We tried to use data mining method to find out the new model of factors that affects the progression of diabetes mellitus.

The advanced development of medical technology had led to the accumulation of a large amount of medical data. Unfortunately, these useful and meaningful clinical data were not properly managed or analyzed. Therefore, data mining technology has become an essential instrument for hospital management and medical research. Possible meaningful information can be obtained from the database through data mining. About 2,000 diabetic patients had been enrolled in Diabetes Care Program since Oct 2002 in our hospital. We intend to look for the determining factors that affect the prognosis of diabetes mellitus in our patients group.

Inclusion Criteria:

• diabetes mellitus who had been enrolled into the Diabetes Care Program from year 2002 until 2007.

Exclusion Criteria:

• none

This retrospective cohort study included 2,000 patients who visited our Division of Endocrinology and Metabolism outpatient department in this hospital with the diagnosis diabetes mellitus who had been enrolled into the Diabetes Care Program from year 2002 until 2007.