Yes IoB = Insulin on Board.
This can be simple or complex.
The simplest method would be to assume all bolus insulin is used up (acts) evenly over a period of 4 or 5 hours. On pumps (and IoB enabled meters?) the user sets this value for the total duration of effect, to suit their type of QA insulin and their own response to it. Usually the range is 4-6 hours and a typical value would be 4.5 or 5.
So if 4 hours is selected, in a simple straight line model we assume 25% is used up per hour, or if 5 hours total duration, 20% per hour.
This is a simplification as in reality the insulin action is a curve, not a straight line. Pumps store a standard curve and rescale that curve according to the user-entered value for total duration of action, eg 4-6 hours. Action curves are published by the drug manufacturers. (If you want to get really sophisticated, the action curves are dose-dependent, but most IoB calculators ignore this.)
If using a straight line model rather than a curve, it is best to use a slightly shorter assumed total duration of action, just because the last hour of action of a QA insulin is usually only a few % of the total, so a slightly shorter assumed duration gives a better fit to the real curve.
The other aspect is how to handle the fact that IoB is cumulative.
In a very simple model, when a new bolus dose is given while there is still a non-zero IoB, you can just add the new dose to the current IoB and keep a single running total. That's simplistic but not terrible. A more accurate method is to track each bolus separately, until it expires, and sum all active boluses together to determine IoB at any given point in time.
So as I say you can make it as simple or as complex as you feel like.
The purpose of tracking IoB is to allow people to safely take multiple bolus injections of QA insulin that are spaced more closely than the total duration of action of that QA insulin. In the absence of IoB information, we are advised not to inject more frequently than that, and certainly not to correct more frequently than that.
