It should be clear from this discussion that coma cocktails are a bad idea and should be immediately abandoned. In fact, the indiscriminate use of the coma cocktail may indeed harm patients, EMS has evolved to a point where any EMS provider should be able to reasonably determine the most likely cause of coma, or, in a worst-case scenario, narrow the cause to but a few possibilities. Certainly, patients with bona fide hypoglycemia should receive IV glucose. Because the consequences of prolonged hypoglycemia are severe, if there's a doubt about whether hypoglycemia is present, then glucose should be empirically administered. Naloxone should be used only for those cases in which a narcotic overdose appears likely. Similarly thiamine administration should be limited to patients suspected of chronic alcohol abuse and who exhibit at least one of the three symptoms of WE described above. Flumazenil has no role in the routine treatment of coma unless the patient is known to not be benzodiazepine dependent and the overdose is known to result only from benzos--two very difficult requirements to verify in the back of an ambulance at 2 a.m. Coma cocktails are bad medicine. Let's banish them from our EMS armamentarium.