Pan-disease clustering analysis of the trend of period prevalence

Prevalence is of essential importance in biomedical and public health research. In the “classic” paradigm it has been studied for each disease individually. Accumulating evidence shown that diseases can be “correlated.” Joint analysis prevalence potentially provide important insights beyond individual-disease but not well pursued. this study we take advantage unique Taiwan National Health Insurance Research Database (NHIRD) conduct first pan-disease period trend. The goal to identify clusters within which have similar trends. A novel penalization pursuit approach applied an intuitive formulation preferable numerical performance. data values are computed using records on close one million subjects 14 years observation. With 405 diseases, 35 with sizes larger than 27 identified. clustering results sound interpretations differ significantly from those alternatives.