You have a patient in the ER in septic shock. You are about to administer resuscitation fluid but notice that the patient is severely hyponatremic. You don't want to correct too rapidly and risk osmotic myelinolysis. What do you do? I've seen this situation many times. This problem is addressed in a post at Academic Life in Emergency Medicine and some references provided.
One thing you could do, says the author, is use lactated ringers which is hypotonic to normal saline (though it may still be hypertonic to a patient with severe hyponatremia). After all, there's been a shift toward lactated ringers in sepsis resuscitation for entirely different reasons anyway so why not?
If you do the math (a link to a calculator is provided) you find that, all other things being equal, lactated ringers raises sodium in a severely hyponatremic patient a lot less than saline. And, using the calculator, you can reliably predict how fast the sodium will rise with a given rate of fluid resuscitation, right? Wrong. Because all other things are not equal. You can bet that patient is pouring out vasopressin, driven by non-osmotic stimuli (volume receptor mediated). When you replete the patient's volume you will turn off that signal and a water diuresis may ensue. That is a variable that cannot be anticipated in the calculation. The only way I know to deal with that situation is to recheck the patient's chemistries frequently, pay attention to the urine output and urine osm in order to catch any trend toward overcorrection early, then deal with it however your sound clinical judgement may dictate.
That is not to take away from the point that lactated ringers may be preferable.