An air embolism would only be a problem if it was in a blood vessel. Those injectors were intramuscular and the air just diffused back out through the tissue.
I can't think of a case where this can be a problem by accident. Back when I was in the medical school (Bulgaria, Eastern Europe) we sometimes used air embolism to kill off experimental animals. I remember that the quantity of air needed is substantial. Something like 20-50 cubic centimeters needed to kill a rabbit reliably by venous (right heart) embolism. I imagine the quantity needed to kill a human is quite a lot bigger.
Of course, it's not the same if you insert air in the carotid artery. I guess that quite smaller quantity there will kill a human. I speculate something in the range below 1 cm3.
>apparently it had some problems
Air emboli?