Biochemical verification of tobacco use in ICU patients using point-of-care urinary cotinine.

Objectives: Obtaining an accurate smoking history can be challenging in critically ill patients. The objective of this study was to determine the prevalence of nicotine exposure using a point-of-care urine cotinine test in a heterogeneous cohort of critically ill patients and to compare this to the smoking history in the medical record. Design setting and participants: In 481 adult patients we prospectively undertook a urine cotinine test within 24 h of admission using a point-of-care binary colorimetric assay with a detection threshold of 200 ng/mL. Smoking history was retrieved from the electronic medical record. Main outcome measures and results: Urinary cotinine was detected in 151 of 481 patients (31.4%), exceeding the 134 participants with a documented history of current smoking by 3.5% (relative increase 12.7%). A positive urine cotinine test was reported in 12% (40/347) of participants not known to be actively smoking, including 18 'former smokers', 12 'non-smokers', and 10 in whom smoking status was unknown. Cotinine-positive participants were ventilated for longer than cotinine-negative patients; a median of 71 h vs. 44 h (median difference of 30h, 95% CrI: 5.6 to 56); they were more likely to have a "code-grey" event; 34 of 151 (23%) vs. 24/330 (7.3%); median relative risk of 3.1 (95% CrI, 1.9-5.1); and were more likely to "discharge against medical advice"; 12/125 (10%) vs 3/275 (1%), with a median relative risk of 9.7 (95% CrI, 3.0-46). Conclusions: Urinary cotinine testing offers a point-of-care determination of current tobacco smoking status in the critically ill.