EVIDENCE LIBRARY
Did Acute Alcohol Significantly Change Blood Clotting in Healthy Volunteers?
A 20-person ROTEM study found a small early fibrin-clot change before statistical correction, but the overall result did not survive correction for multiple comparisons. It measured neither crash causation nor bleeding outcomes.
A small volunteer study found a modest early fibrin-clot change after drinking, but its overall result did not remain statistically significant after correction for multiple comparisons. The experiment measured clot formation in sampled blood, not driving impairment, clinical bleeding or the cause of a crash.
What was measured
Eismann and colleagues studied 20 healthy volunteers, 11 men and nine women, before and during one hour of alcohol consumption. Participants chose wine, whiskey or vodka and consumed an average of approximately 51 grams of ethanol.
The intended alcohol concentration was 1‰, but the final mean reported BAC was 0.76‰. Laboratory ethanol was measured in serum by headspace gas chromatography and converted to BAC. Breath readings were used to guide collection timing; they were not identical to the laboratory concentration series.
The investigators tested citrated blood with rotational thromboelastometry, or ROTEM. ExTEM assesses clot formation with platelets contributing, while FibTEM suppresses the platelet contribution to emphasize the fibrin component. These mechanical clot measurements are not alcohol biomarker concentrations.
The headline result requires the corrected p-value
FibTEM clot amplitude ten minutes after clotting began fell from a mean of 14.0 to 13.0 mm between baseline and the final sample. The uncorrected p-value was 0.014. After the paper’s Holm–Bonferroni correction for multiple comparisons, the p-value was 0.112.
The latter does not meet a 0.05 significance criterion. Although the abstract describes a significant change, the table shows that the overall finding did not survive the stated correction. The observed one-millimeter mean difference remains a descriptive observation; it is not strong confirmatory evidence of an effect across the full cohort.
Maximum FibTEM clot firmness also declined descriptively, from 16.2 to 13.7 mm, but was not statistically significant before or after correction. ExTEM parameters and fibrinolysis measures likewise showed no significant change. No hyperfibrinolysis was observed under the study’s criterion.
The male subgroup does not prove a sex-specific alcohol effect
In the 11 men, the early FibTEM amplitude change remained significant after correction, p = 0.042. The nine women did not show a significant change. However, the men also reached higher mean alcohol concentrations than the women: approximately 0.87‰ versus 0.62‰.
A significant result in one subgroup and a nonsignificant result in another do not themselves establish a statistically significant difference between subgroups. Exposure differed, and the study was not powered to establish specific sex effects.
Healthy volunteers were not injured crash patients
The study removed many clinical complications by enrolling healthy people without recent medication or alcohol use. That helps isolate an acute exposure question, but it also limits application to a trauma patient with shock, tissue injury, medications, transfusion or preexisting disease.
The participants did not experience crash injuries. The researchers did not measure bleeding, transfusion need or survival. A change in a laboratory clot trace cannot be converted directly into a probability that alcohol worsened an individual person’s bleeding.
Different beverages were self-selected, there was no parallel nonalcohol drinking control group, and follow-up lasted only about one hour. The planned alcohol target was not reached on average. Results from ethanol added directly to blood in a laboratory also cannot be assumed to reproduce drinking in a living person.
What this means in a DUI-related crash review
A clotting result addresses hemostasis, while an ethanol result addresses concentration in a specified matrix. Neither substitutes for evidence about the driver’s behavior or crash mechanics. This paper cannot date drinking, establish a driver’s BAC from ROTEM, or diagnose alcohol-caused driving impairment.
For a clinical alcohol result, see the site’s serum alcohol and measurement uncertainty analysis. The trauma clinical examination study addresses a different inference: whether observed appearance establishes a low alcohol concentration.
Primary source
Eismann H and colleagues. Influence of alcohol consumption on blood coagulation in rotational thromboelastometry (ROTEM): an in-vivo study. Korean Journal of Anesthesiology. 2020;73:334–341. https://doi.org/10.4097/kja.20071.