Healthcare professionals perspectives surrounding the use of medicinal ethanol for the management of acute alcohol withdrawal in hospital settings.

Alcohol Use Disorder (AUD) may lead to alcohol-related presentations to acute healthcare services, with conditions such as alcohol withdrawal syndrome (AWS) and alcohol related seizures, or presentations associated with chronic alcohol-related ill health, for example decompensation of alcohol related liver disease, or acute pancreatitis. As a result, patients with AUD often form part of high intensity healthcare service user populations. At a population level, nearly 6% of all hospital admissions in England were due to an alcohol-related diagnosis or a had an alcohol-related condition directly contributing to their admission [ 1 , 2 ]. Furthermore, alcohol-specific deaths rose by 63.8% between 2006 and 2023 [ 1 ], and alcohol related hospital admissions in 2023–2024 cost in excess of £351 million/year [ 3 ] in England. An estimated 5% of adults in England [ 4–6 ]) consume alcohol at harmful levels (seven UK units (8 g of absolute ethanol = 1 UK unit) or more on drinking days, or > 35units per week [ 7 ]). The Institute of Alcohol Studies predict, that left unaddressed these changes in consumption may lead to an extra 14,7982 additional cases of alcohol-related diseases, 9,914 additional premature deaths and a cost to the UK National Health Service (NHS) of an additional £1.2 billion in England by 2035 [ 8 ]. AWS reflects the rapid onset of a range of symptoms, behaviours and psychological features following sudden and unplanned cessation of the consumption of alcohol in an individual who is physically dependent on alcohol [ 9 ]. Symptoms range from agitation, tremor, sweating, fast heart rates to disorientation, hallucinations and seizures. Of those presenting to hospital with alcohol dependence, around 30% may develop AWS [ 10–12 ]. Furthermore, 1.7% and 1.66% of patients on general medical and surgical wards in the UK are assigned an ICD code for AWS or AWS with delirium, respectively [ 13 ]. Among hospitalised patients (<60 years), AWS is as common, or more common, than complications of other long-term conditions, including diabetes and hypertension [ 14 ]. The majority of AWS cases can be successfully managed using benzodiazepines (BZPs) [ 15–17 ]. Some (up to 23% [ 18–20 ]), particularly those who are severely physically dependent on alcohol, require escalating doses of BZPs or the use of adjunct medications such as phenobarbital, baclofen or clonidine [ 21 , 22 ]. As such, they may be considered refractory to standard approaches [ 23 ]. Large doses of BZPs can cause excessive and delayed sedation, and other pharmacological medications for AWS management are limited by a lack of high-quality evidence surrounding their efficacy [ 24–26 ] or require a level of monitoring only available in high dependency settings. As such, many patients presenting to hospital with AWS who experience severe AWS are admitted for extended periods [ 27 ], exposed to additional medication-associated side-effects, intubation or invasive blood-pressure monitoring, and require high dependency or intensive care [ 27 ]. Medicinal ethanol may represent an alternative treatment option in the management of patients who are severely physically dependent on alcohol, and not responding effectively to standard medicines whilst also avoiding some of the adverse side effects of other adjuncts [ 28 , 29 ]. We have recently reviewed the current evidence-base surrounding the use of medicinal ethanol for AWS. We found that whilst the evidence was largely of poor quality, and from a heterogenous group of study designs, it was as effective in the majority of studies as alternative treatments [ 30 ]. Furthermore, when we reviewed the prescribing practices relating to medicinal ethanol for AWS within an acute UK NHS healthcare trust we found that whilst the majority of AWS presentations are successfully treated using standard approaches (BZPs), a minority (∼5-10%) of patients are managed using oral ethanol (provided as UK unit measures of 37.5%ABV vodka, diluted in fruit juice). This approach is typically reserved for patients who 1) have a history of alcohol consumption consistent with severe physical dependence (>25 units/day for women, >30 units/day for men), or 2) have a history of AWS related seizures/delirium tremens, or 3) are known to the Alcohol Care Team (ACT) to experience difficult to manage AWS. Among this group, oral ethanol led to a reduction in the requirement for unplanned admission to the Trust when compared with BZPs [ 31 ], was not associated with an increased occurrence of subsequent alcohol related re-presentations or admissions at follow-up [ 32 ], led to effective symptom control measured with the CIWA-Ar tool, and was well received by service users [ 33 ]. Since these publications, Panneerselvam et al. (2025) reported effective implementation and positive patient outcomes of a monitored therapeutic alcohol administration programme in a perioperative setting [ 34 ]. Whilst these findings relating to the utility of medicinal ethanol in AWS management are promising, further evidence is needed prior to committing resource into a definitive randomised controlled trial and wider service-level roll out. One significant gap in knowledge relates to healthcare professionals perspectives around the use of medicinal ethanol. The perspectives, experiences, hesitancies and recommendations of professional stakeholders represent a vital part of assessing the acceptability and feasibility of a new, innovative intervention [ 35 ]. Here we describe the outcomes from a mixed-methods approach to understanding healthcare professionals opinions of future recommendations surrounding the use of oral ethanol for the management of AWS. We adopted two methods of investigation, a trust-wide survey and semi-structured in-depth qualitative interviews. This approach facilitated the determination of a broad sense of understanding, confidence and acceptability of oral ethanol among the workforce as a whole, whilst also allowing in-depth experiences and perspectives to be explored for future implementation.

75 questionnaire responses were received. Fifty-two percent of participants worked within medicine ( n  = 39), 24% in emergency medicine ( n  = 18), surgery ( n  = 7 (9.3%)), ITU ( n  = 6 (8.0%)), anaesthetics ( n  = 5 (6.7%)). Respondents were medical or surgical doctors ( n  = 43 (57.3%)); nurses ( n  = 26 (34.7%)); advanced clinical practitioners ( n  = 3 (4.0%)), a pharmacist ( n  = 1 (1.3%)) and a specialist doctor ( n  = 1 (1.3%)). Most reported a good understanding of why oral ethanol is a logical treatment option for the management of AWS (Yes n  = 42 (53.2%); No n  = 17 (21.5%); Maybe n  = 20 (25.3%)). The majority of healthcare practitioners considered oral ethanol to be safe for patients ( Table 1 ). Trust wide survey respondents. Resident doctors were generally less confident in prescribing oral ethanol when compared with consultant physicians (resident doctor very confident oral ethanol = 16%; consultant physician very confident oral ethanol = 28%). Despite 32% of consultants and 42% of resident doctors reporting either being confident or very confident to prescribe oral ethanol, 46% of consultants and 72% of resident doctors report either not feeling very confident or feeling not confident at all when prescribing oral ethanol. The frequency of common themes among responses to free-text questions was determined. These relate to the understanding of the rationale for ethanol in the management of AWS, safety and concerns surrounding its use ( Table 2 ). Trust wide survey content analysis. Generally, the rationale for using ethanol for AWS management was perceived positively. Sixty-one percent of responses outlined that ethanol’s safety profile is superior to alternatives or that it is effective for the management of AWS. A number of actionable statements relating to future practice and research recommendations were developed from this analysis. For example, increased education for both healthcare professionals and patients (and their carers) surrounding the use of ethanol for AWS, and reviewing the appropriateness of the CIWA-Ar tool to capture the effects of ethanol on AWS symptoms. Overarching topics among the survey free-text responses from the healthcare professionals relating to the use of ethanol for AWS management were derived ( Table 3 ). Trust wide survey free-text responses. Twenty-seven participants registered an interest in participating in research interviews in their responses to the trust-wide survey. Thirteen interviews were conducted with healthcare professionals from Sandwell and West Birmingham NHS Trust. Five were medically trained, two were pharmacists, six were in nursing roles; two also held senior leadership positions within the trust and a further two currently, or previously, held practice development and training lead roles ( Table 4 ). Semi-structured interview participant characteristics. Prescribing Administration Medical oversight Administration Training Storage, governance, audit Patient care Prescribing Medical oversight Administration Training Patient care Prescribing Complaint management Guideline review and approval Prescription checks Auditing, monitoring and reporting of controlled drugs management Pharmacy issuing Prescribing Management of patients Fielding queries from other specialties and ACT members Complaint management Guideline review and approval Administration Training Storage, governance, audit Patient care Administration Patient care Prescribing Administration Management of patients Fielding queries from other specialties and ACT members Education and training other healthcare professionals Consent and patient education Prescription checks Five overarching topics of discourse emerged; these related to 1) the utilisation of ethanol, 2) impact, 3) the future, 4) perspectives and 5) handling and management or oral ethanol for AWS ( Figure 1 ). Topic map arising from semi-structured interviews. Impact on patients: Various healthcare professionals with direct experience of the prescribing or administration of oral ethanol for AWS discussed the relative effectiveness of ethanol to address symptoms of AWS compared with standard treatment approaches. This was mainly reported in relation to the speed of resolution of symptoms and the group of patients experiencing severe AWS or that are considered very heavily dependent on alcohol. “I’d say for a majority of patients, I think it probably controls their symptoms a lot better than diazepam…It works very, very quickly as long as a patient consents to having alcohol…… And having the ethanol, it works very, very quickly… And it can be really safe and effective way to sort of…you know, control their symptoms and facilitate a safe discharge.” P45 “I’d say for a majority of patients, I think it probably controls their symptoms a lot better than diazepam…It works very, very quickly as long as a patient consents to having alcohol…… And having the ethanol, it works very, very quickly… And it can be really safe and effective way to sort of…you know, control their symptoms and facilitate a safe discharge.” P45 The impact of oral ethanol compared with benzodiazepines of sedation and cognitive engagement was discussed. Four healthcare professionals made reference to patients being more “‘with it’ than if they were treated with benzodiazepines” (P26) and less drowsy: “I think in terms of the patient who’s given, you know, vodka. Generally, I don’t see that drowsiness as much really. They seem a lot more, you know, they just seem a lot more, like I’d say, yeah, they’re not drowsy.” (P45) “I think in terms of the patient who’s given, you know, vodka. Generally, I don’t see that drowsiness as much really. They seem a lot more, you know, they just seem a lot more, like I’d say, yeah, they’re not drowsy.” (P45) Linked to this, and related to discharge from hospital, risk reduction relating to an absence of benzodiazepine induced functional and cognitive impairments that lead to accidental trauma were discussed: “if you’re giving some people loads of benzos, they’re functioning skills are going to be…depressed…they’re going to walk out and fall over, bang their head, walk in front of a bus. So it kind of doesn’t make sense” (P17). “if you’re giving some people loads of benzos, they’re functioning skills are going to be…depressed…they’re going to walk out and fall over, bang their head, walk in front of a bus. So it kind of doesn’t make sense” (P17). Furthermore an improved clarity for patients around discharge and the safety around resumption of alcohol consumption, in the absence of receiving a medically assisted withdrawal, was discussed: “once we’ve been using a benzodiazepine and then they’re going back home to alcohol I think there’s been a lot of concern sometimes from the patient about ‘do I need to drink immediately’? Like where am I at? Whereas [if] they [the hospital team] have only been using alcohol, they know. I think they know the feeling better in themselves…how much they need to drink when they need to drink, where they’re at with their alcohol intake, which helps” (P43) “once we’ve been using a benzodiazepine and then they’re going back home to alcohol I think there’s been a lot of concern sometimes from the patient about ‘do I need to drink immediately’? Like where am I at? Whereas [if] they [the hospital team] have only been using alcohol, they know. I think they know the feeling better in themselves…how much they need to drink when they need to drink, where they’re at with their alcohol intake, which helps” (P43) Impact on services: A reduction in the likelihood to develop a need to intensive care or high dependency involvement was reported. The latter is often associated with high medication and intervention burdens and deconditioning to patients: “where we didn’t have it [ethanol prescribing] and you know we had all sorts of problems with patients in ITU withdrawing and was struggling to sedate them with all sorts of other pharmacological agents…at the same time we had the period of patients on the ward being looked after by outreach who were alcoholics who start to withdraw, and we all knew full well that in two or three days time they’d be coming to ITU… On ITU, it wasn’t like just a few days, it was several weeks usually that they need ventilating for…But since we’ve gone back to having alcohol. Those problems when we can get it and we can get the dose right, have largely disappeared, I think. Yes, we still have problems because of the personalities of those kind of patients. But as regards the acute physiological effects of the withdrawal. Mostly they’ve got a lot less problematic” (P10). “where we didn’t have it [ethanol prescribing] and you know we had all sorts of problems with patients in ITU withdrawing and was struggling to sedate them with all sorts of other pharmacological agents…at the same time we had the period of patients on the ward being looked after by outreach who were alcoholics who start to withdraw, and we all knew full well that in two or three days time they’d be coming to ITU… On ITU, it wasn’t like just a few days, it was several weeks usually that they need ventilating for…But since we’ve gone back to having alcohol. Those problems when we can get it and we can get the dose right, have largely disappeared, I think. Yes, we still have problems because of the personalities of those kind of patients. But as regards the acute physiological effects of the withdrawal. Mostly they’ve got a lot less problematic” (P10). Furthermore, for individuals whose primary presenting complaint was alcohol withdrawal related, ethanol prescribing, in collaboration with early engagement with the ACT, helped reduce over medicalisation of patients, lengthy inpatient unplanned detoxification, and facilitate safe discharge: “I sometimes think the people that we see that are being prescribed ethanol it sort of opens up and focuses everyone’s brain…so they’re not then withdrawing in as an inpatient, which is inherent with risk and complication…ideally, we’re getting this patient home quickly, so almost it kind of influences you not to over medicalize them…that’s not to say that I would not do the medical side of things…but I think if I can think of a safe way of managing things as an outpatient, I would. And I think that this seems to work. it seems to be well tolerated.” (P16) “I sometimes think the people that we see that are being prescribed ethanol it sort of opens up and focuses everyone’s brain…so they’re not then withdrawing in as an inpatient, which is inherent with risk and complication…ideally, we’re getting this patient home quickly, so almost it kind of influences you not to over medicalize them…that’s not to say that I would not do the medical side of things…but I think if I can think of a safe way of managing things as an outpatient, I would. And I think that this seems to work. it seems to be well tolerated.” (P16) Finally ethanol prescribing was reported to be linked with a reduction in nursing workload burden, secondary to a decrease in the frequency of medication dispensing, which for controlled drugs can be burdensome: “The CIWA scoring can be very kind of, you know, can be time consuming and it’s quite a big ask when you’ve got a bay of patients you know…it can be quite a lot for nursing staff. I think with the vodka, they’re still on an hourly CIWA…but they’re not having to, you know, keep keep keep keep giving them diazepam every hour necessarily, because vodka is usually given every four to six hours. Sometimes we give it three hourly but generally it’s given on a regime every four hours or six hours.” (P45) “The CIWA scoring can be very kind of, you know, can be time consuming and it’s quite a big ask when you’ve got a bay of patients you know…it can be quite a lot for nursing staff. I think with the vodka, they’re still on an hourly CIWA…but they’re not having to, you know, keep keep keep keep giving them diazepam every hour necessarily, because vodka is usually given every four to six hours. Sometimes we give it three hourly but generally it’s given on a regime every four hours or six hours.” (P45) Adoption of ethanol prescribing: The role of the ACT at Sandwell and West Birmingham NHS Trust in relation to adoption and acceptance of ethanol prescribing was reiterated by various interviewees. This related to 1) their engagement in education of staff throughout the trust: “I think if it wasn’t for the fact that you’d got the [alcohol care] team so early leading it…and educating and some of the consultants guiding particularly the team in ED, we wouldn’t be where we are now. I think actually they’ve been paramount to that journey. I think without the understanding and the time taken to try and explain it, because even I was a bit like why would we give alcohol to an alcoholic? It seems a bit backwards, but actually that time…taken to sort of, teach and educate and guide the team of why we do it, that we’re not given a bottle of vodka. We’re giving a small amount that’s controlled, etcetera. I think that’s massive. And then that’s obviously gone into acute Med and then other areas in the organisation.” (P27) and 2) their facilitation of ethanol prescribing’s implementation and day-to-day delivery: [when asked: “What do you think has been central to the success of implementing and using alcohol in this group of patients at Sandwell?”] “The team around it, the right MDT team…to design this, manage the patients, review the patients and advise the ward staff. If we just sent out a protocol sent to all the wards, ‘this is what you do’ and left them to it I think would have been more of a disaster” (P18) Provision: Several topics emerged relating to the provision of ethanol as a medicine. These were knowledge-based or practice-based. Knowledge surrounding the rationale for ethanol use was largely good, and mimicked outcomes from the trust wide survey conducted. This largely revolved around the ethos that unplanned detoxification is largely not effective and that most patients requiring ethanol prescribing are not ready to stop drinking and so should not be forced into stopping. Alcohol was viewed as a useful tool to control their withdrawal symptoms in the short-term in a safe and effective manner, and thus help facilitate a fast and safe discharge. “Forcing them into a detox is really inhumane and so risky to their health that actually we’re not enabling them. We’re keeping them safe.” (P30) “I think if it wasn’t for the fact that you’d got the [alcohol care] team so early leading it…and educating and some of the consultants guiding particularly the team in ED, we wouldn’t be where we are now. I think actually they’ve been paramount to that journey. I think without the understanding and the time taken to try and explain it, because even I was a bit like why would we give alcohol to an alcoholic? It seems a bit backwards, but actually that time…taken to sort of, teach and educate and guide the team of why we do it, that we’re not given a bottle of vodka. We’re giving a small amount that’s controlled, etcetera. I think that’s massive. And then that’s obviously gone into acute Med and then other areas in the organisation.” (P27) [when asked: “What do you think has been central to the success of implementing and using alcohol in this group of patients at Sandwell?”] “The team around it, the right MDT team…to design this, manage the patients, review the patients and advise the ward staff. If we just sent out a protocol sent to all the wards, ‘this is what you do’ and left them to it I think would have been more of a disaster” (P18) Provision: Several topics emerged relating to the provision of ethanol as a medicine. These were knowledge-based or practice-based. Knowledge surrounding the rationale for ethanol use was largely good, and mimicked outcomes from the trust wide survey conducted. This largely revolved around the ethos that unplanned detoxification is largely not effective and that most patients requiring ethanol prescribing are not ready to stop drinking and so should not be forced into stopping. Alcohol was viewed as a useful tool to control their withdrawal symptoms in the short-term in a safe and effective manner, and thus help facilitate a fast and safe discharge. “Forcing them into a detox is really inhumane and so risky to their health that actually we’re not enabling them. We’re keeping them safe.” (P30) Governance: Governance concerns around ethanol prescribing were described as largely practical and operational. Some healthcare providers felt storage and handling were relatively unproblematic, noting that “there’s no reason… we couldn’t store it on an ITU,” (PID 10) and that once prescribed, it is simply managed within existing systems where “it’s prescribed like a CD [controlled drug]… two nurses check it out and it’s controlled.” (PID 20) Ethanol was reported to be easily stored within controlled drugs cupboards and electronic drugs management systems, and whilst like with all controlled substances, anxiety around it’s abuse was reported, from a governance perspective, no audit-based issues were reported to date ( “I’ve not had any alcohol go missing” (PID 14)). Public perception and senior leadership: Public perceptions were acknowledged as a potential limitation and barrier to the onward use of oral ethanol, however most clinicians viewed them secondary to the clinical value of ethanol prescribing represents. For example, negative press was a concern but members of the senior executive team interviewed, reported strong confidence in the practice itself with one stating: “I would be happy to defend it as the right medical plan.” (PID 18) “I would be happy to defend it as the right medical plan.” (PID 18) Relatedly, the senior executive staff interviewed report escalation of “… no patient safety incidents… and I haven’t had any patient complaints,” (PID 18). Occasional concerns from communications teams about potential misinterpretation—such as headlines claiming, “ we’re feeding alcohol to our Alcoholics ” (PID 27)—were acknowledged, but senior leaders felt these were mitigated by providing “ the right amount of information… the pros and cons… the benefits… the research .” (PID 27). Overall, senior leadership described a system with no escalated incidents, stable organisational confidence, and preparedness to manage reputational issues should they arise. Alcohol as medicine: Participants expressed mixed views about whether alcohol was understood and experienced as a medicine by staff. Cultural or religious beliefs were a factor discussed by some. One participant reported this potentially representing a barrier for the following groups: “…it [alcohol] is seen as a taboo topic, especially people of my background, your like Islamic background for example, if they’re Muslims, even within Indian communities…it’s societally taboo to be drinking.” (P13) “…it [alcohol] is seen as a taboo topic, especially people of my background, your like Islamic background for example, if they’re Muslims, even within Indian communities…it’s societally taboo to be drinking.” (P13) This related to: “…religious texts basically, especially relating to Islam. The translated version of the word ‘intoxicant’ was culturally associated to alcohol, but in reality it’s not stated as alcohol it’s ‘intoxicant’. So depending on how you culturally translate that. Most people would associate it as alcohol.” (P13) “…religious texts basically, especially relating to Islam. The translated version of the word ‘intoxicant’ was culturally associated to alcohol, but in reality it’s not stated as alcohol it’s ‘intoxicant’. So depending on how you culturally translate that. Most people would associate it as alcohol.” (P13) Despite this, and some reports from staff that they “don’t mind using alcohol hand gel,” but “if they are seen to be pouring a tot of vodka… they won’t want to do that,” (PID 17), this did not represent a frequent issue delivering oral ethanol as an intervention: “…when it comes to being religious, but in my current working environment, I haven’t seen that as being a barrier to providing or dealing with alcohol as a medication” (P13) “…when it comes to being religious, but in my current working environment, I haven’t seen that as being a barrier to providing or dealing with alcohol as a medication” (P13) Other staff members felt there were concerns among some of their peers around enabling or promoting a drinking behaviour. Some struggled with the principle itself: “why are we giving this person alcohol?” (PID 26) because it is “literally the thing they have been misusing,” (PID 43) unlike methadone where “we’d never give a heroin user actual heroin.” (PID 43) “why are we giving this person alcohol?” (PID 26) because it is “literally the thing they have been misusing,” (PID 43) unlike methadone where “we’d never give a heroin user actual heroin.” (PID 43) Yet overall, most participants concluded that within clinical contexts, “it’s a medicine, not a drink…It’s a medicine first” (PID 43) Across these examples, participants repeatedly emphasised that perceptions changed “with the education and the rationale behind it,” explaining that staff “could rationalise it as a medication” (PID 27) once they understood its purpose. Many reported that teams now “see it as a treatment,” (PID 30) noting that even those with reservations “at the beginning” came to view it as “seen as a treatment now.” (PID 30) In relation to patient perspectives, some felt patients “see it as a medicine… the medicine that they need to make them feel better,” (PID 30) particularly during withdrawal, whereas others believed patients would see it “as a drink,”; packaging and presentation were central to this distinction. Ethanol seeking behaviour: Across the dataset, participants consistently reported no evidence that patients attend hospital seeking alcohol. A minority of clinicians acknowledged the possibility of a “very minute population” who might use the hospital as a “contingency plan” if they had “run out of their means,” but emphasised these cases were “very few and far between.” (PID 26). Even when isolated requests occurred (“I’ve seen people… asking for alcohol; it doesn’t happen often” (PID 45)) participants agreed this was not the driver of attendance. Several described initial anxieties: “the general consensus… was absolutely not, this is going to be catastrophic, like there’s going to be a queue around the door” (PID 30) “the general consensus… was absolutely not, this is going to be catastrophic, like there’s going to be a queue around the door” (PID 30) However, the participants report, unequivocally, that these predictions generally “clearly haven’t been the case.” (PID 30). Clinicians repeatedly stated that: [they had] “never had someone come in… seeking alcohol,” (PID 17) “never seen a patient and thought you’re here to get a vodka,” and “I don’t think so [in relation to seeking behaviours] because it’s not a given and it doesn’t always happen. And a lot of stars have to align for the vodka to be the choice [of treatment]” (PID 16). “I’ve definitely not seen people coming to the hospital solely just to get vodka, you know. Yeah. Never had that.” (P45) [across] “four years I’ve never had someone come back that I’m aware of and said either you gave me vodka the last time or…Someone’s told me I can get vodka…which were the two main concerns at the time.” (PID30) [they had] “never had someone come in… seeking alcohol,” (PID 17) “never seen a patient and thought you’re here to get a vodka,” and “I don’t think so [in relation to seeking behaviours] because it’s not a given and it doesn’t always happen. And a lot of stars have to align for the vodka to be the choice [of treatment]” (PID 16). “I’ve definitely not seen people coming to the hospital solely just to get vodka, you know. Yeah. Never had that.” (P45) [across] “four years I’ve never had someone come back that I’m aware of and said either you gave me vodka the last time or…Someone’s told me I can get vodka…which were the two main concerns at the time.” (PID30) Instead, patients were typically requesting treatment and aid: “Oh, I don’t. I don’t think from my perspective [seeking behaviours are seen], I think they just want treatment” (PID26), for example something “to treat the shakes” (PID 16), as opposed to ethanol specifically. “Oh, I don’t. I don’t think from my perspective [seeking behaviours are seen], I think they just want treatment” (PID26), for example something “to treat the shakes” (PID 16), as opposed to ethanol specifically. Staff stressed that obtaining alcohol outside hospital is far easier: “there are much easier ways to get alcohol than coming to hospital…why would you sit in ED for eight hours for somebody to control how much alcohol you get?” (PID 17). “there are much easier ways to get alcohol than coming to hospital…why would you sit in ED for eight hours for somebody to control how much alcohol you get?” (PID 17). Other clinicians report interesting perspectives around the rationale and provision of alcohol treatment and ethanol seeking behaviours: “My view on that would be, well, these are a group in society who’ve got a problem, who are unwell, and if what we’re providing them for providing for them is the best treatments that they can get and this is better…then why shouldn’t they come here? You know, we have an endometriosis centre. We have a Bechet centre. Shouldn’t it be the same that patients with an alcohol problem? Go to a centre…that looks after them properly.” (P10) “My view on that would be, well, these are a group in society who’ve got a problem, who are unwell, and if what we’re providing them for providing for them is the best treatments that they can get and this is better…then why shouldn’t they come here? You know, we have an endometriosis centre. We have a Bechet centre. Shouldn’t it be the same that patients with an alcohol problem? Go to a centre…that looks after them properly.” (P10) And instead of alcohol seeking behaviours, that the trust has potentially observed a reduction in benzodiazepine seeking behaviours: “actually, interestingly, I’ve never thought about that…when we didn’t have alcohol…we probably had more patients presenting asking for diazepam than we’ve ever had, presenting asking for alcohol.” (PID30) “actually, interestingly, I’ve never thought about that…when we didn’t have alcohol…we probably had more patients presenting asking for diazepam than we’ve ever had, presenting asking for alcohol.” (PID30) Overall, clinicians strongly rejected the idea of alcohol seeking occurring as a result of the initiation of ethanol prescribing. Patient perspectives: Patient perspectives on receiving ethanol (and the type of ethanol they were in receipt of) were varied, shaped by personal history, cultural context and the circumstances of presentation. For some it was thought by staff that familiarity mattered: participants described patients who reacted to taste and brand in the same way people prefer their usual or specific cigarette brands: “if I drank super tenants every day and then you gave me vodka… it doesn’t hit the spot.” (PID 14). “if I drank super tenants every day and then you gave me vodka… it doesn’t hit the spot.” (PID 14). Others highlighted that the form of alcohol could carry psychological significance; with some patients “prefer… a sterile approach” (PID 26) being adopted to their treatment in order to avoid reminders of their usual drinking, while others: “may feel more comfortable drinking something that’s familiar…in an otherwise unfamiliar environment.” (PID 26) “may feel more comfortable drinking something that’s familiar…in an otherwise unfamiliar environment.” (PID 26) One member of staff reported a case of a single patient who responded negatively to receiving alcohol: [they were] “absolutely horrified,” saying, “I’m trying to give up alcohol. Why would you be giving me alcohol?” (PID 26) [they were] “absolutely horrified,” saying, “I’m trying to give up alcohol. Why would you be giving me alcohol?” (PID 26) However, most clinicians felt that acutely unwell patients were not concerned with type or taste, reporting that patients are: “scared and unwell…they know what’s coming and…want to avoid it,” and therefore “I can’t imagine any patient being overly bothered…in terms of alcohol percentage.” (PID 30) “scared and unwell…they know what’s coming and…want to avoid it,” and therefore “I can’t imagine any patient being overly bothered…in terms of alcohol percentage.” (PID 30) It was also felt that the presentation of ethanol being used and provided by healthcare professionals also shaped meaning for patient; using a syringe or measured medicinal volume “has a different connotation…a reminder this is a thing that’s being prescribed,” whereas a “glass of vodka” (PID 43) can blur the line between treatment and drinking. Overall, while individual preferences and discomfort appeared, participants generally described patients as “quite positive about it” (PID 27) and focused primarily on symptom relief. Wrap around services: Participants described the alcohol care team (ACT) as essential to safe and confident ethanol prescribing, repeatedly emphasising that difficulties arose when the alcohol care team aren’t around. In their absence, clinicians reported “a reluctance…” (PID 14) and a tendency to default standard treatments. Many highlighted the ACT’s accessibility and reliability, noting they were a team they would “never struggle to get a hold of,” (PID 14) and that clinical confidence increased because “you know you’ve got the backup and the support.” (PID 14) This wrap around support was described as central to the service’s success: “the right MDT team,” was regarded vital and the reason why a protocol alone “would have been more of a disaster,” with nurses and doctors otherwise “stopping it, starting it” [inconsistently] (PID 18). Across accounts, a “really robust alcohol care team” was seen as essential not only for ethanol prescribing but for alcohol withdrawal pathways more broadly, with several clinicians believing: “we’d have loads more instances of badly damaged alcohol withdrawal if we didn’t have them” (PID 16) “we’d have loads more instances of badly damaged alcohol withdrawal if we didn’t have them” (PID 16) Wrap around support also extended beyond acute care, with the ACT arranging community follow up for motivated patients —”we’ll call them… or they can come back to clinic” (PID 37)—ensuring continuity after discharge. Service improvement: Education and training were repeatedly described as essential, with staff highlighting the need for “proper teaching and support” (PID 43) before widespread adoption could occur. Participants described a range of possible formats, from “a brief training package, half hour session” to “some hand holding for the first few cases” (PID 18) and subsequent audit of outcomes. Adoption Recommendations: Participants described three core elements as essential for adopting ethanol prescribing in other trusts 1) specialist leadership, 2) clear governance, and 3) phased, education led implementation. A “really robust alcohol care team” (PID 16) was repeatedly identified as the foundation of safe practice and cultural acceptance, with many emphasising that introducing ethanol prescribing and confidence in its use “wouldn’t happen overnight” (PID 16) and that an alcohol team should “take all the responsibility for it first and then slowly disseminate that learning.” (PID 16). Alongside this, participants stressed the need for a clear SOP, internal governance approval, and robust supervision. A meet the team approach for other NHS Trusts or healthcare partners wishing to adopt oral ethanol prescribing to observe the model, and see the evidence base in action recommended as an implementation and translation aid. While some felt toxicologists were helpful, most agreed oversight “does not need to be a toxicologist,” (PID 18) but rather “someone…involved in acute management ” (PID 26) who is interested, engaged, and willing to take responsibility for patients with alcohol related ill-health: “it’s understanding what you’re doing and and probably giving a shit about your patients”. (PID 27) “it’s understanding what you’re doing and and probably giving a shit about your patients”. (PID 27) Change in formulation: Recommendations surrounding changing the current formulation and presentation of oral ethanol were made by various participants. However, these recommendations reflected mitigating organisational risk, and barriers with healthcare professionals’ adoption of the intervention and addressing stigma associated with alcohol use, rather than adaptations that may lead to improvements in patient care. For example, many felt patients “wouldn’t say that the taste matters that much,” (PID 10) especially when withdrawing. In contrast, several clinicians described strong emotional reactions among staff to: “opening the CD cupboard and seeing a bottle of Smirnoff,” (PID 18) noting that it “doesn’t look like a medicine… it looks like something someone’s going to sit in the bar and drink.” (PID 14). This “ juxtaposition ” (PID 43) made some staff uncomfortable as previously outlined in earlier themes. “opening the CD cupboard and seeing a bottle of Smirnoff,” (PID 18) noting that it “doesn’t look like a medicine… it looks like something someone’s going to sit in the bar and drink.” (PID 14). This “ juxtaposition ” (PID 43) made some staff uncomfortable as previously outlined in earlier themes. Largely it was felt that decanting into a pharmacy-labelled bottle may reduce the polarising nature of oral ethanol in the controlled drugs dispensary: [changing the presentation] “would definitely impact… nursing staff” and avoid “compromising your values.” (PID 26). [changing the presentation] “would definitely impact… nursing staff” and avoid “compromising your values.” (PID 26). Relatedly, presentation was discussed as an important factor influencing public interpretation; the presence of “a bottle of Smirnoff” in a cupboard was seen as something that could prompt “screenshots going on social media,” (PID 18). There was debate around replacing vodka with absolute ethanol. Some believed there may be an argument for absolute ethanol because it “would lead to slightly more acceptance,” and would avoid the impression that “you just turn the hospital into a bar.” However, most rejected the idea due to cost, complexity, and limited practicality, noting that absolute ethanol is “ more expensive ,” comes in unfamiliar vials, and is “ not typically a product used for oral consumption .” Future research: Participants were asked about their perspectives on future research to improve our understanding and confidence around the use of oral ethanol for AWS. When asked about the conduct of a randomised controlled trial, participants consistently highlighted a set of outcomes they believed should be measured spanning both patient and service level indicators: Length of stay, accident and emergency repeat attendances or admissions Avoidance of care escalation requirements “whether or not someone needed to be admitted to ICU,” Incidence of harm for patients, including avoidance of severe withdrawal and delirium tremens, “patient safety incidents,” treatment failure, and staff, for example measuring a reduction in “violence and aggression,” with some suggesting improved management might make “staff feel safer.” Measures of functional recovery “how close to your baseline you get back to,” Cost savings: relating to improved flow in emergency and acute areas, and avoidance of admission or escalation to high dependency areas Length of stay, accident and emergency repeat attendances or admissions Avoidance of care escalation requirements “whether or not someone needed to be admitted to ICU,” Incidence of harm for patients, including avoidance of severe withdrawal and delirium tremens, “patient safety incidents,” treatment failure, and staff, for example measuring a reduction in “violence and aggression,” with some suggesting improved management might make “staff feel safer.” Measures of functional recovery “how close to your baseline you get back to,” Cost savings: relating to improved flow in emergency and acute areas, and avoidance of admission or escalation to high dependency areas Overall, participants described a future trial needing to capture both acute clinical stability and broader system impacts, with outcomes offering clear evidence of safety, efficiency and patient centred benefit.

We have reviewed healthcare professionals’ perspectives of the role that medicinal ethanol may play in the management of AWS in acute healthcare settings. The outcomes from a large trust-wide survey and a smaller series of in depth semi-structured interviews were comparable in their findings; these included: Medicinal ethanol is largely considered as an appropriate medicine by staff at Sandwell and West Birmingham NHS Trust representing the best, and most logical treatment option for a select group of patients; Ethanol has reduced resource burden associated with the management of AWS without leading to alcohol seeking behaviours by service users; From a governance and public profile perspective, oral ethanol has not caused any issues in terms of complaints, abuse, or diversion thus far, and; An effective and fully integrated ACT played a vital role in the implementation and utilisation of medicinal ethanol; this included extensive healthcare professional and service user education surrounding the intervention. These educational efforts helped both overcome initial hesitancies relating to its use and improve confidence to use oral ethanol as a medication among prescribers Medicinal ethanol is largely considered as an appropriate medicine by staff at Sandwell and West Birmingham NHS Trust representing the best, and most logical treatment option for a select group of patients; Ethanol has reduced resource burden associated with the management of AWS without leading to alcohol seeking behaviours by service users; From a governance and public profile perspective, oral ethanol has not caused any issues in terms of complaints, abuse, or diversion thus far, and; An effective and fully integrated ACT played a vital role in the implementation and utilisation of medicinal ethanol; this included extensive healthcare professional and service user education surrounding the intervention. These educational efforts helped both overcome initial hesitancies relating to its use and improve confidence to use oral ethanol as a medication among prescribers In addition to the points outlined above, staff also report that medicinal ethanol reflects a treatment strategy that takes into account the service users’ wishes, drinking goals, and importantly, requires understanding and consent to the intervention; all of which are aligned with core ethical principles [ 40 ]. Ethanol for AWS, implemented within the pathway at Sandwell and West Birmingham NHS Trust (described in detail [ 31 ]), allows for a unique opportunity to effectively identify those at risk of severe AWS and prevent it’s escalation, rather than responding to severe AWS once it has developed (often in someone whose refractoriness to escalating doses of standard treatments may have been predicted and avoided given an effective and in-depth history). Relatedly, there was an understanding that large doses of benzodiazepines, over lengthy hospital admissions were not always in the best interests for patients’ long term and that medicinal ethanol, may help prevent this. A recent cross-sectional survey reviewed healthcare professional practices surrounding the management of severe AWS [ 41 ]. Haloperidol, clonidine, propofol and phenobarbital were some of the most commonly employed adjuncts. However, only 33% of physicians felt comfortable prescribing phenobarbital due to fears around respiratory depression and reduced consciousness. This resulted in 55% of those surveyed waiting until refractoriness to benzodiazepines had been demonstrated, 12% waiting for seizures to develop, and 8% waiting until intensive care admission was required before initiating phenobarbital [ 41 ]. In addition to healthcare professional confidence limiting early intervention with severe AWS, a recent review of the evidence surrounding the efficacy of adjuvant agents in the management of AWS failed to demonstrate a clear case in favour of dexmedetomidine, phenobarbital, dopamine antagonists or GHB [ 24–26 , 42 ]. This report also highlighted the need for intensive care or accident and emergency resuscitation level monitoring and expertise when using many of these adjuncts, further limiting their use [ 42 ]. As a result, small amounts of ethanol supplied to patients as part of a preventative, regulated and holistic assessment pathway may represent an effective alternative to improve outcomes among those presenting to acute services at risk of developing severe AWS. A unique feature related to the provision of medicinal ethanol was relayed by healthcare professionals, namely improved cognitive awareness or engagement, when compared with alternative medications. Ethanol has previously been shown, albeit in healthy controls, to have less impact on cognitive performance, concentration and mood, compared with some benzodiazepines such as alprazolam [ 43 , 44 ] and diazepam [ 45 ]. This feature is employed by the team at Sandwell and West Birmingham NHS Trust in some clinical contexts, for example post-stroke, where there is a desire to avoid sedative agents due to GCS monitoring, and in some patients, a need to concomitantly manage AWS. Furthermore, the ACT report using the period of stabilisation following administration of medicinal ethanol as a ‘teachable moment’ with patients. We have previously demonstrated that the number of units of alcohol required to effectively address the symptoms of AWS is far less than the number of units routinely consumed by patients outside of the hospital. In this circumstance, the ACT capitalise on the moment of cognitive engagement, to reflect how very small amounts of ethanol, when compared to how much the service user drinks outside the hospital, have been used to effectively address their symptoms; resources and information for safe reduction are then also provided. Whilst provisional and anecdotal in nature, this feature warrants further exploration, because it may facilitate an opportunity to provide both medical management of AWS symptoms, and a brief intervention to reduce alcohol consumption going forward.

We acknowledge the potential medical and ethical concerns associated with the use of oral ethanol as a medicine. Ethanol may be considered to have a variable pharmacokinetic profile [ 46 ], is contraindicated in patients receiving metronidazole or disulfiram, and may, like many other hyperosmolar solutions, cause phlebitis if it extravasates [ 47 ]. However, benzodiazepines also exhibit marked variability in pharmacokinetics and dynamics [ 48 ], may be associated prolonged impairments in cognitive function [ 49 ], and are sedating, which may interfere with clinical assessment in some case presentations. Furthermore, we demonstrate in the context of a comprehensive and holistic patient pathway, where consent, understanding and capacity is paramount, ethanol as a medicine may represent and acceptable and feasible intervention for patients and healthcare providers. Whilst promising, and demonstrating the need for further translational work, we acknowledge this study has limitations. Our interviews and trust-wide survey have been completed among service providers working in a single UK NHS organisation. Sandwell and West Birmingham NHS Trust provides acute medical care to over 500,000 patients via a large inner-city service and hosts specialist expertise in the management of the poisoned patient and alcohol dependence. Its service users are culturally diverse [ 50 ], stem from some of the UK’s most deprived neighbourhoods [ 51 ] and consume greater than average levels of alcohol (the number of dependent drinkers and alcohol related mortality exceeds national averages in Sandwell and West Birmingham NHS Trust [ 52 ]). As a result, we recognise that 1) the necessary infrastructure for implementation of medicinal ethanol might not be in place in smaller organisations, 2) severe AWS may represent less service-level demand in other areas, and therefore the ‘need’ to better manage this condition could be considered less pressing, 3) for some the notion of providing alcohol dependent patients medicinal ethanol as a lifesaving intervention may be considered taboo.

Several recommendations evolved from our studies relating to the use and onward implementation of medicinal ethanol for severe AWS; these have been divided into practical, organisational and research recommendations ( Table 5 ). Recommendations regarding the use of medicinal ethanol. Whilst further work is needed, we demonstrate that following transparent implementation practices, healthcare professional education, and clear governance pathways, medicinal ethanol is an acceptable intervention to treatment providers:

A trust wide perspectives survey developed for dissemination on Microsoft Forms. The survey contained a mixture of single and multiple-choice, free-text and Likert scale (5-point scales; for example where 1 = not confident at all and 5 = very confident) response questions relating to: participant role and department, previous experience of ethanol prescribing, the perceived patient safety, confidence to prescribe and future recommendations surrounding the use of oral ethanol for AWS. The survey was circulated via internal mailing lists three times between 24/05/2024 − 10/06/2024. Mailing list recipients included: resident doctors, consultants, junior and senior nursing staff, pharmacists, psychological services, occupational therapy and nutrition teams providing care across accident and emergency, medical, surgical, psychiatric and anaesthetics specialties. Semi-structured interviews were conducted remotely between March and July 2025 using Microsoft Teams. The record and transcribe function was used; the transcript was proofed alongside the audio recording of the interview to correct any errors in automated dictation. All interviews were conducted by Darren Quelch. Darren Quelch is a senior researcher fellow with experience in clinical practice and conducting qualitative research interviews with healthcare professionals surrounding intervention evaluation and development. An interview schedule was developed informed by outcomes of the trust-wide survey and evolved through discussion of the study team and the broader alcohol care team at Sandwell and West Birmingham NHS Trust. The interview schedule covered: staff background and experience of oral ethanol prescribing, complications, barriers and routes to overcome barriers associated with ethanol prescribing, perceptions of patient responses and outcomes, additional resource requirements both at Sandwell and West Birmingham NHS Trust and elsewhere should oral ethanol be implemented more broadly, and recommendations surrounding future research relating to oral ethanol for AWS management. We specifically asked all participants about 1) alcohol seeking behaviours, and 2) patient and healthcare professionals’ perceptions of alcohol as a ‘medicine’. A within-subject iterative approach was used during interviews, such that each participant’s responses guided subsequent questions within their own interview. The process was not iterative across participants; each interview was conducted independently of others’ responses. A qualitative descriptive approach allows for low-inference findings to be drawn directly from the data [ 36 ] while promoting agreement among researchers. This technique also supports inductive, “real-world” dissemination of the interview material, unshaped by prior constructs or theoretical frameworks [ 37 ] and without the stringency associated with more commonly used qualitative analytic traditions [ 38 ]. Unlike theory-driven approaches such as grounded theory, qualitative descriptive methods do not strictly require data saturation or exhaustive testing of theoretical constructs. Our interviews were informed directly by data from our survey exercise and an evolving logic model surrounding the use and wider implementation of oral ethanol for AWS. As such our analysis was context specific, placed emphasis on low-inference, inductive reporting of participants’ accounts, uninfluenced by external constructs or theories, and kept as close as possible to the raw data to present a perspectives and experiences summary. This approach also allows a pragmatic stance in which rich, detailed accounts can be provided despite a modest sample size. Analysis was performed by Josh Molina and Nyle Davies and reviewed by Darren Quelch. The following process was adopted for analysis; where possible recommendations relating to objectivity, dependability, credibility and transferability were adhered to [ 37 ]: An initial coding manual was created a priori based on evidence reviewed during the development of the interview schedule. This was updated following review of the interview transcripts by Darren Quelch and Josh Molina. The codes were organised into topics (a designated meaningful unit of content or group of units) and sub-topics (independent to, but semantically related to the parent topic). The coding manual was then applied to the interview transcripts using NVivo (version 14) [ 39 ]. Where new topics emerged the coding manual was revised throughout the analysis period; previously coded data was recoded using revised coding manual in an iterative process. Comparable statements expressing similar ideas were grouped into meaningful units of discourse and categories, creating sub-categories where patterns or contextual distinctions emerged. Construct descriptive narratives for each category and sub-category using marginal notes and analytic memos and select representative quotations to illustrate key findings were made. Quality assurance of coding and category development was performed by a second reviewer; discrepancies were resolved by consultation with a third party to reach consensus. Data displays and matrices were developed to organise and present outcomes that were appropriate to the research and practice audience and the aims and objectives of the study. An initial coding manual was created a priori based on evidence reviewed during the development of the interview schedule. This was updated following review of the interview transcripts by Darren Quelch and Josh Molina. The codes were organised into topics (a designated meaningful unit of content or group of units) and sub-topics (independent to, but semantically related to the parent topic). The coding manual was then applied to the interview transcripts using NVivo (version 14) [ 39 ]. Where new topics emerged the coding manual was revised throughout the analysis period; previously coded data was recoded using revised coding manual in an iterative process. Comparable statements expressing similar ideas were grouped into meaningful units of discourse and categories, creating sub-categories where patterns or contextual distinctions emerged. Construct descriptive narratives for each category and sub-category using marginal notes and analytic memos and select representative quotations to illustrate key findings were made. Quality assurance of coding and category development was performed by a second reviewer; discrepancies were resolved by consultation with a third party to reach consensus. Data displays and matrices were developed to organise and present outcomes that were appropriate to the research and practice audience and the aims and objectives of the study. The study was registered as a service evaluation with the Sandwell and West Birmingham NHS Trust research and development department; approval from the Health Research Authority and Health and Care Research Wales was provided (IRAS number = 348774). Ethical approval was provided by the Faculty of Lifesciences and Education low risk research ethics committee at the University of South Wales (approval number = P250901LR). This study was carried out in accordance with the ethical principles of the Declaration of Helsinki. No minors or vulnerable adults were included in this study. Informed consent was gained from study participants by both written (electronically signed) consent forms, and verbal confirmation prior to any study related interview recordings.