Statistical insights into ethanol testing: demographic variations and laboratory performance – A core laboratory experience

Introduction

Alcohol is a type of drug that affects the central nervous system and can alter behavior, mood, cognition, and perception, and about 2.3 billion adults worldwide consume it at least once a year [1]. Even though it is legal and commonly consumed in many societies, it can be harmful and addictive if used excessively. Ethanol is the only type of alcohol that is suitable for oral consumption. Its use may result in the development of dependence, tolerance, and withdrawal symptoms in certain individuals. Moreover, chronic alcohol use can cause severe health problems such as liver damage, cardiovascular issues, and an increased risk of accidents or injuries [1], 2]. Ethanol affects the brain and nerve tissues depend on the dose consumed. At 20 mg/dL, one may feel euphoria and relaxation. At 50 mg/dL, hot flashes and reduced caution may appear. Above 100 mg/dL, motor coordination and judgment may be affected, while at 200 mg/dL, confusion may set in. Above 400 mg/dL, undesired effects like stupor or coma may occur [3].

Due to its widespread consumption, measuring blood ethanol level (BEL) across drinkable alcohol types is an increasingly important test in routine biochemistry laboratory practice [4]. The statistical analysis of BEL will aid in managing clinical and laboratory processes. More than real-time data presented in the literature regarding ethanol analysis is required. The Başakşehir Çam Sakura City Hospital in Istanbul is one of Türkiye’s largest healthcare service centers, with 2,500 beds, an annual total of 1.1 million emergency room visits, and 3.8 million outpatient visits. It is anticipated that a high number of ethanol samples will be processed at a center providing services at this scale. Between 2020 and 2023, we conducted 9,521 ethanol measurements in our laboratory. In this study we aim to use statistical methods to summarize the extensive data and present it in the literature. Within the scope of this study, our aim is to present the distributions of blood ethanol values measured by age, gender, and positivity rates. We will also include turnaround time values, test rejection statistics, and other test analytical performance evaluations such as total analytical error in the literature. Our foresight suggests that sharing data on laboratory processes and demographic information will enrich the literature.

Materials and methods

The ethical approval for the study numbered KAEK/24.07.2024–144 conducted at the Medical Biochemistry Laboratory of Istanbul Başakşehir Çam and Sakura City Hospital, was obtained from the relevant hospital ethics committee on March 06, 2024. The Helsinki Declaration was referenced in the planning and execution of the study. 9,521 ethanol test results conducted between 2020 and 2023 were retrospectively reviewed and utilized. In this study, demographic data such as age and gender distributions, positivity rates, BEL in positive cases, and the distribution of test request times were evaluated. Test turnaround time, test rejection rates, test repetition rates, and the rate of blood ethanol level decline were calculated under laboratory processes. To calculate the ethanol elimination rate, patient selection was conducted among those with consecutive measurement results recorded in the hospital information system database. Patients with a clinical suspicion of methanol intoxication were excluded from the selection.

In Türkiye, the processes of ethanol measurement have been legally mandated to be extensively defined and required to be adhered to through the directive “Principles and Procedures for Ethanol Measurement in Blood Samples” [5].

The process for ethanol sampling and testing in our hospital is as follows.

Blood collection was performed by experienced phlebotomists following the recommendations of Clinical and Laboratory Standards Institute Collection of diagnostic venous blood specimen’s guideline (CLSI GP41 ED7) [6]. A gray-capped VACUETTE^® 4 mL FE Sodium Fluoride/K3EDTA tube (Greiner Bio-One, Kremsmunster, Austria Lot no: A211238A) was chosen as the blood collection tube. After blood collection, the tubes were centrifuged in a centrifuge device (NF1200, Nüve, Türkiye) at 1,500 g for 15 min, in accordance with the manufacturer’s recommendations. There is no consensus in the literature regarding the terminology used to describe alcohol concentration in blood. Terms such as blood alcohol concentration (BAC), plasma alcohol concentration (PAC), blood ethanol level (BEL), and plasma ethanol level are utilized for this purpose. In our study, we measured ethanol level from plasma samples, and we have used the term BEL for simplicity throughout our study. BEL were measured by using the Cobas 8,000 modular analyzer series–c702 module with manufacturers’ original reagent (Roche Diagnostics, Switzerland). The spectrophotometric method uses alcohol dehydrogenase as an enzymatic technique for measurement. Ethanol and nicotinamide adenine dinucleotide (NAD) are converted to acetaldehyde and reduced nicotinamide adenine dinucleotide (NADH) by alcohol dehydrogenase. According to the Roche technical report, the assay method used to determine the level of ethanol provides an analytical measurement range of 10.1–498 mg/dL. The report also declares that the assay guarantees a lower limit of quantification of 10.1 mg/dL, and a precision ranging between 0.8 and 2.4 %.

Internal quality control was performed using two levels of manufacturer-produced controls: Ammonia/Ethanol/CO2 Control N and Control A. External quality control was assessed using data obtained from the 9th and 10th cycles of Ammonia/Ethanol external quality control samples provided by Randox (Antrim, UK). To ensure the precision and dependability of our ethanol testing, we implement internal quality control on a daily basis and external quality control on a monthly basis. By analyzing the data collected from these studies, we were able to calculate the total analytical error (TAE) of our laboratory for the ethanol test using the following formula [7].

TAE: total analytical error.

CV: coefficient of variation.

IBM SPSS Statistics 22.0, Microsoft Excel and GraphPad Prism 4 for Windows (GraphPad Software, San Diego, California) programs were used in the study. As descriptive statistics, means (X), standard deviations (standard deviation SD), coefficients of variation (%CV) and percentage changes (%) between group averages were calculated. The Chi-square test was used to compare categorical variables. The rate of blood ethanol level reduction was calculated, and separate regression analyses were performed for males and females. For statistically significant differences, a p-value of <0.05 was considered significant.

Results

Table 1 displays the demographic data and summary statistics of the assessed parameters.

Age distributions are presented in Figure 1. Individuals between 30 and 50 years old are the most represented demographic in the study. The chi-square test results revealed a chi-square statistic (χ²) of 173.18 with a p-value <0.001 and degrees of freedom (df) equal to 5. These findings indicate a statistically significant difference in the distribution of positive and negative cases across age groups. Moreover, the proportion of positive cases increases with age, suggesting an age-related trend in positivity rates. The pairwise comparisons reveal which age groups differ significantly in terms of positive cases. Age group 0–10 is significantly different in positive cases compared to many other groups (e.g., 21–30, 31–40, 41–50). Age groups 21–30 and 31–40 are significantly different from groups ≥71 and 61–70. The smallest p-values indicate the most significant differences, such as between 0-10 and 21–30 or 21–30 and ≥71 years.

Figure 1:

Distribution of test results by age group with positive percentages displayed above the bars. Each bar represents the total count of tests within an age group, divided into three categories: Reject (red, representing samples that were not accepted for analysis), negative (light blue, representing negative test results), and positive (pink, representing positive test results). The percentages shown above the bars indicate the proportion of positive test results within each age group. Age groups are categorized as follows: 0–10, 11–20, 21–30, 31–40, 41–50, 51–60, 61–70, and ≥71 years.

The gender distribution of patients whose BEL were measured and the positivity rates by gender are displayed in Figure 2. Women and men showed similar percentages of positivity, with values of 16.6 and 17.0 %, respectively (x²=0.041, p=0.840).

Figure 2:

The bar chart illustrates the distribution of test results by gender. Each bar is divided into three segments representing reject (red, tests not accepted for analysis), negative (light blue for males, pink for females), and positive (dark blue for males, deep pink for females). The total height of each bar corresponds to the total number of tests performed for each gender. Percentages shown within the bars represent the proportion of the respective category (e.g., positive, negative, or reject) relative to the total number of tests for that gender. The categories are labeled as male (M) and female (F) on the x-axis. Gender distributions and positivity numbers by gender. (16.6 % males and 17 % females, positive results ≥10.1 mg/dL).

The measured BEL in patients with positive results are presented in Figure 3. It has been observed that the concentration of BEL is more apparent within a range of 100–300 mg/dl in individuals who have consumed ethanol.

Figure 3:

BEL distribution of the ethanol positive patients.

The data in Figure 4 shows the reasons for the rejection of ethanol testing, presented as a Pareto analysis. The rejection rate for the ethanol test was calculated to be 2.5 %. Hemolysis was identified as the main cause of rejection, accounting for 66.5 % of all rejections. Inappropriate transfer conditions, insufficient sample, and incorrect test requests were the other leading reasons for rejection, accounting for 14.2 , 4.6, and 4.6 % respectively.

Figure 4:

Pareto chart classifying test rejection reasons.

The graph in Figure 5 displays the number of test requests by time zone. The highest frequency of test requests occurs between 5:00 PM and midnight.

Figure 5:

Test request time scatter plot for ethanol testing.

According to the study on reordered ethanol tests, it was found that the repetition rate is 4.9 %. Additionally, the correlation between the repeated test results was evaluated, and the correlation coefficient was calculated as 0.998. The graphical representation of this analysis is shown in Figure 6. It has been observed that repeated tests in the laboratory analytical phase extend the Turnaround Time (TAT) by an average of 37 min.

Figure 6:

The correlation between the repeated BEL test results. *The two values shown on the graph in the linear range (>498 mg/dL) represent the results of repeated measurements conducted with dilution.

The decrease in ethanol levels was calculated for 7 female and 20 male patients who were admitted to or hospitalized in the hospital. The analysis revealed that the average rate of ethanol elimination was 16.8 mg/dL/h for females and 20.5 mg/dL/h for males. The rates of ethanol decrease for individual patients are presented in Figure 7. The graph illustrates that as the initial blood ethanol level (BEL) increases, males exhibit a faster rate of ethanol elimination compared to females (16.8 mg/dL/h for females and 20.5 mg/dL/h for males). Separate regression analyses were conducted for females and males based on the ethanol elimination rate data. In females, the regression analysis evaluating the relationship between elapsed time (minutes) and ethanol reduction (mg/dL) showed no statistically significant association. The model produced an R² value of 0.061, indicating that elapsed time had low explanatory power for ethanol reduction. The regression coefficient for elapsed time was 0.089 mg/dL per minute, suggesting a positive trend; however, this relationship was not statistically significant (p=0.281). The AIC and BIC values for the model were 234.1 and 236.2, respectively. Similarly, the regression analysis for males revealed no statistically significant association between elapsed time and ethanol reduction (R²=0.102, p=0.485). The regression coefficient for elapsed time was 0.0393 mg/dL per minute, indicating a slight positive trend; however, this relationship was not statistically significant. The adjusted R² value suggests that elapsed time accounted for only a minimal proportion of the variance in ethanol reduction.

Figure 7:

BEL decrease rates by gender..

These findings indicate that while males tend to have a higher rate of ethanol elimination than females, the time elapsed between blood draws does not significantly explain the observed changes in BEL in either group. Further investigations with larger sample sizes and additional variables are needed to better understand the factors influencing ethanol elimination rates.

Discussion and conclusion

Ethanol is not only the most widely consumed substance, but it is also the most extensively analyzed drug. BEL results and their appropriate statistical interpretation are crucial in clinical and forensic settings. Determining trends in ethanol consumption can provide valuable demographic data and contribute to the enhancement of relevant laboratory processes. One approach to achieve this goal is to assess the laboratory results. The Başakşehir Cam Sakura City Hospital in Istanbul is one of Türkiye’s largest healthcare centers, with millions of annual visits. Between 2020 and 2023, ethanol testing requests were made for 9,521 patients. Majority of requests were from emergency department patients (n=9,115), while the rest were from our toxicology service (n=376). In our study, the average age of evaluated patients was 32.42 years, with 85 % male individuals. In Türkiye, 22 % of driver’s license holders are female and 78 % are male [8]. When analyzing gender distribution in hospital applicants for ethanol testing, it’s important to consider that most apply for traffic-related issues. Relying solely on this distribution may lead to inaccurate assumptions about overall gender distribution of ethanol consumption. These data indicates that the demographics of patients presenting to the hospital due to alcohol use vary across different regions and cultures. In the literature, the gender distribution is reported to range from 69 to 95 % in favor of the male population [9], [10], [11].

In these studies, hospital admissions due to ethanol increased for ages 25–55, with variation based on population. According to Yılmaz et al., the gender distribution in their study was 89–11 % [12]. When we consider their findings along with ours, it becomes apparent that trends can vary even between two different cities of the same country. Our study’s positivity percentage was similar for males and females (16.6 and 17 % respectively). Only 1.94 % of positive results were below the legal limit of 50 mg/dl. The average BEL in positive individuals was calculated to be 169.9 mg/dl. Similarly, Klein et al. reported an average ethanol concentration of 213 mg/dL in their study [10]. These levels are associated with neurological findings such as confusion.

The testing process involves two stages: clinical and laboratory. The clinical stage begins with a test request, while the laboratory stage starts with accepting the sample in the laboratory. One of the most challenging aspects of working with ethanol testing is transporting the samples to the laboratory [13]. When our laboratory data is evaluated, it is observed that the average time covering the period between sample collection and sample acceptance (clinical stage) is 43 min, with a ‘Turn Around Time’ (TAT) within this range being 74 min. Our hospital has noticed that the clinical stage duration for ethanol samples is longer compared to the other emergency biochemistry group tests that undergo almost similar processes. The average time for the clinical stage of other emergency biochemistry group tests is 32 min, with 90 % of samples taking 53 min. It appears that transporting ethanol samples to the laboratory takes longer than expected due to the requirements for using a locked bag. Our hospital’s pneumatic system, which is mainly used for emergency samples, does not accommodate this, resulting in manual steps (such as manual placement of the sample in the bag, locking, hand delivery to transport personnel, and manual opening of the bag) that lengthen the process [14]. We propose that the production of locked pneumatic capsule systems could be a solution to expedite the process for hospitals that operate on a similar scale as ours. Similarly, the laboratory stage of ethanol analysis takes longer than other emergency biochemistry test groups. The test results for ethanol have an average turnaround time of 76 min, with 90 % of the tests completed within 125 min. This is significantly longer than other emergency biochemistry tests, which have an average turnaround time of 44 min, and 90 % are completed within 62 min. This difference in time is because the re-run rate for ethanol tests is 4.9 %, which is higher than other tests with less than a 1 % re-run rate. Furthermore, it has been noticed that ethanol test results often require more time for specialist approval. These factors indicate that ethanol analysis poses unique challenges despite being a reliable emergency test.

Regarding blood samples drawn for ethanol testing, it’s crucial to remember that they are not only medical specimens but also legal evidence, making the responsibility for test rejection even more significant. It is widely known that the majority of laboratory errors, approximately 65–70 %, occur during the pre-analytical phase of the testing process [15], 16]. Similarly, many test rejections also happen during this stage. Our hospital has observed a rejection rate of 2.5 % for ethanol analysis (test/sample). There is limited information in the literature about the reasons and percentages of test rejection due to preanalytical problems in blood BEL measurements. However, for comparison purposes, the rejection rates of other emergency biochemical tests typically range from 1.5 to 2.5 %. Ethanol tests are rejected more frequently than other tests in our laboratory, where the rejection rate for the urgent biochemical test group is 1.5 %. When analyzing the causes of rejections, it is found that most of the rejections are due to hemolysis. It has been reported that the concentration of plasma ethanol, when measured with an enzymatic assay, is lower in hemolyzed specimens [17]. A recent study found that between 10 and 18 % of total errors were caused by hemolysis [18]. Although hemolysis may not exceed the total allowable error limit of 20 % on its own, it is still essential to consider it as a significant variable along with other factors. Therefore, using auxiliary devices that can detect hemolysis at the patient’s bedside can prevent rejections caused by hemolysis [19].

Measuring the analytical performance of a test is crucial for medical laboratories. TAE is a key measure used to assess the overall performance of a laboratory test. TAE includes random errors (precision) and systematic errors (accuracy), providing a holistic view of the test’s repeatability and accuracy. Our calculated TAE value is 11.4 %, which falls within the allowable TAE of 20 % for ethanol testing in human blood according to Clinical Laboratory Improvement Amendments 2024 (CLIA ‘24) quality requirements [20]. Conventional spectrophotometric methods, which are easier to apply compared to sophisticated techniques like chromatography, ensure sufficient analytical performance. Our study showed a 99 % correlation based on assessing a reordered and retested sample, indicating high repeatability. Another study demonstrated high consistency among repeated results using a different brand of kit that employed a similar method [21]. Furthermore, delays in test results due to repetitions can extend Turnaround Time (TAT) values, leading to prolonged hospital stays in the emergency department. This highlights the importance of preventing unnecessary repeats.

The rate of ethanol elimination varies depending on factors such as age, gender, drinking habits, and body temperature, with reported values in the literature ranging from 10 to 30 mg/dL per hour [22], [23], [24]. In this study, the decline in BEL was monitored in 27 patients – 7 women and 20 men – admitted to the hospital. The analysis demonstrated a notable gender difference in ethanol elimination rates, with females showing an average decrease of 16.8 mg/dL per hour, while males exhibited a more rapid decline of 20.5 mg/dL per hour. Despite these differences, regression analyses performed separately for females and males revealed no statistically significant association between elapsed time and ethanol reduction in either group. These findings indicate that elapsed time alone is insufficient to explain the observed changes in BEL. The observed gender differences in ethanol elimination rates may reflect physiological variations, such as differences in alcohol dehydrogenase enzyme activity or other metabolic factors between genders. An additional consideration is that some individuals in this study may have been in the absorption phase of ethanol metabolism at the time of measurement. Ethanol absorption is influenced by multiple factors, including the quantity of alcohol consumed, concurrent food intake, age, gender, and history of chronic alcohol use.

These findings, obtained under real-time hospital conditions, contribute to the existing literature by highlighting individual variability in ethanol elimination and corroborating results from previous studies [12], 25], 26]. To better understand the factors influencing ethanol metabolism, further studies with larger sample sizes and the inclusion of additional variables are required to identify the factors influencing ethanol elimination rates.

Our research emphasizes the importance of accurate and timely ethanol testing, especially considering its significant implications in clinical and forensic settings. The results highlight the difficulties associated with sample transport and the need for repeated testing, indicating potential areas for profcess improvement in large healthcare facilities. The demographic information gathered from ethanol tests provides valuable insights into regional and cultural alcohol consumption patterns, showcasing the usefulness of such analyses in improving our understanding of public health trends and laboratory practices.