ABSTRACT
When we observe the construction of data-irrespective what kind of data-the construction is choreographed by technical and technological solutions, “tools” for effi cient and valid data collection-at least defi ned as effi cient and valid by system designers and end product users. One such tool is Diagnosis Related Groups (DRG), a classifi cation system which groups in-hospitalized patients on the basis of primary diagnoses, secondary diagnoses, and complications, surgical procedures, age, and discharge routines (Aas et al. 1989). It is a management-oriented case mix system which aggregates upwards (away from the details of medical practice; Bloomfi eld 1991). In Norway, DRG has since 1997 fi gured as a payment system in activity-based fi nancing (ABF) of four regional national health enterprises.1 ABF effectuates the fi nancing of specialist health services with a 60 percent basis allocation and a 40 percent activity-based allocation (Norwegian Directorate of Health 2009). The individual DRG groups, about 500 altogether, are meant to mirror the average cost of treatment in that group. The primary diagnosis is key to the grouping outcome.
