The measurement of exhaled carbon monoxide is influenced by airflow obstruction.

The concentration of carboxyhaemoglobin (COHb) is often estimated from measurements of carbon monoxide in the exhaled air (COexh). This study investigates whether the presence of airflow obstruction significantly alters the relationship between COexh and COHb. Eighty-one regular smokers were prospectively studied and divided in four groups according to the presence and severity of airflow obstruction (none, mild, moderate, severe). In each subject, the authors measured in this order: 1) arterial blood gases; 2) haemoglobin concentration and COHb (by co-oxymetry); 3) COexh; 4) lung volumes; and 5) forced spirometry. The size of the measurement error (deltaCO) was calculated from the difference between COHb and COexh. Neither the smoking history nor COexh were different in the four groups of subjects studied. In contrast, deltaCO increased in parallel to the degree of airflow obstruction. DeltaCO was >2% (a threshold value normally used in the clinic to separate smokers from nonsmokers) only in patients with severe airflow obstruction. A stepwise multivariate analysis showed that both forced expiratory volume in one second (FEV1) (percentage reference) and COHb contributed significantly (p<0.0001) to predict deltaCO. This study shows that the estimation of carboxyhaemoglobin from exhaled carbon monoxide measurements can be inaccurate in patients with severe airflow obstruction. In these patients, the direct measurement of carboxyhaemoglobin seems advisable in clinical practice.