Pierre P Lévy1. 1. Assistance Publique Hôpitaux de Paris (Hôpital Tenon), Unité de Biostatistique et Informatique médicale (INSERM U444) 4 rue de la chine 75970, Paris cedex 20, France. pierre.levy@tnn.ap-hop-paris.fr
Abstract
OBJECTIVE: There is a worldwide consensus to use the diagnosis related groups (DRG) to consider hospital activity. This tool leads to the production of tables with numbers (case mix), the interpretation of which is difficult. The issue is to translate these tables of numbers into an image named Case view. METHOD: It assimilates, in a way, each DRG to a "pixel", the set of the DRGs being an image, the case view. The methods consist of two phases: the first one is to define the reference set while the second one is to visualize data through the reference set. The "DRG-pixels" which constitute the reference set should be organized according to three criteria: medical/surgical, nosological and economic. RESULTS: This method is used to answer theoretical questions or to visualize activity at the level of a hospital or at the level of a department. It generates information of synthetic nature and ought to be used as a complement to existing methods. DISCUSSION: An important advantage of this method compared to the existing ones (DRGs listing, DRGs clustering, histograms em leader ) is that it presents data simultaneously at a global level (sets of DRGs) and at a local level (the DRG). CONCLUSION: We hope this method to be a supplementary step toward the creation of tools capable of eliciting the semantic content hidden in the medical data banks. The purpose of this paper is to explain the underlying general principles that define the graphic representation and to illustrate this model with the use of the French reference set.
OBJECTIVE: There is a worldwide consensus to use the diagnosis related groups (DRG) to consider hospital activity. This tool leads to the production of tables with numbers (case mix), the interpretation of which is difficult. The issue is to translate these tables of numbers into an image named Case view. METHOD: It assimilates, in a way, each DRG to a "pixel", the set of the DRGs being an image, the case view. The methods consist of two phases: the first one is to define the reference set while the second one is to visualize data through the reference set. The "DRG-pixels" which constitute the reference set should be organized according to three criteria: medical/surgical, nosological and economic. RESULTS: This method is used to answer theoretical questions or to visualize activity at the level of a hospital or at the level of a department. It generates information of synthetic nature and ought to be used as a complement to existing methods. DISCUSSION: An important advantage of this method compared to the existing ones (DRGs listing, DRGs clustering, histograms em leader ) is that it presents data simultaneously at a global level (sets of DRGs) and at a local level (the DRG). CONCLUSION: We hope this method to be a supplementary step toward the creation of tools capable of eliciting the semantic content hidden in the medical data banks. The purpose of this paper is to explain the underlying general principles that define the graphic representation and to illustrate this model with the use of the French reference set.