This time, it's an essentially semi-tabloid article about how a dog helps a six-year-old girl with diabetes (article from Hovinky here (http://www.novinky.cz/koktejl/204119-sestilete-divce-trpici-cukrovkou-nekolikrat-tydne-zachrani-zivot-jeji-pes.html):)

I appreciate that the journalist did not overlook the fact that this is DM1 — that is, type one diabetes mellitus, where the patient depends on insulin replacement therapy and is therefore at risk of both hypoglycaemia and hyperglycaemia. Nevertheless, some passages are confusing yet again.

"The young female dog is trained to detect low blood sugar. Since Rebecca has diabetes, Shirley always alerts her when it is time for insulin." — I consider this passage very dangerous, because it could give a layperson the mistaken idea that low blood sugar is treated with insulin. Unfortunately, precisely such attempts to help are documented in the medical literature as well-known cautionary tales.

High blood sugar — hyperglycaemia: the patient needs more insulin

Low blood sugar — hypoglycaemia — impending coma: the patient urgently needs glucose. An injection of insulin in such a case will reliably kill them.

The worst thing is that the original article the journalist used as her source (here (http://www.dailymail.co.uk/health/article-1288589/Diabetic-girl-6-life-saved-times-WEEK-dog-sniff-needs-insulin-injection.html)) contains no such nonsense, so this is no translation error but her very own, entirely original invention.

Any sloppily done job can have nasty consequences, but that goes double for journalism. So please, people, if you know someone with diabetes and find them in a coma, do not try to help them with insulin. Call an ambulance immediately and tell the responders about the person's condition. They will probably give them glucose intravenously.

RIK wrote this when the original blog was first created; not all the information stands up to comparison with current knowledge.