Data items can be aggregated into collections thereof: for example, most readings (vitals, labs, signals) can be aggregated into time series.
A decision may process a collection of items as a whole, e.g. to look for trends, or may process the elements individually - and then define a way to aggregate the outcome