Abstract
Background. Evaluation of health care contacts from first events alone often misses large amounts of potentially important data and may produce different results than evaluation of all data including recurrent events.