If you suspect that a patient's low glucose levels are due to overdosing on his insulin injections, is there any way to prove this?

Yes, actually there is.

The pancreas, of course, makes insulin. In the final stages of this process, proinsulin is formed. This will be cleaved by an enzyme to form insulin and C-peptide. C-peptide can therefore be found on normal human blood samples.

However, artificially manufactured insulin does not include C-peptide. So if the patient is hypoglycaemic from taking too much (artificial) insulin, the C-peptide level will be low. Aha, you say, but if the cause of his/her hypoglycaemia has nothing primarily to do with insulin levels (e.g. alcoholic binge), then the hypoglycaemia will switch off insulin production. So, the C-peptide level will be low in this case too. Yes, but all is not lost - we need only compare the C-peptide level to the insulin level.

A low insulin level and a low C-peptide level together simply suggest the appropriate insulin response to hypoglycaemia of some other cause. However, if the insulin level is high, but the C-peptide level is low, then the insulin must be artificial.

Incidentally, a fairly rare cause of hypoglycaemia is an insulinoma - an insulin secreting tumour of the pancreas. In this case, both the C-peptide and the insulin levels will be high.