Validation of self-reported cigarette smoking in a remote Australian Aboriginal community

Objective: To examine the relationship between self-reported tobacco smoking and urinary cotinine concentrations in the setting of a remote Aboriginal community. Methods: In a remote Northern Territory (NT) Aboriginal community the relationship between self-reported tobacco smoking and urinary cotinine concentrations was examined as part of a cross-sectional survey of cardiovascular risk factors. Current tobacco smoking was assessed as part of a questionnaire. The concentration of cotinine and cotinine/creatinine ratio (CCR) in a spot urine sample were used as a biochemical marker of nicotine exposure. Results: A total of 237 people took part in the survey, although completed questionnaires and urine results were available for 184 people. Current tobacco smoking was reported by 161 (69 [95% Cl 63 to 75]%) people, with higher rates among males (84/104, 81 [95% Cl 72 to 88]%) than females (77/129, 60 [95% Cl 51 to 68]%, p<0.001). There was good agreement of self report with current tobacco smoking using categories based on urinary cotinine (agreement 94%, k=0.84) and urinary CCR (agreement 94%, k=0.86). Quantitative agreement between cotinine measures and self-reported number of cigarettes per day was better for CCR than cotinine, but the correlation was relatively weak for both. Conclusions: Self-reported cigarette smoking is a valid qualitative measure in this environment. The relatively weak correlation between the biomarker and number of reported cigarettes smoked illustrates problems both with questionnaire and urinary markers of nicotine exposure. Implications: As a qualitative measure of tobacco smoking in this setting, self-report appears adequate but the validity of quantitative self-report is unclear.