Administrative data was anonymously linked to a clinical registry, to evaluate the validity of diabetes algorithms in youth. In addition, incident youth with T2DM (n=342) in Manitoba (1-18 years) from Jan.1986-2009 identified from the clinical registry were anonymously linked to healthcare records in order to evaluate complications, compared to youth with type 1 diabetes (T1DM) (n=1011) and non-diabetes (non-DM) controls (n=1710).
The algorithm including 1 or more hospitalizations or two or more outpatient claims over two years was the most valid. Youth with T2DM had a 47% increased risk of any complication and a 2.29 fold increased risk of renal complication. Age at diagnosis, HgA1c and ace inhibitor/angiotensin receptor blocker use (ACE/ARB) were significant risk factors for any complication. Risk factors for renal complications included ACE/ARB use, albuminuria and diagnosis prior to 2000. Survival at 10 years was 91.4% (T2DM) vs. 99.5% (T1DM) vs. 100% (non-DM); p<0.0001.