Table 1 Research on alcohol screening tools among Indigenous Australians
References | Setting (remoteness) | Participants | Screening tool (number of items) | Study type | Some relevant observations |
|---|---|---|---|---|---|
Skowron and Smith [36] | Port Hedland area (Remote WA)a | Homeless Aboriginal people (n = 162) | CAGEb (4 items) SMASTc (12 items) | Quantitative: CAGE compared with clinical interview on drinking | CAGE had reasonable validity in comparison with clinical interview Those with higher CAGE score drank more on previous day, and more frequently SMAST was found hard to understand by participants in a pilot study |
Hunter et al. [24] | Kimberley region (Very remote, WA) | Aboriginal community members (n = 516) | CAGE (4 items) (reworded for local use of English) | Quantitative: compared with clinician interview | CAGE scores were associated with frequency of alcohol consumption Over half of ex-drinkers had a CAGE score of 2 or more |
Brady et al. [29] | ACCHS (Urban) | Health staff (n = 14), clients (20) | AUDITd (10 items); Two questions: average days per week the patient drank, and amount and type of drinks per day | Mixed methods | AUDIT was reported by health workers to be long and intrusive Problems with question comprehension, when using standard English phrasing Preferred two questions on frequency and quantity |
Kowalyszyn and Kelly [28] | Community setting (Remote far north Queensland) | Aboriginal community members (n = 99) | AUDIT (10 items) KATe (12 items) | Quantitative: two screening tools compared | High correlation between AUDIT and KAT responses AUDIT seemed easier for participants to complete and had more face validity Sharing of alcohol led to challenges in quantification of drinking |
Schlesinger et al. [37] | Clinical and non-clinical services (Urban, regional and remote Queensland) | Aboriginal and Torres Strait Islander clients of services (n = 175) | IRISf (13 items) | Quantitative: IRIS compared with AUDIT, two dependence scales (SDSg, LDQh), mental health scales, and interview on consumption by Aboriginal community worker | IRIS was validated as a screen for alcohol and drug, and mental health risk IRIS had good convergent validity with other scales for alcohol risk, substance use, dependence, and mental health, including AUDIT Good sensitivity for detecting men drinking 11+ drinks per occasion; some false negatives for women with 7+ drinks |
Clifford and Shakeshaft [20] | ACCHSs: (One urban and one rural, in NSW) | Health staff (n = 32) and patients (n = 24) | AUDIT (10 items) | Mixed methods: survey and group interviews of staff and (separately) of patients | Staff reported: Staff and patients preferred the first three items of AUDIT (AUDIT-C) over the full AUDIT; Patients screened with AUDIT-C reportedly showed more interest in their drinking risk than those previously screened with a single question on average consumption; and Staff and patients preferred screening to be part of a routine health check rather than opportunistic |
Clifford et al. [38] | Five ACCHSs (Urban and regional NSWi) | Health staff (n = 37) | A range of (typically unvalidated) questions that were being used by health services | Qualitative; (semi-structured group staff interviews) | Except in adult health check, screening was generally selective A range of questions were being used, and often did not quantify consumption clearly |
Conigrave et al. [5] | Aboriginal community-based groups (Urban NSW) | Aboriginal group participants (n = 47) | AUDIT (10 items) Rephrased for local English use | Quantitative plus researcher observation | AUDIT seemed easy to understand, as long as help was available for individuals not comfortable with reading While AUDIT-C may be enough to identify risk, the whole AUDIT provided a chance for the drinker to reflect on impacts of their drinking Community members appeared interested to receive their AUDIT score |
Lee et al. [30] | Alcohol and drug treatment service –mainstream (Urban, NSW) | Staff and Aboriginal patients (n = 21, n = 24 respectively) | AUDIT-C—modifiedj (3 items) | Mixed methods | An interviewer asked the AUDIT-C questions in a conversational style, adapting phrasing as needed Interviewer helped quantify drinking, as challenges with numeracy, a culture of sharing alcohol, and a common interpretation of the term ‘to drink’ as ‘to drink very heavily’ |
Ober et al. [23] | Prisons in Queensland | Aboriginal or Torres Strait Islander inmates (n = 395) | IRIS—modified (13 items); Asking about substance use in 12-months before prison | Quantitative: compared against CIDI | IRIS was compared against ICD10 criteria for substance use disorders (using CIDI) IRIS had a high sensitivity (94%) and low specificity (33%) |
Calabria et al. [32] | Primary care and community-based settings; (urban and regional NSW) | Aboriginal patients and community members (n = 136) | AUDIT-C (3 items) AUDIT-3 (1 item) AUDIT (10 items); Used plain English, no conversion to standard drinks | Quantitative: short forms of AUDIT compared against full AUDIT | AUDIT-C and AUDIT-3 (at appropriate cut-offs) compare favourably to full AUDIT |
Gray et al. [9] | Several sites- ACCHS and community-based (urban through to remote) | Researchers of five studies; plus overview of their results | AUDIT (10 items) | Review of five alcohol studies with Aboriginal people; workshops with the researchers | Few people (Aboriginal or otherwise) have a clear understanding of a ‘standard drink’ and the amounts poured or consumed as ‘a drink’ generally Understanding cultural context is key |
Noble et al. [35] | An ACCHS (Regional NSW) | Service clients (n = 188; of whom 72% were Aboriginal) | AUDIT-3mk; (2 items) Via touch-screen laptop | Quantitative: compared with 7-day retrospective drinking diary | 81% of Aboriginal current drinkers (n = 69) were equivalently classified by the two measures (weighted kappa = 0.77, 95% CI 0.73, 0.83) 7-day diary missed 31% of current drinkers who did not consume alcohol in the past week |