The influence of diagnostic criteria on the mortality findings in the university group diabetic programme study of diabetic therapy

The University Group Diabetic Programme (UGDP) clinical trial of diabetic therapy demonstrated a higher mortality from cardiovascular disease amongst patients treated with oral hypoglycaemic agents. However, these results may have been biased by imbalances between treatment groups in the proportion of patients who were unequivocally diabetic. In order to eliminate this potential bias and to increase the specificity of the diabetic diagnosis, the UGDP mortality findings have been reassessed, using multiple diagnostic criteria as a basis for case selection. The use of multiple diagnostic criteria increases the mortality differences between the patients treated with oral hypoglycaemic agents and those treated by placebo and diet. This reinforces the original findings of the UGDP study.