The development and validation of the Indigenous Risk Impact Screen (IRIS): a 13-item screening instrument for alcohol and drug and mental health risk

Transcription

1 Drug and Alcohol Review (March 2007), 26, The development and validation of the Indigenous Risk Impact Screen (IRIS): a 13-item screening instrument for alcohol and drug and mental health risk CARLA M. SCHLESINGER 1, CORALIE OBER 2, MOLLY M. MCCARTHY 1, JOANNE D. WATSON 1, & ANITA SEINEN 1 1 Centre for Drug and Alcohol Studies, Alcohol and Drug Service, The Prince Charles Hospital Health Service District, Brisbane, Queensland, Australia and 2 Queensland Alcohol and Drug Research and Education Centre, University of Queensland, Australia Abstract The study aimed to assess the psychometric properties of the Indigenous Risk Impact Screen (IRIS) as a screening instrument for determining (i) the presence of alcohol and drug and mental health risk in Indigenous adult Australians and (ii) the cut-off scores that discriminate most effectively between the presence and absence of risk. A cross-sectional survey was used in clinical and nonclinical Indigenous and non-indigenous services across Queensland Australia. A total of 175 Aboriginal and Torres Strait Islander people from urban, rural, regional and remote locations in Queensland took part in the study. Measures included the Indigenous Risk Impact Screen (IRIS), the Severity of Dependence Scale (SDS), the Alcohol Use Disorders Identification Test (AUDIT) and the Leeds Dependence Questionnaire (LDQ). Additional Mental Health measures included the Depression Anxiety and Stress Scale (DASS-21) and the Self-Report Questionnaire (SRQ). Principle axis factoring analysis of the IRIS revealed two factors corresponding with (i) alcohol and drug and (ii) mental health. The IRIS alcohol and drug and mental health subscales demonstrated good convergent validity with other well-established screening instruments and both subscales showed high internal consistency. A receiver operating characteristics (ROC) curve analysis was used to generate cut-offs for the two subscales and t-tests validated the utility of these cut-offs for determining risky levels of drinking. The study validated statistically the utility of the IRIS as a screen for alcohol and drug and mental health risk. The instrument is therefore recommended as a brief screening instrument for Aboriginal and Torres Strait Islander people. [Schlesinger CM, Ober C, McCarthy MM, Watson JD, Seinen A. The development and validation of the Indigenous Risk Impact Screen (IRIS): a 13-item screening instrument for alcohol and drug and mental health risk. Drug Alcohol Rev 2007;26: ] Key words: drug, dual diagnosis, indigenous, IRIS, psychometrics. Introduction and aims issues are of great concern within Australian Aboriginal and Torres Strait Islander (ATSI) communities [1 3]. Indigenous men are five times more likely to die of alcohol-related conditions than non-indigenous men, and Indigenous women are four times more likely to die from these causes than their non-indigenous counterparts [1 4]. Alcohol and other drug problems are also associated with mental health problems (including anxiety, depression and psychosis), inadequate nutrition and social functioning concerns (including domestic violence and so on [5 7]). Rates of hospitalisation for mental disorders due to psychoactive drug use is four to five times higher for Indigenous populations than for non-indigenous populations [1 2]. Although many screening tools have been developed and validated for mainstream Australians to identify those at risk of alcohol and drug and mental health Carla M. Schlesinger BBehSc Hons, PhD, Centre for Drug and Alcohol Studies, Alcohol and Drug Service, The Prince Charles Hospital Health Service District, Brisbane, Queensland, Australia, Coralie Ober RN, Assoc Dip Welfare, Grad Dip Teach, Queensland Alcohol and Drug Research and Education Centre, University of Queensland, Australia, Molly M. McCarthy BA Hons, Centre for Drug and Alcohol Studies, Alcohol and Drug Service, The Prince Charles Hospital Health Service District, Brisbane, Queensland, Australia, Joanne D. Watson BA Hist Hons, Grad Dip Teach, PhD, Centre for Drug and Alcohol Studies, Alcohol and Drug Service, The Prince Charles Hospital Health Service District, Brisbane, Queensland, Australia, Anita Seinen BBehSc Hons, Centre for Drug and Alcohol Studies, Alcohol and Drug Service, The Prince Charles Hospital Health Service District, Brisbane, Queensland, Australia. Correspondence to Carla M. Schlesinger, Centre for Drug and Alcohol Studies, Alcohol and Drug Service, The Prince Charles Hospital Health Service District, Brisbane, Queensland 4000, Australia. Carla_Schlesinger@health.qld.gov.au Received 10 November 2005; accepted for publication 3 October ISSN print/issn online/07/ ª Australasian Professional Society on Alcohol and Other Drugs DOI: /

2 110 Carla M. Schlesinger et al. problems, few screening tools have been validated for Indigenous Australians. Reports such as the Royal Commission into Aboriginal Deaths in Custody, the National Aboriginal Health Strategy, the National Aboriginal Mental Health Strategy and National Drug Strategy Complementary Aboriginal and Torres Strait Islander Alcohol and Drug Action Plan have highlighted the need to develop a culturally appropriate alcohol and drug and mental health screening tool that will encourage early identification for Indigenous people at risk. In line with these recommendations, the Indigenous Risk Impact Screen (IRIS) was developed to screen for both alcohol and drug and mental health issues, with the broad aim of (i) case finding at risk individuals, (ii) providing timely advice to clients about the potential risk and (iii) enabling Aboriginal and Torres Strait Islander and mainstream substance misuse and health agencies to better respond to client needs and provide appropriate and timely referrals. In contrast to other alcohol and drug screening measures, the IRIS screens for both alcohol and drug and mental health risk within the one instrument recognising the high co-occurrence of the two issues [5 7]. In addition, the IRIS measures alcohol and drug use in terms of total substance use, whereby the individual reports on their overall drug use, rather than each drug separately or their primary drug use only. The rational for this departure lies in the observed patterns of addictive behaviour in Indigenous populations. It is well established that Indigenous Australians use a variety of drugs, including alcohol, cannabis and amphetamines [2,3] and evidence of widespread and higher rates of polydrug use among Indigenous populations has been well documented [8 10]. Current paper and pencil methods for drug screening with polydrug users typically involve either (i) assessing the impact of the primary drug of use only or (ii) repeatedly administering the same measure (such as the Severity of Dependence Scale SDS) according to each drug used. As a consequence of such administration, there may be an underestimation of an individual s total drug use problem and an inability to adequately capture the full impact of interacting drug actions. The biological, psychological and social effects of polydrug use have important implications for the individual and although an individual may not be using a single drug at problematic levels, their total drug intake may be classed as problematic. This approach is consistent with Morgenstern et al. s [11] finding that the risk and dependence syndrome must be considered across the substances, as well as Gossop s [12] hypothesis that it may be useful to focus on the overall nature of an individual s drug use and dependence rather than focus on each substance separately. The following study aimed to develop a culturally appropriate measure of alcohol and drug use and mental health which would screen adequately for risk in an Australian Indigenous population. The study also aimed to determine cut-off scores for the IRIS in order to assess presence of risk. Ethical approval for the research was granted by the Queensland Health Research and Ethics Committee in Design and methods Design The study proceeded in two phases. The first phase saw the selection of a number of items known to be predictive of (i) alcohol and drug issues and (ii) mental health issues. The initial items were evaluated for face validity in three iterative stages, with each stage adding further modifications upon the previous stage. In the first stage the item set was focus tested with an academic focus group consisting of 12 experts in the Indigenous health, mental health and alcohol and drug areas. Secondly, a worker focus group containing 46 workers from the alcohol and drug area examined the revised item-set for content and administration feasibility. Finally, the item set was focus tested with 14 Aboriginal and Torres Strait Islanders to gauge acceptability and comprehension of the new item set. The second phase examined the psychometric properties of the IRIS and included examinations for construct validity and internal reliability, as well as checks for convergent validity across a number of well-established alcohol and drug and mental health measures. Sample characteristics Participants were 175 ATSI people (43% male) with a mean age of 35 years (SD ¼ 11.5). To obtain a crosssection of participants, individuals were recruited from a number of clinical (75%) and non-clinical services (25%), and from rural, remote, regional and urban areas across Queensland. Table 1 presents the demographic characteristics of the sample. Individuals were eligible for inclusion if they self-identified as being from an Aboriginal or Torres Strait Islander background and were 18 years of age or older. Exclusion criteria included obvious intoxication and an inability to understand English. Participants recruited were reimbursed for their time and travel expenses. Measures Participants were given a structured interview consisting of demographic questions and questions about their substance use and mental health. They were asked to complete the IRIS, which was rated according to a

3 Development and validation of IRIS 111 Table 1. Demographic characteristics of the sample Education level Graduate or postgraduate: 3.5% Secondary education: 41.8% Senior Certificate (year 10): 40.7% Primary education: 8.1% Did not complete primary education: 3.5% Employment status Student: 28.2% Employed (PT&FT): 28.2% Unemployed: 17.6% Home duties: 17.6% Indigenous status Aboriginal (not TSI): 81.7% TSI but not Aboriginal origin: 8.0% Aboriginal and TSI origin: 7.4% South Sea islander: 1.0% Recruitment location Dirranbandi: 15.0% Brisbane: 61.3% Charters Towers: 5.0% Palm Island: 5.0% Townsville: 5.0% Wuchoppern: 8.8% ensure that results were disseminated among these same communities through regular contact with the IRIS Project Team. The health workers then invited their clients to participate in the study. Participants were reimbursed for each interview attended, with $15 reimbursed for the first assessment session and $10 for the second. Participants were administered the measures according to study protocols, usually by the worker who had recruited them. They were also asked questions about their drug intake, including their alcohol and drug use over the previous 2 weeks. Participants reported on the average quantity of each drug consumed per day and number of days consumed over the previous fortnight. These two figures were multiplied to provide a drug consumption score for each drug. Workers chose a location that preserved confidentiality. During administration, the worker clarified comprehension issues that arose during the interview. Any information that indicated the participant was at risk of harm to self or another was managed according to the service s policies and procedures. Participants were re-interviewed at 4 weeks to establish temporal stability of the IRIS interview. variable Likert scale and were also administered the following screening instruments: Depression, Anxiety and Stress Scale (DASS-21): the DASS is a 21-item questionnaire that measures the three factors during the past 7 days [13]. Self-reporting Questionnaire 20 (SRQ): the SRQ was designed to measure the presence of mental health symptoms within the last 30 days and it has been validated previously with a number of Indigenous populations [14]. Leeds Dependence Questionnaire (LDQ): the LDQ was designed to measure symptoms of dependence without reference to a specific drug, and it has been validated previously with Indigenous populations [15]. Alcohol Use Disorders Identification Test (AUDIT): the AUDIT was designed to detect hazardous and harmful levels of alcohol consumption [16]. Severity of Dependence Scale (SDS): the SDS was designed to measure the degree of dependence experienced for different types of drugs [17]. Procedure In order to test the psychometric properties of the instrument, key community health workers who had self-nominated to conduct the research in north, south and central Queensland were trained in the research protocols. This was part of a community capacity building strategy that sought to increase the research skills of workers within their own communities and to Analyses Principle axis factoring was conducted to examine the underlying factor structure of the full IRIS questionnaire and to examine the internal consistency of the factors. Items were eliminated from the original IRIS according to their factor loadings, face validity and overlap with other items. The new item set was examined for convergent validity with other well established screening tools, and cut-offs for each subscale were determined using a receiver operating characteristics (ROC) curve analysis. The IRIS questionnaire was re-tested on participants from the original sample to determine the reliability of the questionnaire at 4 weeks. Results Of the total sample, 64% had used alcohol or other drugs in the past 2 weeks, with 54% drinking alcohol [mean (M) ¼ 3.5 days in the past 2 weeks], 25% using cannabis (M ¼ 7.6 days in the previous 2 weeks), 2.3% using amphetamines (M ¼ 8.0 days in the previous 2 weeks) and 1.7% using inhalants (M ¼ 6.0 days in the past 2 weeks). Polydrug use was prevalent in the group with up to six drugs being used in the past 2 weeks (M ¼ 3.9; SD ¼ 0.9). Bivariate correlations were calculated between each drug according to frequency of use days during the previous 2 weeks. Significant correlations were found between the use of alcohol and both cannabis (r ¼ 0.24, p ) and amphetamine (r ¼ 0.28, p ). Cannabis was also correlated

4 112 Carla M. Schlesinger et al. significantly with both inhalant (r ¼ 0.20, p ) and amphetamine use (r ¼ 0.25, p ). Factor structure of the IRIS The factor structure of the original IRIS questionnaire was examined using principle axis factoring. Table 2 presents the results of the principal axis factoring, with items selected on the basis of high loadings (above 0.3) on their appropriate factor. The a priori two-factor solution confirmed the existence of two underlying constructs consistent with alcohol and drug and mental health conceptualisations. Factor loadings ranged from moderate to high. The first factor had an eigenvalue of 6.8 (28.5% of the variance) and the second factor had an eigenvalue of 3.3 (13.8% of the variance), with the two factors accounting for 42.2% of the variance. Internal consistency of the factors according to the original item set was examined using Cronbach s alpha, and this analysis revealed high internal consistency ratings for each factor (alcohol and drug factor: a ¼ 0.84; mental health factor: a ¼ 0.81). The IRIS item set was reduced to 13 items to produce a quick screen, with items chosen for inclusion based on high kappa values, face validity and reduced overlap with other factor items. The revised IRIS instrument (see Appendix 1) contains a total of 13 items (1 7 representing alcohol and drug use, and items 8 13 representing mental health symptomatology). The IRIS alcohol and drug subscale is calculated by adding scores 1 7 yielding a score ranging from 7 to 28. The IRIS mental health subscale is calculated by adding scores 8 13, yielding a score ranging from 6 to 18. Convergent validity of the IRIS The IRIS alcohol and drug subscale was significantly correlated with self-reported frequency and quantity of alcohol and drug use over the previous two weeks (r ¼ 0.47; p ) and evidenced strong convergent validity with other well established measures. The IRIS mental health subscale also evidenced strong convergent validity with other mainstream scales (see Table 3). Establishing cut-offs for the IRIS A receiver operating characteristics (ROC) curve analysis was conducted on the revised IRIS to determine optimal cut-off points for the two subscales. The LDQ was chosen as the reference point for problematic use of alcohol or drugs. A cut-off score Table 2. Principle axis factoring of the original IRIS questionnaire Factor Item In last 6 months have you needed to drink or use more to get effects you want? When cut down or stopped drinking or using drugs in the past, have you experienced any symptoms How often do you drink or use more than expected? Has anyone been concerned with drinking or drug use? How often experienced blackouts or unable to remember from use? Do you feel out of control with drinking or drug use? How difficult to stop or cut down your drinking or drug use? What time of day start drinking or using drugs? How often do you find entire day involved drinking or drug use? How often do you have difficulty in paying for food or bills? Have you had any legal strife in past 6 months? In the last week have you felt unsafe in your home? Have you been scared about the safety of your family in household? How often do you feel down in the dumps? How often do you feel that life is hopeless? Do you suddenly feel scared for no reason? How often do you feel nervous or scared? Do you worry much? How often do you feel restless and that you can t sit still? Is it hard for you to fall asleep or stay asleep at night? Have you ever experienced hallucinations from using too many drugs? Have you been situations where you or people close have experienced a form of abuse? Do past events still affect your well being today? In the last day have you seriously thought about hurting yourself or others? Note: Items were chosen for inclusion based on high kappa values, face validity and reduced overlap with other factor items.

5 Development and validation of IRIS 113 Table 3. Convergence between the IRIS alcohol and drug scale and other alcohol and drug screening instruments scales Leeds Dependence Questionnaire Severity of Dependence Scale Alcohol Use Disorders Identification Test Mental Health scales IRIS alcohol and drug risk subscale 0.74** 0.55** 0.62** IRIS mental health and emotional subscale Depression Anxiety and 0.62** Stress Scale Anxiety Depression Anxiety and 0.71** Stress Scale Depression Self Report Questionnaire 0.74** **Pearson s correlation is significant at the 0.01 level (twotailed). Figure 1. ROC curve for the IRIS drug and alcohol subscale. of 1 for the LDQ is suggested to demonstrate low to moderate dependence and so this was used as the true value for risky alcohol and drug use in the population (Alcohol Advisory Council of New Zealand, 1999). Figure 1 shows the ROC for the IRIS alcohol and drug subscale. A score of 10 was chosen as the cut-off for the IRIS alcohol and drug subscale giving a sensitivity of 65% and a specificity of 86%. A ROC curve analysis was similarly conducted to determine cut-offs for the mental health subscale, with the SRQ chosen as the reference point for the presence or absence of significant mental health symptomatology as it correlated highly with the IRIS mental health subscale. The ROC curve revealed a score of 11 to be the cut-off point that obtained the best balance of sensitivity (83%) and specificity (84%). Figure 2 displays the ROC curve for the IRIS mental health subscale. Validation of cut-off scores An independent samples t-test was conducted to examine whether scores determined to be symptomatic and non-symptomatic according to the IRIS alcohol and drug subscale were significantly different in terms of the frequency and quantity of alcohol consumed in the past 2 weeks. Results indicated that those participants classified as being at risk on the IRIS alcohol and drug subscale had significantly higher drinking frequency and quantity than those classified as non-symptomatic, further validating the subscale norms (see Table 4). Considering binge drinking is a significant concern in Indigenous communities, and the majority of the current sample (54%) indicated alcohol as their primary Figure 2. ROC curve of the IRIS mental health subscale. substance of use, a box-plot for males and females was generated to examine whether the IRIS alcohol and drug cut-off was able to capture binge drinking. The NHMRC guidelines for drinking were used to identify those who were drinking at risky levels [18]. As seen in Figure 3, all male participants in the non-symptomatic group were drinking below the NHMRC guidelines for short-term, high-risk drinking (11 or more standard drinks on any one day). Of female participants who were classified as non-symptomatic, seven participants were

6 114 Carla M. Schlesinger et al. Table 4. Frequency and quantity of alcohol consumed according to IRIS cut-offs IRIS drug and alcohol subscale norms n Mean SD t d.f. Significance Quantity of alcohol consumed on average day (standard drinks/day) Non-symptomatic Symptomatic Frequency of alcohol use days in past 2 weeks Non-symptomatic Symptomatic Figure 3. Quantity of alcohol consumed by male participants in an average drinking episode in the previous 2 weeks. drinking above the NHMRC guidelines for short-term, high-risk drinking (seven or more standard drinks in any one day), as shown in Figure 4. A frequency table was generated to determine whether the IRIS drug and alcohol scale captured polydrug use successfully. The test detected that only two participants of the 30 who were using two or more drugs in the past 2 weeks were classified as non-symptomatic. The box-plot displayed in Figure 5 shows that the number of drugs used in the previous 2 weeks (e.g. a person reporting use of alcohol and amphetamine in the previous 2 weeks is calculated as two) is higher for those classified as symptomatic than those classified as non-symptomatic by the IRIS alcohol and drug subscale. A t-test confirmed that this mean difference between number of drugs used in the previous 2 weeks and the IRIS classification was significant (t ¼ 710.8, p , df ¼ 153). Of Figure 4. Quantity of alcohol consumed by female participants in an average drinking episode in the previous 2 weeks. interest, 46 of the 73 participants who used one drug in the previous 2 weeks were classified as symptomatic, indicating that although polydrug users are at increased risk, single-drug use at high quantities is still prevalent and able to be captured using the screen. Temporal stability of the IRIS The 4-week temporal stability of the measure was assessed based on 95 participants who were followedup (54% of the original sample). No significant differences existed between participants who were interviewed and those who had dropped out at the second time-point according to gender, age or drug intake. Subscale analyses revealed significant associations between the two time-points (alcohol and drug subscale, r ¼ 0.79, p ; mental health subscale, r ¼ 0.81, p ). Bivariate coefficients between

7 Development and validation of IRIS 115 Figure 5. Number of drugs consumed by participants classified as symptomatic and non-symptomatic on the IRIS alcohol and drug subscale. Table 5. Test retest coefficients by question at pre-test and 4-week post-test Pre-test items Post-test item 1. In last 6 months needed more to get 0.62* effects you want? 2. When cut down or stopped are there are 0.60* any symptoms? 3. How often drink or use more than 0.53* expected? 4. Do you feel out of control with drinking 0.69* or drug use? 5. How difficult to cut down or stop? 0.69* 6. What time of day start drinking or drug 0.70* use? 7. How often do you find entire day 0.66* involved drinking or drug use? 8. How often do you feel down in the 0.44* dumps? 9. How often do you feel that life is 0.54* hopeless? 10. How often do you feel scared or nervous? 0.61* 11. Do you worry much? 0.67* 12. How often do you feel restless and that 0.63* you can t sit still? 13. Do past events still affect your well being today? 0.65* *p level (two-tailed). each pre-test and post-test item revealed strong relationships, with good subscale coefficients (see Table 5). Discussion and conclusions This study developed and tested the psychometric properties of a screening tool developed specifically for Indigenous Australians to measure comorbid alcohol and drug as well as mental health risk. Factor analysis supported the theoretical constructs developed, with the resulting two-factor solution corresponding to an alcohol and drug subscale and a mental health subscale, evidencing strong conceptual relationships with related measures and strong temporal stability across a 4-week period. The IRIS demonstrates strong internal consistency, convergent validity and evidenced valid cut-offs for determining symptomatic individuals in terms of drug and alcohol use and mental health problems in Indigenous populations. Results showed that the IRIS was able to capture polydrug use accurately, with all but two participants using two or more drugs being classified as symptomatic on the IRIS alcohol and drug subscale. The IRIS drug and alcohol subscale shows promise in terms of its ability to screen for binge drinking, a common concern in Australian Indigenous communities, and one that is often not captured by well-established drug and alcohol screens that assess only dependence. In male participants, the IRIS drug and alcohol subscale classified as symptomatic all participants who were drinking at short-term risky levels, according to NHMRC guidelines. For female participants, the screen was less accurate and further work is needed to improve the IRIS alcohol and dug subscale s ability to identify female binge drinkers. Further research could be conducted with the IRIS alcohol and drug subscale to generate different cut-offs for binge drinking in females. The LDQ was chosen as the reference point for alcohol and drug risk in the population as it measures lower levels of dependence, is applicable across drugs and has been found previously to have utility with Indigenous populations. However, as the LDQ has a cut-off score of 1 of a possible score of 30 indicating the presence of mild dependence, the IRIS cut-off generated would be equally sensitive. For this reason a cutoff with a better balance of specificity against sensitivity was chosen for the alcohol and drug subscale, rather than simply selecting a high sensitivity rating, as would be expected with a screening instrument interested in detecting risk. The lack of a relevant gold standard screening tool to use as reference for the IRIS alcohol and drug subscale is problematic, not only in terms of identifying appropriate cut-offs for the subscale but also for the area of alcohol and drug research generally. Screening tools that (i) could be used across drugs and capture the full impact of polydrug use and (ii) capture risky drug use in addition to dependence, similar to the AUDIT which assesses risky alcohol use [16], would greatly improve

8 116 Carla M. Schlesinger et al. screening and early intervention in communities where multiple drug use is problematic and prevalent. Considering the majority of participants were alcohol drinkers, the IRIS requires further validation in terms of other drug norms. Nevertheless, alcohol misuse at present is a significant issue in Indigenous communities and the availability of a culturally appropriate screening tool would facilitate widespread identification of people in need of intervention for this problem. The IRIS screen would be feasible for use in general health-care settings as it is delivered quickly and easily comprehensible to Indigenous clients. It is expected that use of the IRIS as a risk screening instrument would enable early identification of (i) alcohol and drug misuse and (ii) mental health risks, and enable health personnel to target their response to client needs in a timely manner. In clinical practice, it is recommended that the IRIS be used in a culturally sensitive manner where the Indigenous client has choice about whether to complete the screen. If the client s score indicates risk on either or both subscales, it is recommended that they receive a brief intervention to feed back results of the screen and to work towards goals and strategies to minimise the risks identified. Any brief intervention should highlight that both alcohol and drug and mental health issues are likely to be connected. At a service level, it is recommended that the IRIS form a routine part of clinical practice for Indigenous clients to ensure appropriate identification, intervention and referral. Embedding the screen within service protocols will increase service ownership and will allow for the screen to be utilised in a manner which aids service delivery. In conclusion, the utility of the IRIS as a measure of risk of drug and alcohol use as well as mental health risk has been validated statistically in the present study. While further investigation is needed, it is anticipated that this instrument will be implemented easily and will evidence utility in a range of health settings. Training and support in the appropriate use of the IRIS is available through the Centre for Drug and Alcohol Studies, Alcohol and Drug Service, The Prince Charles Hospital Health Service District. Acknowledgements This research was funded by the Commonwealth Department of Health and Aged Care and the Alcohol Tobacco and Other Drugs Unit who endorsed this expenditure. The authors wish to thank the Academic Committee (including Professor John B. Saunders, Dr Maggie Brady, Dr Sharyn Watts, Professor Dennis Gray, Dr David Cutts, A/Professor Ted Wilkes, Dr Noel Hayman, Dr Mal Miller and Dr Tracey Westerman), the Steering committee members (Mr Mark Fairbairn and Dr Kevin Lambkin) and the Queensland Workers who recruited as part of the study (Alcohol Tobacco and Other Drugs Services Cairns, Cooktown, Longreach, Redlands, Rockhampton, Townsville and Weipa Health Services; Wuchopperen Community Controlled Health Service, Indigenous Community Team, Biala, The Prince Charles Hospital Health Service District and Yulu-Burri-Ba Aboriginal and Torres Strait Islander Community Health Service. References [1] Australian Bureau of Statistics. The Health and Welfare of Australia s Aboriginal and Torres Strait Islander Peoples. Cat. no Canberra: ABS, [2] Australian Institute of Health and Well-Being. Statistics on Drug Use in Australia Drug Statistics Series no. 15. Cat no. PHE62. Canberra: AIHW, [3] Gray D, Saggers S. Substance Misuse. In: Thomson N, ed. The health of Indigenous Australians. Australia: Oxford University Press, 2003:1 10. [4] Brady M, Dawe S, Richmond R. Expanding knowledge among Aboriginal service providers on treatment options for excessive alcohol use. Drug Alcohol Rev 1998;17: [5] Burdekin B. Report of the national inquiry into the human rights of people with mental illness. Human rights and mental illness. Canberra: Australian Government Publishing Service, [6] Swan P, Raphael B. National consultancy report on Aboriginal and Torres Strait Islander mental health. Part 1. Ways forward. National Mental Health Strategy. Canberra: Australian Government Publishing Service, [7] Trudgen R. Why warriors lie down and die. Darwin: Aboriginal Resource and Development Services, [8] Holly C. Review of literature on injecting drug use within urban Indigenous Communities, South Australia: Aboriginal Drug and Alcohol Council South Australia Inc., [9] Burns CB, D Abbs P, Currie BJ. Patterns of petrol sniffing and other drug use in young men from an Australian aboriginal community in Arnhem Land, Northern Territory. Drug Alcohol Rev 1995;14: [10] Gray D, Morfitt B, Ryan K, Williams S. The use of tobacco, alcohol and other drugs by young Aboriginal people in Albany, Western Australia. Aust NZ J Public Health 1997;21:71 6. [11] Morgenstern J, Langenbucher J, Labouvie EW. The generalizability of the dependence syndrome across substances: an examination of some properties of the proposed DSM-IV dependence criteria. Addiction 1994;89: [12] Gossop M. The web of dependence. Addiction 2001; 96: [13] Lovibond SH, Lovibond PF. Manual for the depression, anxiety and stress scales 9, 2nd edn. Sydney: Psychological Foundation, [14] Beusenberg M, Orley JA. User s guide to the Self-Reporting Questionnaire (SRQ). Geneva: Division of Mental Health, WHO, [15] Paton-Simpson G, MacKinnon S. Evaluation of the Leeds Dependence Questionnaire (LDQ) for New Zealand. Research Monograph Series no. 10. Alcohol Advisory Council of New Zealand, [16] Saunders JB, Aasland OG, Babor TF, De La Fuente JR, Grant M. Development of the Alcohol Use Disorders Identification Test (AUDIT): WHO Collaborative Project on early detection of persons with harmful alcohol consumption II. Addiction 1993;88:

9 Development and validation of IRIS 117 [17] Gossop M, Darke S, Griffiths P, Hando J, Powis B, Hall W. Strang, J. The Severity of Dependence Scale (SDS): psychometric properties of the SDS in English and Australian samples of heroin, cocaine and amphetamine users. Addiction 1995;90: [18] National Health and Medical Research Council (NHMRC). Australian alcohol guidelines: health risks and benefits. Canberra: NHMRC, Appendix 1. IRIS item set Question Content domain Response alternatives 1. In the last 6 months have you needed to drink or use more to get the effects you want? 2. When you have cut down or stopped drinking or using drugs in the past, have you experienced any symptoms, such as sweating, shaking, feeling sick in the tummy/vomiting, diarrhoea, feeling really down or worried, problems sleeping, aches and pains? 3. How often do you feel that you end up drinking or using drugs much more than you expected? 4. Do you ever feel out of control with your drinking or drug use? 5. How difficult would it be to stop or cut down on your drinking or drug use? 6. What time of the day do you usually start drinking or using drugs? 7. How often do you find that your whole day has involved drinking or using drugs? 8. How often do you feel down in the dumps, sad or slack? Mental health and emotional 9. How often have you felt that life is hopeless? Mental health and emotional 10. How often do you feel nervous or scared? Mental health and emotional 11. Do you worry much? Mental health and emotional 12. How often do you feel restless and that you can t sit still? 13. Do past events in your family, still affect your today (such as being taken away from family)? Mental health and emotional Mental health and emotional 1 ¼ No 2 ¼ Yes, a bit more 3 ¼ Yes, a lot more 1 ¼ Never when I stop 3 ¼ Yes, every time 2 ¼ Once a month 3 ¼ Once a fortnight 4 ¼ Once a week 5 ¼ More than once a week 6 ¼ Most days/every day 3 ¼ Often 4 ¼ Most days/every day 1 ¼ Not difficult at all 2 ¼ Fairly easy 3 ¼ Difficult 4 ¼ I couldn t stop or cut down 1 ¼ At night 2 ¼ In the afternoon 3 ¼ Sometime in the morning 4 ¼ As soon as I wake up 3 ¼ Often 4 ¼ Most days/every day