Accuracy and reliability of breath alcohol testing by handheld electrochemical analysers

Usefulness of portable, handheld breath analysers equipped with electrochemical sensor was assessed. Breath- and blood-alcohol concentrations in drunken drivers were taken from 370 expert opinions elaborated at the Institute of Forensic Research between January 1st 2002 and February 28th 2007. The results of second and subsequent measurements were re-calculated using mean elimination rates. The readings of portable instruments were in very good agreement with the results of confirmatory analyses performed by stationary devices (r = 0.978, p < 0.001, y = 0.969x − 0.0002). The correlation with the results of blood analysis was weaker (r = 0.940, p < 0.001, y = 1.722x + 0.214), but comparable with the correlation between the readings of stationary devices and the results of blood analyses (r = 0.936, p < 0.001, y = 1.790x + 0.091). The readings of portable and stationary breath analysers were also compared by the Bland–Altman plots. The differences in results were independent of alcohol concentration (absolute difference (mg/L): r = 0.054, p > 0.1, y = 0.011x + 0.013; relative difference (%): r = 0.020, p > 0.1, y = 0.90x + 2.36).