Introduction

In Germany and many other countries, retrograde extrapolation of blood alcohol concentration (BAC) is a routine task in legal medicine, forensic toxicology, and traffic medicine. But, its permissibility and practical handling at very low measured BAC values differs between jurisdictions and expert traditions. In some settings, extrapolation is discouraged or avoided below a defined threshold, because close to zero it becomes increasingly difficult to exclude late-phase elimination behaviour and because relative analytical uncertainty becomes more influential. As a result, borderline cases may be handled inconsistently, despite the fact that such low BAC values can still be relevant in legal contexts (e.g., assessment of recent drinking, plausibility checks of statements, or interpretation of marginal findings).

Retrospective estimation of BAC has traditionally been performed using a linear (Widmark-type) elimination assumption; however, elimination kinetics are not strictly constant across the concentration-time curve and are influenced by the concentration range considered and by absorption conditions. In particular, when calculations are started from very low BAC values, both kinetic-model uncertainty and analytical imprecision become increasingly influential, which may affect the robustness of retrospective estimates.

In forensic expert practice in Germany, retrograde calculation of the minimum BAC is generally not considered reliable when the initially measured value is below 0.15‰. The present study addresses this practical gap by re-analysing controlled drinking data with measurements extended to complete ethanol elimination. We quantify elimination behaviour in the low-BAC range and derive a conservative, easy-to-apply correction approach for retrograde estimation below 0.15‰, including a suggested lower limit for application.

A review of the scientific basis commonly cited for omitting retrograde extrapolation at very low BAC values indicates that several practical rules originate from experimental work in which elimination rates were estimated over specific post-peak concentration ranges and observation windows. Because the apparent slope of the concentration-time curve depends on the concentration interval considered and on absorption conditions, elimination estimates may vary and should be applied with caution when calculations are initiated from very low BAC values (Lewis 1986; Mihan et al. 2025) [1, 2].

At present, gas chromatography is the method most widely used. According to the “Guidelines for determining blood alcohol concentration (BAC) in blood for forensic purposes,” jointly issued by the German Society of Legal Medicine (DGRM), the German Society of Toxicological and Forensic Chemistry (GTFCh), and the German Society for Traffic Medicine (DGVM) [3], determination of BAC in criminal proceedings requires measurement of the ethanol concentration in whole blood (or measurement in serum/plasma with subsequent conversion of the result to an equivalent whole-blood concentration) using two different analytical methods, each performed in duplicate. The legally relevant BAC corresponds to the mean value of the four individual measurements. The range of values must not exceed 10% of the mean value (for mean values < 1‰, a maximum deviation of 0.1‰ is allowed). Calculating the mean of four individual determinations increases statistical reliability and compensates for random measurement errors. To ensure and verify reproducibility and accuracy, regular calibrations and controls using quality-assured samples (internal quality control) as well as proficiency testing schemes (external quality control) are routinely performed.

Controlled drinking studies with serial sampling continued to complete ethanol elimination permit a direct evaluation of elimination behaviour in the low-BAC range, including the interval below 0.15‰. For this purpose, the raw data of four studies of controlled drinking experiments, carried out at the Frankfurt Institute of Legal Medicine, were pooled and analysed. Blood sampling was continued until the end of the elimination curve (at least until a negative breath alcohol test was obtained). While the general principles of post-peak BAC decline are well described, practical back-calculation at very low measured BAC values remains debated, because uncertainty increases as concentrations approach zero and absorption-related effects can influence the curve near the peak. The focus of the present study was to investigate alcohol elimination during the final phase of the curve, and to assess the elimination process in this range. The central question was whether, in cases of measured BACs below 0.15‰, retrograde calculation of the minimum BAC for the time of the incident might be possible – and if so, under which prerequisites or conditions.