S-Grouper: A Semantic Approach to Increase Quality Performance by Innovating Information Systems in Healthcare Organizations

In Italy, as in many other countries, hospital acute activities are classified and measured using the DRG system. DRG system was originally created in the early 80’s within the MEDICARE scheme of USA HealthCare System. DRG system is even used, at least partially, to finance hospitals. The measuring process implemented by many hospitals is based on a specific software, well known as DRG-Grouper, (Grouper) that, starting from clinical records, is able to automatically generate the specific DRG code. The efficacy of the coding process requires a direct involvement of physician in collecting and analyze clinical records. Clinical data needed for classification are collected on the base of clinical records measured by physicians that are often required to manually codify diagnosis and clinical procedures under the international ICD-IX-CM standard. As it not difficult to imagine, physicians do not appreciate administrative activities because they consider administration as a waste of clinical time. For this reason, in many cases, the measuring process leading to DRG identification is characterized by lot of mistakes and to serious problems for public hospitals. These problems are related both to economic issues and to quality issues. Focusing on quality two elements should be highlighted: a) On a first basis quality of the administrative process. Classification mistakes lead to conflict with financing bodies and agencies that end in a relevant increase in not planned control activities within hospital organizations and in a waste of value; b) On a second basis clinical quality. Classification mistakes are often related to elements such as effectiveness of care, appropriateness of care and outcome evaluation of hospital activities directly affecting the overall quality of care. The paper, based on the results of an in site research project carried out by Trento University with GPI and Expert System, aims to analyze and measure how DRG classification processes can be improved and sped up by implementing and integrating semantic analysis in traditional information systems. Semantic analysis is, in fact, a tool that let automatically emerge from traditional documents the information that are useful and needed to implement an automatic process for monitoring and verifying the quality and relevance of their content in order to classify DRGs.