The impact of a continuum model of alcohol problems on clinical practice: a double-edged sword? Commentary on Morris et al. ‘Should we promote alcohol problems as a continuum? Implications for policy and practice’

Morris and colleagues provide an interesting commentary on the policy and practice implications of adopting a continuum model of alcohol use and harms (Morris et al., 2023). This model departs from the hitherto dominant categorical model, which applies a threshold to determine ‘use disorder’, ‘harmful use’ or ‘dependence’ according to the severity of alcohol problems experienced by an individual, according to the number of symptom criteria endorsed.

A major benefit of a continuum approach, as raised by the authors, is the potential this offers for destigmatisation. Alcohol use disorder (AUD) is highly stigmatised (Kilian et al., 2021), with the risk of potential social exclusion and marginalisation of people when viewed through the traditional biomedical lens of ‘alcoholism’ (Room, 2005). As such, a continuum approach could foster an individual’s problem recognition by blurring the distinction between heavy alcohol use and AUD. This serves to circumvent the clinical labelling often rejected by people wanting to change their drinking (Sobell et al., 1996) and normalises help-seeking. However, research by Morris et al. suggests continuum models alone do not lead to enhanced identification of problems (Morris et al., 2022). Nevertheless, destigmatisation is key to driving help-seeking (May et al., 2019) and reducing the prolonged duration of untreated illness for AUD globally. Indeed, in an Australian study, the time between problem onset to first treatment engagement for AUD was as high as 18 years (Chapman et al., 2015), and likely longer for women and marginalised or vulnerable groups facing heightened stigma.

Whilst it may be desirable to lower the threshold so more people seek treatment at an earlier (and less severe) stage in most jurisdictions, alcohol services are under-funded and under-resourced. Thus, in reality, any increased demand could exacerbate waitlists and bottlenecks, and ultimately impede or delay access to treatment. Indeed the US 2021 National Survey on Drug use and Health indicates that fewer than 9% of people aged 18+ years with AUD in the past 12-months had received AUD treatment (Substance Use and Mental Health Services Administration (SAMHSA), 2022). In Australia, it is estimated that the existing alcohol and drug treatment system is currently meeting the demand of only around 27–56% of those needing care (Ritter et al., 2019). The long-standing underinvestment in alcohol treatment is entirely disproportionate to its significant harms (Griswold et al., 2018). Thus, it is critical, we reserve scarce treatment resources for those with the greatest need, whilst balancing opportunities for early intervention or indicated prevention.

Globally, alcohol treatment services and systems have largely been orientated to meet the needs of individuals with moderate to severe alcohol problems (Tuithof et al., 2016). There remains very limited research on what works for lower severity problems. Indeed, given the recognition that many people who do not seek treatment recover ‘naturally’ from AUD without treatment (Kelly et al., 2019; Witkiewitz et al., 2019), it remains unclear whether support or formal intervention is even necessary for many individuals. For example, applying a continuum approach to existing symptom criteria, it is easy to envisage that many who drink regularly might endorse one or two items on the DSM-5- such as Criteria 1 ‘if alcohol is often taken in larger amounts or over a longer period than intended’, and Criteria 4 ‘craving, or a strong desire or urge to use alcohol’. Yet, it is unlikely such individuals require the level of psychological, medical and pharmacological interventions typically offered within current treatment. Despite growing research into early intervention and screening and brief intervention (SBIRT) approaches, translation of such research into practice has been slow, with limited uptake by primary care or community health sectors (Thoele et al., 2021). Key questions that will need to be answered include the identification of pathways to indicate where people with low severity problems go, and expansion of effective scalable and accessible digital health interventions, designed to meet the needs of individuals with lower severity problems or harms (Kaner et al., 2017). In addition, to circumvent the overburdening of existing treatment services, we need to develop clear evidence-based pathways that channels individuals into care matched to their needs, with varied duration and intensity, including appropriate self-help and psychoeducation, brief and low-intensity counselling approaches, peer-based recovery support, and digital health.

Importantly, the existing evidence base for interventions for AUD are drawn from studies in populations with high severity use disorders, with demonstrated effectiveness comparable to other health disorders (Witkiewitz et al., 2019). One way to compare effect sizes of interventions in healthcare is to examine the number needed to treat (NNT), which refers to the number of patients who need to receive the treatment to prevent one additional bad outcome (i.e. relapse to drinking, or binge drinking for instance). Meta-analyses suggest the NNT for firstline alcohol use pharmacotherapies naltrexone and acamprosate is in the range of 12–20 (depending on binge drinking or abstinence as an outcome) (Kranzler & Soyka, 2018). This is favourable in comparison to medications prescribed for other chronic health problems such as antihypertensives for stroke prevention (estimated NNT 67), or aspirin to prevent cardiovascular disease (estimated NNT 50) (Fairbanks et al., 2020). This challenges the widely-held assumption that medications for AUD are ineffective, with this myth driving under-prescribing and international estimates suggesting that less than 10% of people with AUD receive medication treatment (Foulds et al., 2018; Kranzler & Soyka, 2018). Despite the low uptake of medication treatments for AUD, unsurprisingly they are up to sixteen times more likely to be prescribed to individuals with dependence and higher severity problems, rather than individuals with lower severity problems (Han et al., 2021). Similarly, a meta-analysis of cognitive behavioural therapy for drug and alcohol use disorders found a moderate and significant effect size, with improvements up to six months across a range of outcomes (Magill et al., 2019); but again, these effect sizes were based on treatment seekers, and predominantly from specialist alcohol and other drug treatment providers, where individuals met criteria for use disorder or dependence diagnoses. Providers draw on this body of evidence to guide clinical decision-making, but the generalisability of this literature to lower severity populations remains to be clarified.

Further, clinical decision-making in alcohol treatment is typically informed by case formulation, taking into account both the classification systems and individual symptom criteria, as well as the biopsychosocial context and range of comorbidities that an individual presents with (Gastfriend & Mee-Lee, 2022). Given this multidimensional process, it is worth acknowledging that a threshold model for identifying use and harms does not detract from the value of individual symptom criteria to inform individually tailored and holistic treatment responses.

In summary, continuum-based models raise a number of key challenges for the conceptualisation of alcohol use problems, not least how such a model can be applied within treatment and clinical practice to challenge existing decision-making thresholds and paradigms. Perspectives of individuals with lived experience of alcohol use, alcohol use disorder treatment, and ‘natural’ recovery will be essential to guiding progress in this field, with the prospect of a less stigmatized and more accepting framework that engages people effectively across a spectrum of need.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.