Using SNOMED CT to identify a crossmap between two classification systems: a comparison with an expert-based and a data-driven strategy.

A crossmap between successive versions of classification systems is necessary to maintain the continuity of health care documentation. A reference terminology can serve as an intermediary to support this task. Within this study we evaluated the use of SNOMED CT to create a crossmap between two versions of an intensive care classification system. Firstly, the SNOMED CT crossmap was compared with an expert-based and a data-driven crossmap. Next, the influence of these crossmap strategies on the health care outcome was evaluated. For 50% of the analyzed cases, the three mapping strategies resulted in the same crossmaps. In other cases, there was an overlap between the SNOMED CT crossmaps and the crossmaps provided by one of the two other strategies. Differences in the crossmap results had however no significant influence on the health care outcomes. SNOMED CT can be used as an intermediary to solve the problem of crossmapping between versions of classification systems.