^■ Introduction and background for ER alcohol and injury studies

Although a relatively large body of literature on alcohol’s role in injury occurrence existed internationally prior to the latter part of the 20^th Century, most studies focused on specific types of injury due to motor vehicle collisions and on fatalities rather than on non-fatal injuries (reviewed in [1, 2]). The existing literature suggested that injuries were more likely to occur among those with a history of “alcoholism” (alcohol dependence) or heavy drinking and the potential risk of injury might be heightened by an even moderate consumption of alcohol [3]. The data linking alcohol to injury derived from reports of trauma histories among “alcoholics” (those with alcohol dependence) or heavy drinkers or case reports of specific causes of injury though good epidemiological studies were largely lacking. One notable exception was a large-scale study of alcohol across the whole range of emergency room (ER) injury cases in the US [4] that found 22% of those admitted to the ER for an injury were breathalysed positive for alcohol.

An ER study was undertaken at a large trauma centre in San Francisco, US [5] because of the lack of large rigorous epidemiological studies on alcohol’s role across the entire range of injuries, serious injuries being relatively rare events in the general population and ERs (and their equivalent across cultures) being the primary entry point for events of this kind. The purpose of this study was not only to expand our understanding of the role of alcohol in injury occurrence and in the risk of injury, but also to serve as a pilot study for evaluating the feasibility and refining methodology for subsequent ER studies since the nature of ER visits and the milieu of the ER could create circumstances under which obtaining data and maintaining study integrity might prove difficult. Additionally, studies across countries and settings needed to rely on common indicators and definitions of alcohol consumption, injury type, demographics and context-specific variables among other factors to be comparable and lead to a better global understanding of the relationship between alcohol and non-fatal injuries.

This first study served as a model for subsequent ER studies globally (detailed in [6] as the ‘Cher­pitel Model’). A probability sample of all patients was drawn, equally representing all times of the day and days of the week. (Subsequently, most of the ER studies limited sampling to injured patients only.) Blood-alcohol levels of patients who consented to participate in the study were determined by a breathalyser as soon as possible following admission to the ER, and patients were interviewed for 20 to 30 minutes. Patient interviews elicited data on the reason for the ER visit (injury or non-injury), the type and cause of injury, drinking prior to the event, feeling ‘drunk’ at the time, attributing a causal association between drinking and the event, quantity and frequency of usual drinking and of higher consumption times (heavy episodic drinking), symptoms of alcohol use disorders based on ICD-10 and DSM-IV criteria for alcohol dependence and harmful use of alcohol/alcohol abuse prior ER visits and demographic characteristics.

As the ER studies expanded to other countries and cultures, contextual-level data were added including ER type (trauma centre), legal drinking age, legal level of intoxication while driving, level to which alcohol use is stigmatised and country-level detrimental drinking pattern (DDP). DDP was developed by the World Health Organization (WHO) from country-level surveys and key informant data across 51 countries based on indicators of alcohol’s integration in society [7] and included indicators of heavy drinking occasions (quantity per occasion, proportion of daily drinking, proportion of drinking occasions resulting in drunkenness, proportion of festive drinking occasions), drinking with meals and drinking in public places.

The International Alcohol Policy and Injury Index (IAPII) [8] was later added to the array of contextual-level variables to predict alcohol-related injury. This measure was developed on cross-sectional data (validated by member states and included in the Global Information System on Alcohol and Health, GISAH, of WHO) from 156 member countries in the WHO 2012 Global Survey on Alco­hol and Health and included ten policy items grouped into four regulatory domains: 1) physical avail­ability – legal minimum drinking age, government monopoly of retail sales, restrictions on outlet density, restrictions on outlet hours and days of operation, 2) road specific – random breath testing, legal blood alcohol concentration (BAC) limits, penalties for exceeding the maximum BAC, 3) advertising/promotion – a composite measure of restrictions on the majority of media advertisements and 4) drinking context – community mobilisation programmes, mandatory server training. Based on these domains, a composite indicator was developed to assess the association between policy and deaths due to alcohol-related injury, controlling for total per capita consumption, unlike other alcohol policy measures which predict consumption or patterns of drinking.

Following the San Francisco ER study, the first international conference on alcohol and injury was held in Scarborough, Ontario, sponsored by the US National Institute of Alcohol Abuse and Alcoholism (NIAAA), WHO, the Pan American Health Organization (PAHO) and several Canadian organisations [9]. This conference served as a springboard for several subsequent ER studies, all following the protocol established in the San Francisco study and included Mexico, Spain, Italy, Canada, Australia, Poland and Argentina. These studies, together with additional ER studies in the US included 32 ERs in 8 countries and was called the Emergency Room Collaborative Alcohol Analysis Project (ERCAAP).

In 2000, WHO elected to conduct ER studies in 12 countries following the established methodology, which led to an international conference on new knowledge on alcohol and injury from ER studies in Berkeley, California in 2005 and included 50 participants from 15 countries [10]. In 2008 PAHO initiated ER studies of alcohol and injury in eight countries in Central and South America and the Caribbean. In 2018, in collaboration with the Interamerican Development Bank (IDB), PAHO initiated a second set of ER studies on alcohol, drug use and road traffic injuries (which added to the protocol oral fluid testing for drug use). The US NIAAA intramural programme also supported ER studies in several Asian and African countries as did other governmental funders. The final compilation of these studies along with the ERCAAP studies resulted in a data set of over 34,000 injured ER patients from 101 ER sites in 33 countries covering 6 continents and is known as the International Collaborative Alcohol and Injury Study (ICAIS).

^■ Logistics and implementation of ER studies in a multi-cultural context

While implementing studies using a common protocol was necessary for sound epidemiological
cross-cultural comparisons, it also faced many organisational and logistic difficulties. Ethical approval of the common protocol was needed for all countries, and depending on the country, appro­val at the country level and at the facilities where the study would be implemented was also required, in addition to approval through the organising institution (university, research centre or Ministry of Health). Each ethical submission could give rise to new questions which needed to be addressed before the study was able to begin. Standardisation of training of investigators on the protocol and procedures, and for the sampling of ERs and their location was also necessary, which required coordination between trainers, funding organisations (PAHO, WHO, NIAAA and others) and the local teams. The selection and use of breathalysers and oral fluid tests were also standardised; central purchasing was implemented to obtain better prices as was the need to address export and import rules and taxes. Each country also needed to adapt a common protocol for eliciting country-specific data on local alcohol beverage type, definition of a standard drink, permitted-drinking locations, local and national policies related to drink driving and other contextual-level data required to enable later comparisons. Identification of funding and issuing of contracts for the principal investigator in each country along with transferring funds to the appropriate WHO/PAHO country office or other institution depending on the support for the study was also required. The coordination needed across the country researchers, WHO/PAHO offices or other institutions and trainers added to the complexity of logistics for a timely implementation of each ER study. Each dataset had to be reviewed, checked and then integrated into a global dataset for comparisons but also returned to the country team for their own analyses and report writing. Authorship guidelines also needed to be developed and shared with each participating country prior to the initiation of data collection.

Implementation of the studies at each site also faced challenges. Some were common across sites and intrinsic to the nature of ER studies, while others were idiosyncratic and related to each site’s cultural and socio-economic context. Among the first set of challenges were difficulties in integration of study procedures into the fast-paced ER environment with urgent clinical care needs taking priority. Research procedures therefore had to be flexible while maintaining data integrity. Collaboration from ER and hospital staff at all levels from administrative to clinical was also critical. Regarding the second set of challenges, some were related to the cultural, legal and socio-economic milieu of each site. For instance, different national and local norms regarding research involving human participants, as well as the legal implications of individual data related to alcohol and/or other substance use at the time of injury, made each administrative and ethical approval process highly site-specific. The same applied to the acquisition of breathalysers and oral fluid tests as some sites had to undergo complex import procedures. In low- and middle-income settings, hospitals or ERs faced greater difficulty accommodating research activities due to limited space and resources. In high-income settings, interviews could often be carried out in private rooms, whereas at other sites they were conducted in busy hallways, making privacy a challenge. Additional economic constraints affected funding for logistics, data management and staff compensation.

^■ Differential findings and dissemination of findings from ER studies in a multi-cultural context

The most important outcome of interest ana­lysed in the ICAIS data was alcohol-related injury, defined as a patient reporting drinking within six hours prior to the injury event. The data across multiple countries and cultures provided an opportunity to analyse contextual variables which might influence validity of self-reported alcohol consumption, a topic which has been of ongoing interest with mixed findings across alcohol-use studies. When self-reports were compared to breathalyser readings, validity averaged 92%, ranging from 85% for drivers to 95% for those with violence-related injuries. Validity was found to be highest for high IAPII (more restrictive alcohol policy) countries, ranging from 92% to 99%, while countries with high DDP scores (more DDP) and those with lower IAPII scores (less restrictive alcohol policy) were more likely to deny drinking while registering positive in breathalyser testing [11].

Using multi-level analysis and controlling for demographic characteristics, individual-level drink­ing and study-level volume, the IAPII was found to be significantly associated with the proportion of both self-reported drinking prior to the injury (p < 0.001) and the patient attributing a causal link between their drinking and the injury (p < 0.01), with DDP having little effect on these associations [12]. Each of the four IAPII domains were significantly predictive of self-reported drinking, while only the road domain was significantly predictive of causal attribution, suggesting that the more restrictive the alcohol policy in a country, the lower the rate of alcohol-related injury, with country-level drinking pattern having little effect on this relationship.

A second multi-level analysis of cause of injury found that while both frequent and infrequent heavy drinking were strong predictors of injuries related to violence and falls, DDP was a significant predictor only of traffic (OR = 1.54) and violence-related injuries (OR = 1.38), but lost significance when the IAPII was included [13]. IAPII was a significant predictor only for traffic injury (OR = 0.97) and each domain, with the exception of context, was also significant, suggesting the importance of targeted policies for specific causes of injury.

Additional multi-level analyses examined whether the societal context of country-level frequency of drinking was associated with rates of alcohol-related injury in ER studies from the same countries. The societal context of drinking variable was based on questions on the frequency of drinking in public contexts and the frequency of drinking in private contexts, taken from the multi-country Gender, Alcohol and Culture: An International Study (GENACIS) survey [14]. Controlling for demographic characteristics, individual-level volume and drinking pattern, study-level volume, DDP and the IAPII, societal public drinking context was significantly predictive of an alcohol-related injury (OR = 1.08) while societal private drinking context was not [15]. By injury cause, public drinking context was predictive of violence-related injury (OR = 1.09) while private drinking context was predictive of injuries from falls (OR = 1.01), and neither context was predictive of traffic-related injuries.

Using case-crossover analysis, the effects of DDP and IAPPI on the dose-response relationship of alcohol and injury were examined for specific causes of injury. Risk was higher at all volume levels in higher DDP countries compared to lower DDP countries and for each cause of injury [16]. Risk of a traffic injury was significantly greater in higher than lower DDP countries at 3.1-6 drinks (OR = 2.65), and at ≤ 3 drinks for falls (OR = 2.51) and injuries from other causes (OR = 1.72). Countries with more restrictive alcohol policy were at a lower risk of injury at lower levels of consumption (≤ 3 drinks) for all injuries (OR = 0.72), and at a lower risk of traffic injuries at higher levels of consumption (≥ 10 drinks). At higher levels of consumption, countries with greater alcohol policy restrictiveness were at greater risk for injuries from violence (OR = 9.02) and those from falls (OR = 4.29). These findings suggest that countries with high DDP are at higher risk of injury from most causes at a given level of consumption, while countries with low restrictiveness of alcohol policy are at higher risk of injury at lower levels of consumption and at higher risk of traffic injuries at high levels of consumption.

Beyond challenges in implementation of the studies, dissemination and visibility of these findings posed additional challenges, closely linked to the socio-political sensitivity of the topic in a country and the available resources. On the one hand, cultural attitudes toward alcohol – whether drinking is stigmatised or socially normalised – influenced how results were received. On the other hand, local teams’ resources varied considerably, including technical expertise, personnel availability, and institutional capacity to translate evidence into policy recommendations. To enhance dissemination of results and policy recommendations, in addition to the multitude of scientific papers published from these data, two books were produced, one edited by WHO [10] and the other by PAHO [17] as well as a fact sheet produced by PAHO. The role of multinational research initiatives such as ICAIS, and of agencies such as WHO and PAHO, was therefore essential, not only for fostering cross-national collaborative research and building research capacity, but also for helping to reduce disparities in the dissemination and visibility of results across settings.

^■ Outcomes of ER studies in a multi-cultural context

While these findings on the association of aggregate-level societal factors with alcohol-related injury across multiple countries and cultures underscore the importance of data of this kind in better understanding the alcohol-injury relationship and the development of effective intervention and prevention strategies for reducing alcohol-related injury, these data have also contributed to a broader agenda of alcohol studies. In calculating risk of injury from alcohol and differential risk by gender, these data have been instrumental in determining the attributable fraction of alcohol to injury morbidity which informed the Global Burden of Disease estimates for establishing safe drinking guidelines. While risk of injury was found to be similar for females and males at ≤ 3 drinks prior to injury (OR = 2.74 vs. 2.76 respectively), at higher volume levels females were at significantly greater risk than males [18]. Consuming between 3.1-6 drinks prior to injury put females at an 8-fold increased risk of injury (OR = 8.29) compared to a 5-fold increased risk for males (OR = 4.99), while consuming over 6 drinks put females at double the risk of males (OR = 14.69 vs. 7.38 respectively). Based on risk of injury morbidity from drinking, these data suggest that females are at a higher risk than males at a given level of consumption and this should be taken into consideration in recommending gender-specific safe drinking guidelines.

These data have also been used to inform the two diagnostic schemes used globally to identify alcohol use disorders, the Diagnostic and Statistical Manual of Mental Disorders, 4^th Edition (DSM-IV) [19] and the International Classification of Diseases and Related Health Problems 10^th edition (ICD-10) [20]. To inform the 5^th revision of the DSM, the ICAIS data were analysed combining criteria within the dependence and abuse domains, removing the criteria assessing legal problems and adding a criterion to assess alcohol craving that had not previously been part of the DSM but was a criterion in the ICD for alcohol dependence [21, 22]. In the absence of the criterion assessing legal problems alcohol abuse and dependence were found to form an unidimensional continuum [23] while the craving criterion did not contribute significantly to identification of those with AUD (alcohol use disorder) for any country [24].

To inform the 11^th revision of the ICD in developing a new diagnostic category of a “single epi­sode of harmful drinking,” the ICAIS data were analysed to evaluate the extent to which patients admitted to the ER with an alcohol-related injury would go undetected if signs of intoxication or withdrawal, a positive BAC, or signs of alcohol dependence or harmful drinking were relied upon as indicators of alcohol involvement in the injury event. While 18.8% reported drinking in the six hours prior to injury, and 47.1% of these attributed a causal association of their injury to drinking, 6.3% of those reporting drinking and 10.3% of those attributing a causal association were negative for dependence or harmful drinking [25]. The vast majority of both groups reported no heavy (5+) drinking episodes during the last year and had a BAC of < 0.05 or negative on ER admission, suggesting that some individuals may have alcohol-attributable harm to health due to an episode of drinking without a history of harmful use or dependence. These data underscored the clinical and public health relevance of a new diagnostic category of a single episode of alcohol use resulting in harm as proposed in the 11^th revision of the ICD-11.

^■ Summary and conclusions

The facilitating pattern of international coopera­tion for this research was a common protocol that could be adapted at country level to fit indivi­dual norms and culture (in relation to alcohol use, drink sizes, etc.) and demographics (education, religion, etc.), training and coordination provided by international experts and coordination by the involved governmental and other agencies with local researchers including ethical approval processes. The centralisation of the datasets allowed for integration into a single database researchers could use as well as provided access to their own cleaned dataset for country reports, analyses and publications. Multiple funding sources from the institutions involved also facilitated the work rather than the project relying on a single source of funding.

The framework for such research cooperation as this has recently become more difficult with funding cuts at PAHO/WHO and the NIH (National Institutes of Health), as well as reductions in staff at global, regional and country offices, limitations on travel due to visa issues, a shift in the prio­rity given for research and development of research capacity in low and middle income countries and a moratorium on funding international research in the US a period. Additionally, following the Covid pandemic, there has been a sharp increase in anti-science attitudes globally affecting funding and research initiatives and has impacted policy development and implementation as the lack of data to inform needs and priorities can be a major barrier to obtaining future support.

Despite the complex nature of international collaboration using a common protocol and comparable indicators, the number of studies undertaken and the data collected and analysed were a critical step forward to the global understanding of the relationship between alcohol consumption, culture, alcohol policy and injuries. Findings from these studies across multiple countries and cultures highlight the importance of aggregate-level factors that need to be considered in developing effective intervention and prevention strategies for reducing alcohol-related injury. They also emphasise the high prevalence of drinking in ER caseloads and the large association of drinking with injury, pointing to the need for integrating screening and brief interventions for those at risk in the ER context as part of a public health approach to the prevention and management of alcohol problems.

The presented complexities, challenges and rewards of undertaking a large scale multi-national research study like this demonstrated the unique opportunity and potential it provided. Findings from ICAIS highlight the immense value of an undertaking of this kind not possible to achieve with individual studies.

Conflict of interest/Konflikt interesów

None declared./Nie występuje.

Financial support/Finansowanie

None declared./Nie zadeklarowano.

Ethics/Etyka

The work described in this article has been carried out in accordance with the Code of Ethics of the World Medical Association (Declaration of Helsinki) on medical research involving human subjects, Uniform Requirements for manuscripts submitted to biomedical journals and the ethical principles defined in the Farmington Consensus of 1997.

Treści przedstawione w pracy są zgodne z zasadami Deklaracji Helsińskiej odnoszącymi się do badań z udziałem ludzi, ujednoliconymi wymaganiami dla czasopism biomedycznych oraz z zasadami etycznymi określonymi w Porozumieniu z Farmington w 1997 roku.

References/Piśmiennictwo