Assessment of a simple colorimetric procedure to determine smoking status in diabetic subjects.

The performance of a colorimetric assay for nicotine metabolites to validate self-reported smoking classification (nonsmoker, ex-smoker, and smoker) was assessed in a group of diabetic patients (n = 201). Comparison of the results with those of cotinine immunoassay (ELISA), by comparing respective areas under receiver operating characteristic curves, established the superiority of the cotinine immunoassay method. Adjusting the urinary concentrations of nicotine metabolites for creatinine excretion significantly lowered test performance. The sensitivity and specificity for the assay of nicotine metabolites to discriminate smoking classification within the diabetic patients at a threshold of > or = 28 mumol/L "cotinine carboxylic acid equivalents" were 68.4% and 98.6%, respectively; the corresponding sensitivity and specificity for urinary cotinine at a cutoff of > or = 0.5 mumol/L were 94.7% and 100%. The low sensitivity of the colorimetric urinary nicotine metabolites assay precludes its application as an objective assessment of smoking status in our patient population.