Factors associated with higher alcohol concentrations in emergency department presentations: PACE study
Abstract
Introduction
The health impact from alcohol is of recognised concern, from acute intoxication as well as increased risk of chronic health issues over time. Identifying factors associated with higher alcohol consumption when presenting to the emergency department (ED) will inform public health policy and enable more targeted health care and appropriate referrals.
Methods
Secondary testing of blood samples collected during routine clinical care of 1160 ED patients presenting to the Royal Brisbane and Women's Hospital in Queensland, Australia, for 10 days between 22 January and 1 February 2021. Alcohol was measured by blood ethanol (intake in recent hours) and phosphatidylethanol (PEth; intake over 2–4 weeks). Zero-inflated negative binomial regression was used to identify demographic and clinical factors associated with higher alcohol concentrations.
Results
Males were found to have 83% higher blood ethanol and 32% higher PEth concentrations than females (adjusted rate ratio [ARR] 1.83, 95% confidence interval [CI] 1.37–2.45 and ARR 1.32, 95% CI 1.04–1.68, respectively). Blood ethanol concentrations were 3.4 times higher for those 18–44 years, compared to those aged 65+ (ARR 3.40, 95% CI 2.40–4.82) whereas PEth concentrations were found to be the highest in those aged 45–64 years, being 70% higher than those aged 65+ (ARR 1.70, 95% CI 1.19–2.44). Patients brought in involuntarily had eight-times higher blood ethanol concentrations than those who self-attended.
Discussion and Conclusions
This study used two alcohol markers to identify factors associated with higher alcohol concentrations in emergency presentations. The findings demonstrate how these biomarkers can provide informative data for public health responses and monitoring of alcohol use trends.
Key Points
- The health impact from alcohol is of recognised concern. Understanding the factors associated with higher alcohol use using reliable biomarkers can inform patient-tailored interventions and pre-emptive patient management.
- Younger patients, males, injury-related presentations and those brought in by ambulance or involuntarily, were significantly more likely to have higher blood ethanol concentrations on emergency department presentation than the rest of the cohort.
- Higher phosphatidylethanol concentrations were identified for middle aged, males and in the days following the festive period highlighting a need for screening and assessment for harmful alcohol consumption, particularly in post-festive periods.
- Developing novel, minimally invasive methods for capturing alcohol involvement in emergency health presentations is a valuable contribution to the ongoing public health challenges of addressing alcohol-related harms in the community.
1 INTRODUCTION
Alcohol consumption at high levels is known to be associated with greater risk of injury from acute intoxication and disease from ongoing, frequent consumption [1, 2]. Patients with high levels of alcohol use are also known to experience greater risk of complications, increased care costs and higher mortality when being treated in acute care settings [3-5]. There is an acknowledged public health need to reduce the burden, social and economic costs, and consequences from harmful alcohol use [6, 7]. In 2020, the World Health Organization raised concerns regarding the harmful use of alcohol which causes approximately 3 million deaths per annum globally, and called for the development of an action plan (2022–2030) to address this as a public health priority [7]. Similarly, the National Alcohol Strategy 2019–2028 in Australia identified the need to prevent and minimise alcohol-related harm by reducing harmful levels of alcohol consumption [8]. In Australia between 2020 and 2021, more than half (57%) of drug-related hospitalisations were attributed to alcohol [2]. These targets and concerns are likely more critical given the impact of COVID-19, which has led to unprecedented job losses, lockdowns, work from home, isolation, increasing domestic violence and reports of increased alcohol sales and consumption [9, 10].
The age group at greatest risk of alcohol-attributable hospitalisations in Australia are those aged 35–64 years, followed by those aged 15–34 years [11]. In 2017, the leading three alcohol-attributable health conditions were neuropsychiatric conditions (e.g., alcohol dependence/abuse, alcoholic polyneuropathy), unintentional injuries (e.g., road traffic crashes, falls, poisoning) and digestive diseases (e.g., alcoholic gastritis, alcoholic liver cirrhosis), with males being hospitalised more than females for all three causes [11]. In an Australian multi-centre study, Miller et al. [12] analysed the 10 most common principal diagnosis codes (International Classification of Diseases, 10th Edition, Australian Modification, ICD-10-AM) from patients who had self-reported alcohol consumption in the previous 12 h. The leading three principal diagnoses blocks were mental and behavioural disorders due to psychoactive substance use, injuries to the head, and symptoms and signs involving the circulatory and respiratory systems. The sample from this study was skewed to be predominantly male across all age groups but more pronounced in the 35–44 and 55–64-year-old age groups. There will be patients presenting, however, who are not coded for alcohol intoxication or a secondary injury related to alcohol, and it is difficult to get a true prevalence within the limitations of hospital coding systems.
The abovementioned studies demonstrate that the hospital and emergency department (ED) are unique settings for collection of data that may identify the nature and extent of alcohol-related health harms. With EDs often being the first point of contact for alcohol-related issues, identifying the factors associated with ongoing higher alcohol consumption when presenting to the ED may enable early recognition of specific subgroups that are at higher health risk. This will allow for more tailored or targeted interventions for patients, or adjustments in clinical risk assessments. Despite these potential benefits, objective data are not routinely collected in Queensland, Australia, to determine the involvement of alcohol and other drugs in ED presentations [13], unless required by police for traffic-related incidents [14].
Unlike other tools for screening alcohol use (e.g., Alcohol Use Disorders Identification Test [15], self-screening questionnaires or indirect markers such as carbohydrate-deficient transferrin [16]), the combined use of direct markers of alcohol such as blood ethanol and phosphatidylethanol (PEth) may provide a more complete picture of alcohol consumption to better reflect both short-term and ongoing alcohol-use in patients. Blood ethanol has a short half-life in blood [17] and can be used to detect alcohol intake in recent hours. Blood ethanol is a standard measure when acute intoxication is the focus of investigation, however, investigations of long-term unhealthy and risk-associated alcohol consumption require alternative measures [18]. PEth is an abnormal phospholipid that accumulates with repeated exposure to alcohol [19] and blood levels indicate alcohol intake in the previous 2–4 weeks [20, 21]. PEth is a marker with high sensitivity [22], suitable for differentiating harmful use from moderate/average consumption, and considered more reliable than other indirect alcohol markers such as carbohydrate-deficient transferrin, gamma-glutamyl transpeptidase or alanine aminotransferase [16, 20, 23, 24]. Our results using blood ethanol and PEth to measure alcohol intake in ED presentations have shown that overall prevalence for a positive blood ethanol test of 9.3% increased four-fold to 22.2% for injury presentations [25]. The overall prevalence for a positive PEth test increased from 32.5% overall to 41.4% for injury presentations. Over 25% of cases that tested negative for acute blood ethanol showed PEth concentrations indicative of significant to heavy medium-term alcohol consumption [18, 25].
While the prevalence of blood ethanol and PEth positive samples have previously been reported from these data [25], this analysis aimed to determine the risk factors associated with higher PEth and ethanol concentrations in ED presentations. More specifically, to determine the clinical and demographic factors associated with increasing PEth and ethanol concentrations such as sex, age and clinical factors (e.g., primary diagnoses, triage category). Understanding the factors associated with higher ongoing alcohol use from reliable biomarkers can aid in clinical risk assessments and pre-emptive patient management, inform patient-tailored interventions, priorities for policy development and resource allocation.
2 METHODS
2.1 Study design and setting
The Prevalence of Alcohol Consumption in Emergency presentations (PACE) study involved secondary testing of blood alcohol concentrations from existing blood specimens of patients (N = 1160) presenting to the ED at the Royal Brisbane and Women's Hospital (RBWH) in Queensland, Australia, over a 10-day period. Further information about the study is published elsewhere [25, 26].
RBWH is a large metropolitan hospital and during 2020–2021 had an average of 393 patients presenting to the ED each day [27]. As part of routine clinical care, prior to COVID-19, whole blood samples were collected from approximately 40% of patients presenting to RBWH ED and subsequently sent to Pathology Queensland for testing [28].
During the data collection period, blood samples were intercepted, and an aliquot separated and frozen, without the patient or treating clinician's knowledge, to avoid any impact on patient care. Blood samples were not tested until 3–4 months after the data collection period ended to avoid any interference with patient management. The study team received the de-identified test results from Pathology Queensland and clinical data extracted from the ED Information System (EDIS).
2.2 Data collection
Blood samples were collected across the full 24 h between 22 January and 1 February 2021. The collection period included a 4-day national festive long weekend (Australia Day), when there were no COVID-19-related lockdowns in the local area and minimal public restrictions in place.
Demographic data, that is, sex, age and Indigenous status (identifying as Aboriginal and/or Torres Strait Islander peoples) and clinical data were extracted from EDIS. Clinical data included the principal diagnosis code (ICD-10-AM) assigned by the ED clinician (with EDIS including both standard ICD-10-AM diagnosis codes and a selection of ICD-10-AM external cause codes as a principal diagnosis), patient mode of arrival, departure status, triage code, presenting problem (recorded in free text), presenting date and time, and the patient's compensation status (health care attendance potentially compensable under a transport, workers or liability insurance scheme) [25, 26].
2.3 Biomarker testing
Blood ethanol concentration was measured in serum samples using an enzymatic assay, with a positive test above 1.6 mmol/L [29]. Blood ethanol concentrations can be grouped into clinically meaningful levels of consumption, with five levels: no/very light blood ethanol (<1.6 mmol/L), light blood ethanol (1.6 to <10.9 mmol/L), moderate blood ethanol (10.9 to <32.6 mmol/L), significant blood ethanol (32.6 to <65.3 mmol/L) and potential for coma or death (65.3 mmol/L or greater) [30].
PEth was measured in whole bloods using a validated ultra-performance liquid chromatography tandem mass spectrometry method used in the range of 0–2000 μg/L [22]. Of the various PEth molecule species or homologues, the homologue 16:0/18:1 had excellent correlation with total PEth (r2 = 0.98, p < 0.001) and was found to be the most dominant species, consistent with previous literature [23, 25, 31]. Hence, the homologue 16:0/18:1 was used as the biomarker to quantify PEth [21].
Based on the literature, PEth values were defined as: no/light consumption (<20 ng/mL), significant consumption (20–199 ng/mL); and heavy consumption (≥200 ng/mL). PEth is accumulated over repeated intake. PEth levels between 20 and 199 ng/mL could indicate an average moderate level of drinking or possible isolated recent binge drinking. However, heavy consumption (PEth levels of ≥200 ng/mL) indicates likely frequent heavy drinking with these levels unlikely due to isolated binge drinking [18, 21].
2.4 Ethics
Ethics approval was obtained by the RBWH Human Research Ethics Committee (LNR/2019/QRBW/56859). In accordance with section 2.3.10 of the National Statement on Ethical Conduct in Human Research, a waiver of consent was also obtained. Approval from the Director General of Queensland Health was received for the release of de-identified patient demographic and clinical data, under section 280 of the Public Health Act of Queensland.
2.5 Data management and analysis
Data cleaning and analyses were conducted using SPSS 27.0.1.0 and SAS 9.4. Missing data and outliers were cross-checked and verified with the original data sources by the relevant data custodians.
2.6 Data management
Age was categorised as <18 years, 18–44 years, 45–64 years and 65+ years. The data collection period was dichotomised for analysis, with the first 5 days being ‘festive’ and including the public holiday weekend, and the remaining days, also including a weekend, being ‘non-festive’. Triage codes range from 1 to 5, where 1 represents a life-threatening illness or injury requiring immediate attention and 5 represents a non-emergency health concern to be treated within 2 h [32]. Triage was further grouped for analysis (1–3 and 4–5). Mode of arrival included walk-ins, ambulance (road or air) or an Emergency Examination Authorities (EEA) via ambulance or police. An EEA is a legal mechanism by which a person may be taken involuntarily under Chapter 4A of the Public Health Act 2005, to a public health service facility, if their behaviour indicates they are at immediate risk of serious harm, which appears to be a result of a major disturbance in the person's mental capacity [33].
Based on the primary ICD-10-AM diagnoses of patients, the presentations were categorised into four primary diagnostic groups for analyses. (i) General medical—included diagnoses such as chest pain of various forms, abdominal pain, genitourinary and digestive issues, which included all ICD-10-AM codes in Chapters 1–4, 6–16, 18 and 21 (with the exception of codes K29.2, K29.7, R78.0, Z72.1) and included codes T80–T88; (ii) Injury—included all ICD-10-AM codes in Chapter 19, except for codes T80–T88 related to complications of surgical and medical care. The external cause codes captured in EDIS (W46, W54, W57 and X84) were also included in the injury categories (external causes of injury, e.g., needle, animals, insects); (iii) Mental health—ICD-10-AM codes included were in the range F01–F99 excluding F10.0 and F10.3; and (iv) alcohol-related—ICD-10-AM codes included were F10.0, F10.3, K29.2, R78.0, Y90 and Z72.1.
2.7 Data analysis
Zero-inflated negative binomial regression was used to quantify demographic and clinical factors associated with higher blood ethanol and higher PEth concentrations. Due to the high proportion of zeros and the skewed distribution of the data, examining the goodness of fit showed that the model estimated using zero-inflated negative binomial regression was the best fit for both the blood ethanol and PEth data. Zero-inflated negative binomial regression models provide a comprehensive analysis of data where excessive zeros and overdispersion exists, by combining two parts: (i) a binary component to handle excess zeros; and (ii) a count component to model the non-zero concentrations from the tests. Given blood ethanol concentrations were reported to one decimal and PEth concentrations to two decimal places, both were scaled to whole integers prior to analysis to more closely resemble count data.
Univariable analyses were conducted to assess the relationship between each demographic/clinical factor with blood ethanol concentration and PEth concentration separately. Factors that were statistically significant (p < 0.05) in univariable analysis were then entered into a multivariable model. Backwards elimination was used to remove non-significant factors and reduce the model to be the most parsimonious and the one with best fit. Adjusted rate ratios (ARR) and 95% confidence intervals (CI) were then calculated for the factors remaining in the final model.
3 RESULTS
During the data collection period, there were 3912 ED presentations. COVID-19 testing presentations were excluded from analysis (n = 872). Of the remaining 3040 ED presentations, this data collection intercepted 1160 patient samples (38.2%). This proportion aligns with the approximately 40% of patients expected to have bloods taken for clinical reasons based on hospital records [28], suggesting few were missed. From the 1160 samples, 12 were missing serum required for ethanol tests and 4 had insufficient whole blood for PEth testing.
The prevalence, baseline characteristics and sample representativeness of the study cohort have been previously reported [25]. In summary, while fewer than 10% of the cohort had a positive blood ethanol test (i.e., a concentration above 1.6 mmol/L), nearly one-third of the cohort had PEth levels indicative of significant to heavy alcohol consumption in the previous 2–4 weeks. The mean age of the cohort was 48 years (SD = 21). Just over half of the sample were female (54.3%), with 4.2% of the total sample identifying as Aboriginal and/or Torres Strait Islanders. Over half the sample presented to the ED via road or air ambulance (53.2%), 44.2% were walk-ins and 2.6% presented under an EEA via ambulance or police. Almost a third (29.7%) of the presentations had a triage priority 1–2, 40.9% priority 3 and 29.3% priority 4–5. About 79.4% of the presentations received a primary diagnosis in the general medical category which included chest or abdominal pain, genitourinary and digestive issues. Other presentations had injury (16.0%), mental health (3.2%) and alcohol-related (1.4%) primary diagnoses. The study sample included more females, more patients aged 45 years and above, and more patients triaged to a higher category, compared to ED presentations not included in the study [25].
3.1 Factors related to (i) a positive blood ethanol test and (ii) higher blood ethanol concentrations
Table 1 provides an overview of the frequency of positive blood ethanol tests (i.e., ≥1.6 mmol/L) and their median blood ethanol concentrations stratified by demographic and clinical factors. A higher percentage of males had a positive blood ethanol test than females (12.0% vs. 7.1%) and people aged 18–44 had the highest percentage of positive blood ethanol tests compared to people aged 65 years and over (12.8% vs. 3.4%) (Table 1). The percentage of injury-related presentations with a positive blood ethanol test was four times higher than the percentage of persons presenting for a general medical reason with a positive test (22.2% vs. 5.3%). For most factors, the highest median blood ethanol concentration among their categories would be regarded as significant (i.e., blood ethanol concentration between 32.6 and 65.3 mmol/L) (Table 1).
| Ethanol positive (≥1.6 mmol/L)a | PEth positive (≥20 ng/mL)b | |||
|---|---|---|---|---|
| N (%) | Median mmol/L (IQR)d | N (%) | Median ng/mL (IQR)e | |
| Sex | ||||
| Female (n = 630) | 44 (7.1) | 30.4 (18.2–45.3) | 161 (25.6) | 107 (42–289) |
| Male (n = 530) | 63 (12.0) | 33.8 (14.5–52.1) | 215 (40.8) | 139 (48–378) |
| Age groups, years | ||||
| <18 (n = 23) | c | c | 6 (26.1) | 58.5 (32–160) |
| 18–44 (n = 524) | 66 (12.8) | 32.2 (15.7–50.4) | 201 (38.5) | 118 (48–334) |
| 45–64 (n = 314) | 29 (9.3) | 40.0 (22.4–57.5) | 109 (34.7) | 175 (64–433) |
| 65+ (n = 297) | 10 (3.4) | 20.5 (8.7–32.3) | 59 (15.7) | 71 (37–241) |
| Missing = 2 | ||||
| Aboriginal and/or Torres Strait Islander | ||||
| No (n = 1106) | 94 (8.6) | 32.4 (16.8–51.4) | 351 (31.9) | 117 (46–329) |
| Yes (n = 49) | 9 (18.4) | 19.5 (13.4–40.0) | 21 (42.9) | 275 (56–436) |
| Missing = 5 | ||||
| Festive period | ||||
| Yes (n = 534) | 57 (10.9) | 35.0 (25.2–53.0) | 175 (32.8) | 124 (46–305) |
| No (n = 626) | 50 (8.0) | 23.4 (13.4–47.5) | 201 (32.3) | 119 (48–376) |
| Triage codef | ||||
| 1–3 (n = 820) | 88 (10.9) | 30.7 (15.8–51.6) | 282 (34.5) | 123 (44–338) |
| 4 and 5 (n = 340) | 19 (5.6) | 32.5 (19.5–54.9) | 94 (27.7) | 120.5 (52–331) |
| Mode of arrival | ||||
| Walk-in (n = 513) | 18 (3.5) | 23.0 (13.1–54.6) | 151 (29.5) | 85 (38–246) |
| Ambulance (road and air) (n = 617) | 77 (12.6) | 32.5 (18.7–51.4) | 207 (33.7) | 160 (50–457) |
| Emergency examination (ambulance/police) (n = 30) | 12 (41.4) | 30.3 (18.9–50.2) | 18 (60.0) | 226 (145–380) |
| Primary diagnosis categories | ||||
| General medical (n = 921)g | 48 (5.3) | 24.1 (11.5–42.5) | 271 (29.6) | 98 (42–277) |
| Injury (n = 186)h | 41 (22.2) | 34.8 (19.5–51.4) | 77 (41.4) | 179 (75–389) |
| Mental health (n = 37)i | c | c | 12 (32.4) | 320.5 (44.5–530.0) |
| Alcohol-related (n = 16)j | 13 (81.2) | 44.1 (39.4–56.6) | 16 (100.0) | 1149 (379–1471.5) |
| Departure status | ||||
| Not admitted (n = 451) | 37 (8.3) | 30.1 (14.5–51.1) | 145 (32.2) | 103 (43–285) |
| Admitted or transferred (n = 689) | 64 (9.4) | 31.6 (16.1–50.9) | 222 (32.4) | 142 (48–380) |
| Did not wait (n = 20) | 6 (30.0) | 49.9 (22.4–77.9) | 9 (45.0) | 153 (81–512) |
| Compensation status (work or motor vehicle) | ||||
| Non-compensable (n = 1131) | 103 (9.2) | 31.9 (16.3–51.8) | 366 (32.5) | 125 (48–338) |
| Compensable (n = 29) | c | c | 10 (34.5) | 63 (33–311) |
- Abbreviations: ICD-10-AM, International Classification of Diseases, 10th Edition, Australian Modification; IQR, interquartile range; PACE, Prevalence of Alcohol Consumption in Emergency department presentations; PEth, phosphatidylethanol.
- a Missing 12 blood ethanol tests.
- b Missing 4 PEth tests.
- c Information suppressed where cell counts ≤5.
- d Positive blood ethanol categories: Light blood ethanol (1.6 to <10.9 mmol/L), Moderate blood ethanol (10.9 to <32.6 mmol/L), Significant blood ethanol (32.6 to <65.3 mmol/L) and Potential for coma or death (65.3 mmol/L or greater).
- e Positive PEth categories: Significant consumption (20–199 ng/mL), and Heavy consumption (>200 ng/mL).
- f Triage codes 1–5. 1 = life-threatening illness or injury requiring immediate attention, 5 = non-emergency health concern.
- g All ICD-10-AM codes in chapters 1–4, 6–16, 18 and 21 (with the exception of codes K29.2, K29.7, R78.0, Z72.1) and included codes T80-T88.
- h All ICD-10-AM codes in chapter 19 were included with the exception of codes T80-T88 related to complications of surgical and medical care. Codes W46, W54, W57 and X84 were also included in the injury categories (external causes of injury i.e., needle, animals, insects).
- i ICD-10-AM codes included were F10.0, F10.3, K29.2, R78.0, Y90 and Z72.1.
- j ICD-10-AM codes included were in the range F01-F99 excluding F10.0 and F10.3.
Univariable analyses found that males, people aged 18–44 and 45–64 years, and presentations occurring during the national Australia Day holiday weekend were significantly associated with higher overall concentrations of blood ethanol (p < 0.01). Arrival mode, triage priority, departure destination and primary diagnostic reason for the ED presentation were also significantly associated with higher overall blood ethanol concentrations (p < 0.01). Indigenous status (p = 0.21), and whether the presentation was potentially compensable under a transport, workers or liability insurance scheme (p = 0.71), were not significantly associated with overall blood ethanol concentrations. Collinearity existed between the mode of arrival and triage priority, and mode of arrival and departure status. As such, only mode of arrival was considered in the multivariable analyses. Using backwards elimination, the festive period no longer contributed significantly to the multivariable model and was removed. The final multivariable model, therefore, included sex (male/female), age (<18, 18–44, 45–64 and 65+ years), primary diagnostic group (general medical, injury, mental health, alcohol) and mode of arrival (walk-in, ambulance, EEA via ambulance/police).
On average, males were found to have 83% higher blood ethanol concentrations than females (ARR 1.83, 95% CI 1.37–2.45), Table 2. Blood ethanol concentrations were found to be the highest for those aged 18–44 years, with concentrations 3.4 times higher than those aged 65+ years (ARR 3.40, 95% CI 2.40–4.82), followed by those aged 45–64 years with 3.12 times higher concentrations than those aged 65+ years (ARR 3.12, 95% CI 2.13–4.56), Table 2. ED presentations with an alcohol-related primary diagnosis were found to have over 16 times higher blood ethanol concentrations (ARR 16.10, 95% CI 5.04–51.47) than those with a general medical diagnosis (Table 2). Injury-related ED presentations had blood ethanol concentrations four times higher than presentations for a general medical reason (ARR 4.00, 95% CI 2.71–5.89). Patients brought in by ambulance had 2.26 times higher blood ethanol concentrations than those who self-presented to ED (ARR 2.26, 95% CI 1.68–3.03).
| Demographics and clinical factors | N (%) | Adjusted rate ratioa | 95% confidence interval |
|---|---|---|---|
| Sex | |||
| Female | 624 (54.4) | Reference | |
| Male | 524 (45.6) | 1.83* | (1.37–2.45) |
| Ageb, years | |||
| 65+ | 295 (25.7) | Reference | |
| 45–64 | 313 (27.3) | 3.12* | (2.13–4.56) |
| 18–44 | 515 (44.9) | 3.40* | (2.40–4.82) |
| <18 | 23 (2.0) | 2.62 | (0.95–7.23) |
| Primary diagnosis categories | |||
| General medicalc | 911 (79.4) | Reference | |
| Injuryd | 185 (16.1) | 4.00* | (2.71–5.89) |
| Mental healthe | 36 (3.1) | 1.84 | (0.82–4.13) |
| Alcoholf | 16 (1.4) | 16.10* | (5.04–51.47) |
| Mode of arrival | |||
| Walk-in | 508 (44.3) | Reference | |
| Ambulance (road or air) | 611 (53.2) | 2.26* | (1.68–3.03) |
| EEA via ambulance/policeg | 29 (2.5) | 8.21* | (3.34–20.16) |
- Abbreviations: EEA, Emergency Examination Authorities; ICD-10-AM, International Classification of Diseases, 10th Edition, Australian Modification; PACE, Prevalence of Alcohol Consumption in Emergency department presentations.
- a Zero inflated negative binomial regression.
- b Missing = 2.
- c All ICD-10-AM codes in Chapters 1–4, 6–16, 18 and 21 (with the exception of codes K29.2, K29.7, R78.0, Z72.1) and included codes T80-T88.
- d All ICD-10-AM codes in Chapter 19 were included with the exception of codes T80-T88 related to complications of surgical and medical care. Codes W46, W54, W57 and X84 were also included in the injury categories (external causes of injury i.e., needle, animals, insects).
- e ICD-10-AM codes included were in the range F01-F99 excluding F10.0 and F10.3.
- f ICD-10-AM codes included were F10.0, F10.3, K29.2, R78.0, Y90 and Z72.1.
- g EEA is a legal mechanism by which a person may be taken involuntarily to a public health service facility for medical assessment.
- * p < 0.05.
3.2 Factors related to (i) a positive PEth test and (ii) higher PEth concentrations
Table 1 provides an overview of the frequency of positive PEth tests (i.e., ≥20 ng/mL) and their median PEth concentrations stratified by the demographic and clinical factors. A higher percentage of males had a positive PEth test than females (40.8% vs. 25.6%), and people aged 18–44 had the highest percentage of positive PEth tests compared to people aged 65 years and over (38.5% vs. 15.7%). People involuntarily brought in for an emergency examination, and those with mental health or alcohol-related diagnoses had the highest PEth concentrations, with the medians all in the heavy consumption category (≥200 ng/mL).
Univariable analyses found males, people aged 18–44 years and presentations occurring during the non-festive period were significantly associated with higher overall PEth concentrations (p < 0.05). Arrival mode and primary diagnostic reason for the presentation were also significantly associated with higher PEth concentrations (p < 0.01). Neither triage code (p = 0.30) nor departure status (p = 0.07) were associated with higher PEth concentrations. Indigenous status (p = 0.48), and whether the presentation was potentially compensable under a transport, workers or liability insurance scheme (p = 0.11), were also not statistically significantly associated with higher PEth concentrations and were therefore not considered for the multivariable regression model. The final multivariable model for PEth, therefore, included sex, age, festive period, primary diagnostic group and mode of arrival.
On average, males were found to have 32% higher PEth concentrations than females (ARR 1.32, 95% CI 1.04–1.68), Table 3. PEth concentrations were found to be the highest in those aged 45–64 years, being 70% higher than those aged 65+ (ARR 1.70, 95% CI 1.19–2.44), followed by those aged 18–44 years (ARR 1.60, 95% CI 1.15–2.23 against those aged 65+), Table 3. On average, patients who attended the ED in the latter period of data collection, after the festive days, had PEth concentrations 38% higher than patients who attended the ED during the festive period (ARR 1.38, 95% CI 1.09–1.76). ED presentations with an alcohol-related primary diagnosis were found to have four times higher PEth concentrations (ARR 4.04, 95% CI 2.19–7.45) than those with a general medical diagnosis (Table 3). Patients brought in by ambulance had 1.67 times higher PEth concentrations than patients who self-presented to ED (ARR 1.67, 95% CI 1.30–2.16).
| Demographics and clinical factors | N (%) | Adjusted rate ratioa | 95% confidence interval |
|---|---|---|---|
| Sex | |||
| Female | 629 (54.4) | Reference | |
| Male | 527 (45.6) | 1.32* | (1.04–1.68) |
| Ageb, years | |||
| 65+ | 295 (25.6) | Reference | |
| 45–64 | 314 (27.2) | 1.70* | (1.19–2.44) |
| 18–44 | 522 (45.2) | 1.60* | (1.15–2.23) |
| <18 | 23 (2.0) | 0.62 | (0.22–1.71) |
| Festive period | |||
| Yes | 534 (46.2) | Reference | |
| No | 622 (53.8) | 1.38* | (1.09–1.76) |
| Primary diagnosis categories | |||
| General medicalc | 917 (79.3) | Reference | |
| Injuryd | 186 (16.1) | 1.03 | (0.77–1.38) |
| Mental healthe | 37 (3.2) | 1.35 | (0.68–2.70) |
| Alcoholf | 16 (1.4) | 4.04* | (2.19–7.45) |
| Mode of arrival | |||
| Walk in | 512 (44.3) | Reference | |
| Ambulance (road or air) | 614 (53.1) | 1.67* | (1.30–2.16) |
| EEA via ambulance/policeg | 30 (2.6) | 1.30 | (0.71–2.36) |
- Abbreviations: EEA, Emergency Examination Authorities; ICD-10-AM, International Classification of Diseases, 10th Edition, Australian Modification; PACE, Prevalence of Alcohol Consumption in Emergency department presentations; PEth, phosphatidylethanol.
- a Zero inflated negative binomial regression.
- b Missing = 2.
- c All ICD-10-AM codes in Chapters 1–4, 6–16, 18 and 21 (with the exception of codes K29.2, K29.7, R78.0, Z72.1) and included codes T80-T88.
- d All ICD-10-AM codes in Chapter 19 were included with the exception of codes T80-T88 related to complications of surgical and medical care. Codes W46, W54, W57 and X84 were also included in the injury categories (external causes of injury i.e., needle, animals, insects).
- e ICD-10-AM codes included were in the range F01-F99 excluding F10.0 and F10.3.
- f ICD-10-AM codes included were F10.0, F10.3, K29.2, R78.0, Y90 and Z72.1.
- g EEA is a legal mechanism by which a person may be taken involuntarily to a public health service facility for medical assessment.
- * p < 0.05.
4 DISCUSSION
This study aimed to identify the clinical and demographic factors associated with higher acute and ongoing harmful alcohol concentrations, with the goal to inform public health policy responses and identify key ‘at risk’ groups of patients, to enable better screening for more targeted health care provision.
Consistent with the literature [11, 12, 34], this study found that males and people of younger age had the highest blood ethanol concentrations, with males having 83% higher concentrations than females, and younger age groups having more than three times higher concentrations than those aged 65+ years. Blood ethanol concentrations at the time of presentation in patients brought in by ambulance or involuntarily for emergency examination were between two- and eight-fold higher than in those who self-attended. Patients presenting with an injury had four times higher blood alcohol concentrations than general medical patients. Patients diagnosed as an alcohol-related presentation had significantly higher blood ethanol concentrations than patients with a medical, mental health or injury diagnosis, which is unsurprising as those whose principal reason for presentation was due to alcohol would be expected to have a higher blood ethanol content than those with other principal reasons for presentation. However, contrary to expectation, in the adjusted regression model, higher blood ethanol concentration was not statistically significantly associated with presentations to the ED over the festive long weekend.
In essence, younger patients, males, injury-related presentations and those brought in by ambulance or involuntarily, were significantly more likely to have higher blood ethanol concentrations on ED presentation than the rest of the cohort. These findings are consistent with one study that found that younger adults who screened positive for blood alcohol on hospital admission were more likely to have trauma-related presentations [34]. Whereas, middle and older age patients with positive blood alcohol on admission were more likely to have alcohol-related diagnosis and greater numbers of underlying comorbidities [34]. In the current study, alcohol consumption in the previous 2–4 weeks (as determined by PEth concentration) was highest for males and those middle-aged (45–64 years) and 18–44 years. Diagnostic reason for presentation, with the exception of alcohol-related presentations, was not significantly associated with higher PEth concentrations, and neither was being brought to hospital involuntarily. While older patients were more likely to arrive by ambulance than other age-groups, the majority of people arriving by ambulance with levels of PEth indicative of heavy consumption were below the age of 65 years.
The current study found that PEth concentrations of patients presenting in the days after the festive period were 38% higher than in the previous festive days. PEth is accumulated over repeated intake and has a half-life of 4 to 10 days but can continue to be detected in blood up to 4 weeks after alcohol has been eliminated from the body [35]. Thus the higher PEth concentrations in the 5 days directly following the festive period are likely to reflect higher alcohol consumption over the festive period, but not presenting to the ED until the days after.
4.1 Strengths and limitations
This study has uniquely used two alcohol markers (blood ethanol and phosphatidylethanol) to identify risk factors for acute and frequent high alcohol consumption in ED patients. Blood ethanol alone will only reflect very recent alcohol consumption and not necessarily identify persons with frequent harmful consumption. The findings offer a new insight to the wider impact of alcohol on emergency care. PEth testing as part of routine clinical care is currently prohibitive due to costs, specialist equipment and expertise required, but with periodic surveillance, can provide informative data for public health responses and monitoring of alcohol use trends. However, it is anticipated that in the future, such screening test will become day-to-day practice, with growing evidence of its utility to inform clinical management of patients in a variety of health care contexts [36-39]. Recent studies have also demonstrated the successful use of dried blood spots for PEth testing which is minimally invasive and reduces some costs associated with collection and storage [40].
While this study was conducted at one of the largest public tertiary hospital EDs in Queensland, it is only one metropolitan site and findings may not be generalisable. In addition, approximately 40% of all presentations to ED had routine blood samples collected. The study sample differed by sex, age and triage category to the rest of the presentations, also potentially affecting the generalisability of the findings. The study has demonstrated the feasibility of this research, and next steps will be to expand the data collection to multiple sites including those outside metropolitan areas and to other state jurisdictions, increasing the sample size and strengthening generalisability. As the use of PEth testing remains limited, there are no Australian population norms for PEth levels which is needed for proper control comparisons.
5 CONCLUSION
These findings demonstrate that both acute and frequent alcohol consumption is associated with clinical and demographic factors for emergency presentations. Importantly, presentations showed higher PEth concentrations in the days following the festive period highlighting a need for screening and assessment for harmful alcohol consumption, particularly in post-festive periods, even if the patient is not under the influence of alcohol at the time of presentation. Understanding the factors associated with higher ongoing alcohol use can aid in clinical risk assessments and pre-emptive patient management of alcohol withdrawal, inform patient-tailored brief interventions to reduce future harm for the individual, as well as inform resource and skills allocation so that optimal care can be provided to patients. Developing novel, minimally invasive methods for capturing both acute and frequent harmful alcohol involvement in emergency health presentations is a valuable contribution to the ongoing public health challenges of addressing alcohol-related harms in the community.
AUTHOR CONTRIBUTIONS
Each author certifies that their contribution to this work meets the standards of the International Committee of Medical Journal Editors.
ACKNOWLEDGMENTS
This study was internally funded by the Jamieson Trauma Institute, Metro North Health. No external funding has been received. Open access publishing facilitated by Queensland University of Technology, as part of the Wiley - Queensland University of Technology agreement via the Council of Australian University Librarians.
CONFLICT OF INTEREST STATEMENT
None to declare.
