Comorbidity coding trends in hospital administrative databases

Abstract

Comorbidities are related to an increase in hospital costs, hospital mortality and length-of-stay. The analysis of inpatient comorbidities is therefore important for hospital management, for epidemiological studies, and for health services research and planning. In this paper we study the annual evolution of coded comorbidities in administrative databases. We used data from Portuguese hospitals over a period of eleven years (2000-2010). We identified comorbidities in 7,034,213 inpatient episodes using both the Elixhauser and the Charslon/Deyo comorbidity methods. Our results clearly evidenced a positive association between the number of secondary diagnosis and coded comorbidities. We argue that the increased number of comorbidities over time is mostly related to an increase in the quality of coded data. Data analysts, researchers and decision makers should always be aware of possible data quality bias, such as completeness, when using administrative databases.