Reporting measurement uncertainties with ethanol results

Abstract Background All measurements are affected by errors whose quantitative expressions are called measurement uncertainty. Blood, urine, breath or serum alcohol concentration is measured to assess if an offence has been committed. In this study, we aimed to calculate the measurement uncertainty for ethanol. Materials and methods Measurement uncertainty was calculated according to the Nordtest guidelines by tracking six steps. All of the patient ethanol results were scanned and recorded between October 2015 and March 2016 by using electronic patient records. Results The measurement uncertainty value was calculated as ±12.25%. The results from 1178 patients referred to our hospital for serum ethanol level analyses were scanned. Of the 1178 ethanol results, 597 were driver. Male persons have higher probability of alcohol-impaired driving than female persons [p = 0.007; odds ratio was 2.239 (1.211–4.141 with 95% confidence interval)]. Conclusion The measurement uncertainty for ethanol concentrations measured in our laboratory was lower than the allowable total error (TEa) of the guidelines (20%). Laboratories that analyze ethanol have to be more transparent about measurement procedures when presenting their ethanol result. Measurement uncertainty should be limited with legislative regulations by country-specific scientific committees.