Study design, recruitment, and baseline characteristics: the Department of Veterans Affairs Dental Diabetes Study.

OBJECTIVES We are conducting a clinical trial of the efficacy of periodontal therapy in the improvement of glycaemic control in veterans with poorly controlled diabetes. This report describes study design, recruitment and randomization and compares baseline characteristics of the sample frame with those randomized into study groups. METHODS Veterans with poorly controlled diabetes were randomized in two groups: immediate periodontal therapy ("early treatment") or usual care followed by periodontal therapy ("deferred treatment"). Half of each group continued care for 12 months; the other half returned to their usual care. We studied baseline patient characteristics, self-reported health measures, and clinical examination data. We examined means for continuous variables, frequencies for categorical variables and compared groups using t-tests and chi(2) tests (alpha=0.05 for both). RESULTS The 193 randomized participants were younger (58 years) and had slightly higher HbA1c (10.2%) than the 2534 non-randomized participants (64 years, HbA1c =9.8%). The deferred treatment group was more likely than the early treatment group to have a history of stroke, transient ischaemic attacks, and less likely to be current or former smokers. CONCLUSIONS The mechanism for randomization was largely successful in this study.