Medical diagnoses showed low relatedness in an explorative mutual information analysis of 190,837 inpatient cases.

OBJECTIVES: We aimed to quantify the shared information between medical diagnoses of an adult inpatient population to explore both multimorbidity patterns and vice versa the unrelatedness of medical diagnoses. STUDY DESIGN AND SETTING: This was a cross-sectional study, performed at a tertiary care center in Switzerland. Diagnoses were routinely coded using the International Classification of Diseases, 10th revision. RESULTS: Among 190,837 inpatient cases, 7,994 unique diagnoses were coded. There were 31.9 million possible diagnosis pairs; the respective mutual information scores in diagnosis pairs were low (range, 10-7 to 0.237). There were 148 pairs of diagnoses with a mutual information score higher than 0.01, which formed several clinically plausible disease clusters; 27.2% of cases did not have a diagnosis that belonged to one of the morbidity clusters. CONCLUSION: In an explorative analysis, we observed a high unrelatedness of diagnoses in a tertiary-care inpatient population. This finding indicates that although multimorbidity patterns can be observed, inpatient cases frequently have further, unrelated diagnoses, which share little information with specific other diagnoses. Therefore, management of multimorbid patients should be individualized and may not be generalized based on a few multimorbidity patterns or clusters.

Secondary use of routine data in hospitals: description of a scalable analytical platform based on a business intelligence system.

We describe a scalable platform for research-oriented analyses of routine data in hospitals, which evolved from a state-of-the-art business intelligence architecture for enterprise resource planning. This platform involves an in-memory database management system for data modeling and analytics and a high-performance cluster for more computing-intensive analytical tasks. Setting up platforms for research-oriented analyses is a highly dynamic, time-consuming, and costly process. In some health care institutions, effective research platforms may be derived from existing business intelligence systems.