most common cause of pseudohyperkalemia of course is due to hemolysis from venipuncture.
There are other, more rare, possibilities however.
In hereditary spherocytosis, for instance, the RBC membrane is susceptible to lysis after collection, and may result in a serum K that is much higher than the in vivo situation.
Familial pseudohyperkalemia is a genetic condition in which RBCs exhibit a leak of K without hemolysis when stored a room temperature (though there is no observed K leak when stored at 4 degrees). Several loci have been identified, but no specific gene has been pinpointed yet.
Extreme leukocytosis (e.g., with blast crisis in various leukemias) may also lead to pseudohyperkalemia.
Finally, there is the phenomenon of "regional pseudohyperkalemia"--the observation that samples which sit around in a cold climate tend to be (on average) very, very slightly more hypokalemic whereas samples which sit around in a warm climate tend to be (on average) very slightly more hyperkalemic.