Use of the UKPDS Outcomes Model to predict all-cause mortality in U.S. adults with type 2 diabetes mellitus: Comparison of predicted versus observed mortality

Aims: The applicability of the United Kingdom Prospective Diabetes Study (UKPDS) Outcomes Model is unknown in populations with type 2 diabetes mellitus (T2DM) outside the United Kingdom. We compared all-cause mortality predicted from the UKPDS model with observed mortality among T2DM subjects in the U. S. Methods: We studied participants with T2DM from the National Health and Nutrition Examination Survey 1988-1994 with characteristics comparable to the UKPDS cohort. The 10-year observed all-cause mortality was compared to the UKPDS model-predicted mortality. The Lifetable method was used to estimate the probability of mortality for 10 years following diagnosis. Results: Among 156 subjects with characteristics comparable to the UKPDS cohort, mean age was 49.6 years, age at T2DM diagnosis was 47.1 years, and T2DM duration averaged 2.6 years, with follow-up for 10.4 years. The UKPDS model-predicted 10-year mortality was 15.7%, similar to the observed mortality of 14.2%. Corresponding 10-year predicted versus observed mortality was 32.7% versus 32.4% including subjects >age 65, 17.0% versus 19.3% including individuals with pre-existing CVD, and 31.1% versus 20.9% including individuals with diabetes duration >= 6 years. Conclusion: All-cause mortality predicted by the UKPDS model was comparable to observed mortality in U. S. NHANES participants with characteristics similar to the UKPDS. (C) 2010 Elsevier Ireland Ltd. All rights reserved.