Why Are Some Male Alcohol Misuse Disorder Patients High Utilisers of Emergency Health Services? An Asian Qualitative Study

Background: Certain alcohol misuse patients heavily utilise the Emergency Department (ED) and Emergency Medical Services (EMS) and may present with intoxication or long-term sequelae of alcohol misuse. Our study explored reasons for repeated ED/EMS utilisation and sought to understand perpetuating and protective factors for drinking. Methods: Face-to-face semi-structured qualitative interviews were conducted. Participants were recruited from an ED in Singapore. Interviews were audio-recorded, transcribed verbatim and underwent manual thematic analysis. Emergent themes were independently reviewed for agreement. Data from medical records, interview transcripts, and field notes were triangulated for analysis. Results: All participants were male (n = 20) with an average age of 55.6 years (SD = 8.86). Most were unemployed (75%), did not have tertiary education (75%), were divorced (55%), and had pre-existing psychiatric conditions (60%) and chronic cardiovascular conditions (75%). Reasons for utilisation included a perceived need due to symptoms, although sometimes it was bystanders who called the ambulance. ED/EMS was preferred due to the perceived higher quality and speed of care. Persistent drinking was attributed to social and environmental factors, and as a coping mechanism for stressors. Rehabilitation programs and meaningful activities reduced drinking tendencies. Conclusion: ED/EMS provide sought-after services for alcohol misuse patients, resulting in high utilisation. Social and medical intervention could improve drinking behaviours and decrease overall ED/EMS utilisation.


Background
Patients who frequently attend the emergency department (ED) for alcohol-related problems are known as alcohol-related frequent attenders [1] (ARFAs). While these patients account for a small percentage of the patient population treated by the ED, they utilise a disproportionately large amount of emergency medical resources, which encompass emergency medical services (EMS), ED, and inpatient services [2,3]. Compared to the average ED patient, ARFAs have lower admission and higher abscondment rates from the ED, suggesting that their visits could be inappropriate and less urgent in nature [4]. In relation to stigmatisation, emergency providers are known to question the severity or validity of their medical complaints [5,6]. Investigative studies [3,7] have shined a light on the high mortality and morbidity of this patient population.

Study Design and Setting
This is a prospective, single-centre qualitative study of alcohol-misuse patients who have frequent attendance in the ED. Participants were recruited via convenience sampling from the ED of Khoo Teck Puat Hospital (KTPH). KTPH is a 795-bed general and acute care hospital serving a population of over 550,000 in Singapore's northern region. Since its operation in 2010, KTPH's ED has seen over 138,000 patients yearly, attending to both walk-ins as well as the more critically ill and trauma cases.
A team of trained male and female researchers conducted face-to-face interviews with study participants from February to July 2021 (6 months) (including the first, fifth, and sixth authors of this publication). Based on existing qualitative, grounded theory studies, an estimated but also practical sample size of 15 to 30 was determined. Study enrolment was pursued until theoretical saturation was achieved, with no new themes emerging from the interviews.

Inclusion and Exclusion Criteria
Eligible participants (male and female) were over 21 years old and spoke either English or Mandarin. They had at least five ED visits in the preceding 12 months of the study enrolment date, of which at least two visits were alcohol-related as defined by SNOMED-CT codes (Supplementary File S1). Patients who were psychologically or physically unable to demonstrate a capacity for informed consent or refused participation were excluded from this study.

Recruitment Process
Participants were first identified by attending medical staff within the ED, before they were referred to the study team. Once the study team assessed a patient to be eligible, the study coordinator then approached them to explain the purpose of the study and that their participation entailed interviews on their chronic alcoholism and the subsequent reasons, including their experiences, for their frequent ED visits. Written informed consent was obtained only when the patient was finally deemed to be medically fit. This consent process was also maintained for patients already admitted to the inpatient wards from the ED, in which case the admitted patient was approached and recruited in the ward. Only when informed consent was obtained were the interviewers introduced to the study participants by the study coordinator. Interviewers introduced themselves as researchers who were interested in learning about their drinking behaviours and ED/EMS usage experiences.

Data Collection
Each participant was interviewed by two primary interviewers with experience in qualitative research methods. A similar research area of interest that the interviewers have previously looked at encompassed understanding citizens' involvement in responding to public medical cases before first responders arrive.
Interviews were conducted in a private room within the hospital grounds either on the same or the next day of the participant's ED visit or on a different day convenient for both the participant and interviewers. No one else was present in the interview except for the two interviewers and the participant; all interviews were also completed on the same day within that single session. The interviews typically occurred in the morning when participants were less likely to drink and/or experience alcohol withdrawal symptoms prior to the interview. A semi-structured questionnaire was employed, covering participants' socio-demographic background, personal alcoholism profile, social alcoholism profile, perceptions of utilising emergency services, and motivation behind corrective behaviours, as well as attempts and attitudes towards interventions (Supplementary File S2). Interviewers were conducted until saturation was achieved, with no new themes emerging from the interviews. All interviews were audio-recorded with the verbal consent of participants.
Patient identifiers were removed, and interviews were transcribed verbatim on Microsoft Word by one of the two interviewers from each interview session. Each participant was assigned a unique subject code prior to the interview to be used throughout the audio recording. To capture essential nuances in communication, interviewers also recorded independent field notes beyond the questionnaire, including non-verbal behaviours of participants not captured by the audio recording. All audio recordings and transcripts were then stored on a password-protected laptop at the study site, accessible only by the research team.
Participants' medical records were retrieved by their treating doctor in the ED during their medical treatment. The treating doctor would provide information on each participant's number of ED attendances and number of alcohol-related ED attendances in the preceding 12 months, as well as any history of chronic and psychiatric conditions.

Data Analysis
Data from multiple sources including interview transcripts, independent field notes, and participants' medical records were triangulated for a more holistic understanding of study participants. Each transcript was additionally peer-reviewed by both primary interviewers retrospectively to ensure consensus of information between the interviewers.
Thematic analysis was then conducted across all transcripts, and relevant data in relation to the study objectives were identified and extracted to be indexed into one or more codes. Analyses were conformed to inductive and deductive codes relating to participants' (i) socio-demographic background, (ii) motivations for drinking and (iii) against drinking, (iv) rationale for the high utilisation of emergency services, and (v) perceptions and/or experiences of this utilisation. Emergent themes were subsequently revisited by a third independent reviewer, and differences and similarities between participants were explored. For the purpose of reporting the key findings and supporting quotations in the present report, the assigned subject codes unique to each participant were used in replacement of their actual names.
Study participation did not affect medical care rendered to the patients, although several participants had an extended ED stay to achieve a greater degree of sobriety. Enrolled participants were reimbursed with SGD$100 (approximately USD$72) supermarket vouchers-which could not be used for alcohol-at the end of the interview.
Ethics approval for the study was secured from the National Healthcare Group Domain Specific Review Board (DSRB 2020/00118). Informed consent was obtained from each participant, including permission to audio-record the interview and publish the data after removing identifiers to ensure anonymity. Our study adhered to COREQ guidelines [17].

Results
Forty patients were referred for the study and a total of 20 participants were enrolled when data saturation was observed ( Figure 1). Interviews lasted an average of 60 min (ranging from 19 to 122 min) in a single session. It was estimated to be at the fifteenth participant where the saturation point was reached.
Thematic analysis was then conducted across all transcripts, and relevant data in relation to the study objectives were identified and extracted to be indexed into one or more codes. Analyses were conformed to inductive and deductive codes relating to participants' (i) socio-demographic background, (ii) motivations for drinking and (iii) against drinking, (iv) rationale for the high utilisation of emergency services, and (v) perceptions and/or experiences of this utilisation. Emergent themes were subsequently revisited by a third independent reviewer, and differences and similarities between participants were explored. For the purpose of reporting the key findings and supporting quotations in the present report, the assigned subject codes unique to each participant were used in replacement of their actual names.
Study participation did not affect medical care rendered to the patients, although several participants had an extended ED stay to achieve a greater degree of sobriety. Enrolled participants were reimbursed with SGD$100 (approximately USD$72) supermarket vouchers-which could not be used for alcohol-at the end of the interview.
Ethics approval for the study was secured from the National Healthcare Group Domain Specific Review Board (DSRB 2020/00118). Informed consent was obtained from each participant, including permission to audio-record the interview and publish the data after removing identifiers to ensure anonymity. Our study adhered to COREQ guidelines [17].

Results
Forty patients were referred for the study and a total of 20 participants were enrolled when data saturation was observed ( Figure 1). Interviews lasted an average of 60 min (ranging from 19 to 122 min) in a single session. It was estimated to be at the fifteenth participant where the saturation point was reached.

Study Participant Characteristics
Demographic characteristics are shown in Table 1. While both male and female participants were eligible, all 20 participants recruited were male. They were mostly Indian, with the remaining coming from different ethnicities prevalent in Singapore. The majority of them were either divorced or single. Many participants had psychiatric conditions and various cardiovascular risk factors, including either, or a combination of, diabetes, hypertension, hyperlipidaemia, and ischemic heart disease.
Most reported that their first drink was between the ages of 10 and 20 years old, with a general preference for beer among other available alcoholic drinks. Aside from alcohol dependency, a majority of them were also tobacco users with an early onset age. Table 2 further detailed the participants' socio-economic characteristics. Educational background differed greatly among the participants, although most had completed only primary school education. Some of them attained post-secondary education and even various pre-tertiary qualifications. Most participants lived in rental apartments or self-owned public housing. Those living in a 1-to 2-room apartment accounted for the largest proportion of interviewed participants, and only one participant was reported to be homeless. Some lived alone, while others were revealed to be living with their friends or family.
The majority of the study participants were unemployed at the time of the interview. More than half of them were previously incarcerated for various criminal offences. A significant minority had at least one previous suicide attempt.

Treatment for Alcohol-Related Health Conditions
The majority of the interviewed participants reported experiencing alcohol withdrawal symptoms or various health conditions aggravated by their chronic drinking problem. They deemed these conditions severe enough to warrant immediate medical intervention, such that they had no choice but to call for an ambulance.

Alcohol Intoxication That Led to Intervention by Concerned Bystanders
Some study participants shared that concerned members of the public, upon seeing them unconscious, were the ones who called for emergency services. Bystanders had judged the need for ambulance services since they were unsure why the participants were unconscious. Study participants themselves admitted that their condition at that present instance did not require any form of emergency intervention.

Preference for High-Quality Service and Care Levels Afforded by Emergency Services and Departments
In general, most study participants felt a sense of gratefulness for the emergency services rendered, and the warmth and genuine care shown by emergency personnel. They appreciated the emotional involvement of these staff who not only recognised them but were also aware of their chronic conditions from multiple occasions of attending to them. The efficiency of the emergency services, which could render immediate medical help, further encouraged participants to continue utilising them instead of turning to other non-emergency options.
The ambulance that comes . . . sometimes it's the same people. So they know me, know what's wrong with me . . . before I explain to them. They spend some time with me, then they see (that) I'm okay. It's not that we are well-known, they just remember-remember-(me). (Participant 018) If I come by ambulance, excellent service. Very fast, everything (gets) done. If I come personally by taxi, or (with) my sister or brother, then I [sic] got to wait and wait and wait. But ambulance is a very fast service. (Once) they come (and) they bring you out of the ambulance, they put you on the bed already. (Participant 015) Interestingly, several study participants also expressed guilt for their constant utilisation of emergency services and repeated ED attendances. They felt disheartened at their use of EMS and going to the ED, because they perceived their medical condition(s) to not require the level of service care provided. Some felt liable for causing delays in treatment for other patients due to their unplanned ED attendances and subsequent hospital stay. Generally, they recognised that emergency services and resources were needed by those who were in critical condition or had acute conditions that required immediate medical attention.
Actually, I don't want to call. Because when you call, your number is already there (in the system). Every time (when) the same person (keeps) calling, (the paramedics) don't even come (anymore], because it is pulling a fast one . . . (When I go to the A&E because of drinking) I feel a bit guilty because I think these emergency services shouldn't be used on me. (Participant 003) Maybe I feel that it's no good [sic], I feel sorry every time because the more people like me come here too often, it will delay and reduce the chances for other people (to use the emergency services or department). (Participant 011) Here, accident [sic] people come down, heart problem, kidney problem, sometimes throat problem, a lot of cases. Sometimes some doctors also see me only [sic] (and they went,) "You again!" (Did) they think (that) I (really) want to come to hospital? (Participant 010) 3.3. Motivations for Drinking 3.3.1. To Cope with Personal Issues Study participants commonly cited alcohol consumption as a coping mechanism for various stressors. Some participants resorted to drinking as a way to release the frustrations that they experienced. Alternatively, some drank as a form of temporary pain relief. Others turned to alcohol to fill the void from their loneliness or drank simply to pass time and cope with boredom. For instance, Participant 010, who was currently divorced, mentioned that loneliness and boredom contributed to his perceived need for alcohol consumption.

To Cope with Symptoms Suggestive of Alcohol Dependence
While some of the study participants persistently returned to drinking to manage alcohol-related withdrawal symptoms, others attributed their chronic alcohol consumption to wanting to experience the intoxicating effects of drinking. These symptoms or consequences of use could be either physical or psychological in nature.
The heavy drinking lifestyle, especially during periods of boredom, might be a reflection of alcohol dependency in study participants. In general, a sense of helplessness was observed in those who had perceived a need to consume alcohol for the abovementioned reasons.
I tried to give up (drinking) on my own, but because of the withdrawal symptoms, then I just . . . I chose to go back to drinking rather than to undergo the withdrawal symptoms. (It feels) terrible, I tell you. Vomiting, shaking, your hands will shiver, your legs will tremble, that kind of thing Last time (it) was (to) drink to socialise, to just hang out with the uncles and pass time. The difference is (that) today, I just want to drink and I want to get drunk. I need to get drunk. (Participant 005)

Encouraging Social Environments
Ease of Obtaining Free Drinks. The vast majority of study participants were unemployed and did not have a financial source to sustain their drinking lifestyle. Despite so, participants mentioned the ease of obtaining free drinks from people around them. Like others, Participant 005 shared that acquaintances were willing to cover the costs of alcohol for them at nearby coffee shops, attributing this ease to his likeable personality. Encouragement from Others to Drink. The existence of a socially reinforcing environment appeared to be a strong enabler in promoting and maintaining one's drinking behaviours, independent of their financial situation. Interestingly, Participant 003 reflected that his spouse had encouraged him to drink at a family event despite efforts at alcohol abstinence for the past few months.
That time (when) I was trying to go off alcohol for a few months, at a Christmas party my wife saw [sic] (that) I (was) very poor thing [sic], so she asked me to have a can of beer. But I was the first one to pass out. (Participant 003) While some study participants, such as Participant 003, received explicit encouragement to consume alcohol, others experienced more subtle nudges or reinforcements that facilitated the continuation of drinking behaviours. Several participants revealed that people around them would not raise the issue of their alcohol use as long as they fulfilled their job requirements and/or remained non-disruptive to others and were not judgmental about their alcohol use.
. . . (I was) late half an hour (for work), but they (are) ok. The office (is) very good to me. I never create trouble, drink [sic] also never argue with them, so they (are) all very good to me. (Participant 006) Father knows I drink. He knows (that when it was) working time I drink. (Although) they know but they never disturb me. (Not even) one day also (did) he say "you don't come [sic] to this church." He never say. (Participant 017) Benefits from Drinking Alone. The decision to drink in solitude was prominent in many study participants. On one hand, it helped them avoid people who might express their disapproval to participants for their alcohol use. On the other hand, it also prevented them from getting into confrontations with others if they become intoxicated. Both benefits could have reinforced the drinking of alcohol.

Alcohol Consumption Habits Were Simply Not Considered to Be Problematic
Several participants did not perceive a problem in their alcohol consumption patterns. In other words, either their drinking behaviours were not seen as an indication of addiction, or they did not view their alcohol consumption to be as bad or serious as other forms of addictions such as drugs. In this case, alcohol addiction in comparison to drug dependency was deemed to be the "lesser of two evils". Participation in alcohol rehabilitation programs at various addiction treatment centres in Singapore was perceived to have some degree of effectiveness in reducing problematic drinking in individuals. Among those who received a previous formal intervention, participants cited the conducive and supportive environment during treatment as helping to suppress the urge to use alcohol, alongside an amplified resolve to improve their lifestyle. However, they returned to drinking upon discharge from these programmes, highlighting a limited effect in promoting alcohol abstinence. While many study participants acknowledged the usefulness of these treatment programmes, many did not express the motivation to take part as a means of managing their alcohol use. They knew that they had to be genuinely interested to help themselves and be committed to the treatment process. However, most were unwilling to reach out to and depend on external formal organisations for help.

NAMS [National Addictions
I prefer to do it [quit drinking] alone . . . I think the programmes [formal interventions] are good, but personally I feel more comfortable helping myself rather than engaging external organisations. (Participant 004) I personally feel that the interventions are helpful but I'm not improving because I cannot accept (that I have to stop drinking) and I cannot stop drinking. I think we addicts, we don't care for ourselves and so we don't take professional advice seriously. (Participant 003)

Deterring Social Environments
One's social environment could be both an enabler of and deterrence from alcohol consumption. Consistent with their inclination to drink alone, the discouragement or objection from important people in their lives could lead study participants to refrain from drinking, albeit temporarily, in their presence. These people may include family and even non-family members. Employment and day-to-day activities offer participants the opportunity to spend time constructively and meaningfully. Consequently, lower alcohol consumption was reported because of the need to work or attend church. Instead of idling away the hours in boredom or being confined at home, participants appreciated the opportunity to be meaningfully engaged, and this in turn reduced the perceived need to turn to drinking to occupy time. Some also indicated adherence to a drinking schedule so as not to jeopardise their employment, which could additionally serve as a method to assist with the management of their problematic drinking habits.

Discussion
To the best of our knowledge, this is the first Asian qualitative study on this patient population. Based on the grounded theory approach, our study uncovered key themes for the high utilization of emergency health services, as well as perpetuating and protective factors concerning ARFAs' alcohol misuse. We found that our study participants shared certain key demographic characteristics and had similar reasons for ED/EMS use.

Demographic Characteristics
All our study subjects were male with a preponderance of Indians. In a recently published database study [18], using comprehensive nationwide Singapore data, the proportion of male patients vastly outnumbered female patients presenting with alcohol-related diagnoses. In the same study, Indians contributed a much higher proportion of patients compared to the other major ethnic groups in Singapore. This could explain why our study sample was mainly Indian males. This highlights the potential existence of certain unique and possibly cultural factors associated with their drinking behaviours. This is also consistent with another Singapore study [19] and warrants further research.
Most of our study subjects had lower education levels and were unemployed. Local data pertaining to unemployment in this population are lacking. Previous studies [20,21] have shown that problem drinking is associated with losing a job and remaining unemployed. This finding suggests that they had lower socio-economic status. This is consistent with other similar populations [5,22]. Low income was not a deterrent to a drinking lifestyle for the study population. This is ascribed to their socially reinforcing environments where obtaining drinks from others, including from their relatives, acquaintances, and friends alike, was effortless. While seemingly harmless, these small acts of 'generosity' created more harm, further entrapping them in a vicious cycle of alcohol misuse. These individuals surrounding the participant are known as "enablers" and may include family and friends. They may not have a conscious intent to encourage alcohol misuse, but their actions may inadvertently encourage our study participants to continue drinking behaviours. Efforts should be directed to involve "enablers" in the alcohol reduction strategies for the study participant.

Reasons for ED/EMS Usage
The reasons for using ED/EMS services fell into 'push' and 'pull' factors. Push factors included participants seeking treatment for alcohol withdrawal and treatment of health conditions not directly related to alcohol use such as chronic pain. In addition, the perceived need for urgent medical attention in alcohol misuse patients for conditions other than acute intoxication is well documented [5,23].
The main pull factor was the perceived high level of service provided by emergency personnel. Most participants in our study group had basic accommodation and seemed less likely to turn to the ED for the purposes of lodging or solely for food. They understood that they would receive high-quality, expeditious care when utilising these services and preferred this approach compared to routine specialist clinic-based alcohol services. Another qualitative study [11] had shown similar findings where ARFA patients had extremely positive accounts of their prior ED attendances.

The Challenge in Managing Behaviour
Perpetuating factors for continued alcohol usage were varied. First, having been exposed to heavy and constant alcohol consumption patterns since youth, our subjects had developed the habit of turning to alcohol as a coping mechanism for varying stressors. Some stressors included physical pain from unresolved physical pain. Most participants recognised that these effects only provided temporary relief. Second, they drank to cope with physical withdrawal symptoms. Third, enabling social environments perpetuated alcohol usage. Lastly, several study participants did not consider their drinking to be problematic.
In keeping with the lack of insight and motivation to seek appropriate help, a UK study [11] suggested that ARFA patients lack the motivation to attend specialist outpatient alcohol-specific treatment. Therefore, obtaining entry into formal interventions such as a detoxification program at an addictions centre is difficult, despite some ARFAs recognising their benefits. However, even with the completion of such programs, recidivism was high, possibly contributed to by enabling social environments.
A paradigm shift from traditional treatment approaches and expectations is indicated. Research evidence points to the benefits of interventions involving case management and conducted primarily in home and community settings. In a pilot randomised controlled [24], Drummond et al. demonstrated that Assertive Community Treatment (ACT), a model of care that involves assertively seeing patients in the community and case-working to assist them in addressing and seeking help for problems in medical, psycho-social, and alcohol-use domains, led to less unplanned healthcare use. In a similar intervention model on the ARFA population, Hughes et al. [25] also showed reductions in both ED attendance and hospitalisations.
An interesting finding from our study was the tendency of concerned bystanders to activate emergency services upon encountering intoxicated and poorly responsive individuals. This may be explained by the fact that Singapore is a densely populated city-state, with public areas such as void decks (communal areas on the ground floor of public housing) and food courts experiencing high footfall. Such study participants with public drinking tendencies may also be vulnerable to exploitation and harm during intoxicated states. Working along a harm reduction approach through psycho-education on safe drinking behaviours and environments may be of benefit.
Overall, the observation that there could be more perpetuating than protective factors for alcohol use indicates a challenging endeavour to treat problematic drinking in our study participants. Alleviating the utilisation of EMS and EDs could be a long-term process. Exploration for interventions beyond traditional treatment approaches and expectations for such patients is indicated.

Limitations
The study was a single-centre qualitative study with a small sample size. As such, findings were exploratory, hypothesis-generating, and may not be generalisable to other ARFA populations. Secondly, findings may be subjected to social desirability, given the sensitive nature of the interview questions asked. Thirdly, the nature of convenience sampling led to selection bias in the enrolment process, since only patients who were available and willing to participate during the recruitment period were recruited. In order to minimise this bias, attempts were made to enrol patients who were unable to consent to be interviewed in an earlier ED visit. Finally, the study depended on self-reported data, which could not be verified for accuracy.

Conclusions
ARFAs may have similar demographics and perpetuating and protective factors for alcohol misuse. Emergency health services could provide sought-after services to ARFAs, resulting in high utilisation. Multipronged social and medical intervention could improve drinking behaviours and decrease overall ED/EMS utilisation. Findings from this study should be validated through a robust, prospective cohort study. Funding: This study is supported by the Alexandra Health Enabling Grant (AHEG2002) and the APC was funded by the same grant.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki, and approved by the NHG Domain Specific Review Board (NHG DSRB 2020/00118) (approval date 23 September 2020).

Data Availability Statement:
The data presented in this study are available on request from the corresponding author. The data are not publicly available due to privacy restrictions.