While we are on the topic, be mindful that D50 isn't the only thing we administer capable of causing local necrosis.
Calcium for example is capable of causing massive necrosis. I've read calcium is actually capable of producing full thickness necrosis.
Even hypo/hypertonic saline can produce necrosis by overhydration/dehydration of cells.
Personally I take a lot of pride in showing up at the hospital with a clean, patent IV site. I know many guys who skip using a lock/flush and just directly attach a drip set but personally I like to always use a saline lock/flush as it provides a detachment point other than the catheter and more importantly it allows me to safely ensure my site is good to go before administering any meds. Unless you extravasate enough normal saline to substantially stretch the connective tissues of the skin you aren't going to cause any adverse effects. Better safe than sorry.
Funny, just today my medic partner and I were talking about old-school ALS, like D50/thiamine/narcan, 3+3 liters of saline for shock, etc, and he mentioned how we used to just run the fluid in and not do a lock. We were saying how we now use 500cc bags instead of liter bags to save some $$$ and not over-bolus someone. If you need more, just spike another 500 bag when the last one goes down below 100.
Anyway, we have protocols that allow us to dilute D50 in 250cc's of NS for the hypoclycemic pt if their vasculature appears poor. We used to do this all the time in the field, but we just wouldn't document it. The North Shore PCR Nazis would haved tripped out for sure. Same thing with the hyperglycemic pt. If we have a BGL > 400 and an ETCO2 of 25 or less, we can run NS 1000 cc's/hr. IIRC, that's 2.77 gtts/min. We used to give hyperglycemics fluid off the record back in the day as well. They're urinating a lot, and can be hyperventilating themselves into dehydration, so some saline should help for the time being.