Social demography of alcohol-related harm to children in Australiaadd_

Transcription

1 bs_bs_banner RESEARCH REPORT doi: /j x Social demography of alcohol-related harm to children in Australiaadd_ Anne-Marie Laslett 1,2, Jason Ferris 1,4, Paul Dietze 3 & Robin Room 1,2 Centre for Alcohol Policy Research, Turning Point Alcohol and Drug Centre, Fitzroy, Victoria, Australia, 1 School of Population Health, University of Melbourne, Melbourne,Victoria,Australia, 2 Centre for Population Health, Burnet Institute, Melbourne,Victoria,Australia 3 and Eastern Health Clinical School, Monash University, Victoria, Australia 4 ABSTRACT Aims This study seeks to establish the prevalence alcohol-related harms to children (ARHC) that occur because of others drinking in the general population and examine how this varies by who was reported to have harmed the child and socio-demographic factors. Design and setting A randomly selected cross-sectional national population telephone survey undertaken in 2008 in Australia. Participants A total of 1142 adult respondents who indicated they lived with or had a parental/carer role for children. Measurements Questions included whether children had been negatively affected in any way, left unsupervised or in an unsafe situation, verbally abused, physically hurt or exposed to serious family violence because of others drinking in the past year. Findings Twenty-two per cent of respondents reported children had been affected because of another s drinking in the past year; 3% reported substantial harm. Respondents most commonly reported that children were verbally abused because of others drinking (9%). Participants in single-carer households were more likely to report ARHC than participants in households with two carers, and participants who drank weekly were more likely to report ARHC than those who did not drink. Conclusions Almost a quarter of those with a caring role for children in Australia reported that a child or children with whom they lived or for whom they were responsible have been affected adversely by others alcohol consumption in the past year. The problem extends across the social spectrum, but children in single-parent homes may be at higher risk. Keywords Alcohol, carer drinking, child abuse, child maltreatment, harm to others, population level data, social location, verbal abuse. Correspondence to: Anne-Marie Laslett, Centre for Alcohol Policy Research, Turning Point Alcohol and Drug Centre, Fitzroy, Victoria 3065, Australia. anne-mariel@turningpoint.org.au Submitted 2 June 2011; initial review completed 14 September 2011; final version accepted 22 December 2011 INTRODUCTION Alcohol is widely used, not always responsibly, and mainly in leisure time. For families with children, most leisure time is also family time [1]. Families with young children tend to have relatively young adults as parents and heavy drinking is common at this life-stage, but beginning to wane [2,3]. While it is considered unseemly for an intoxicated adult to be in charge of young children [4], children are commonly exposed to heavy drinking of their parents and others at social occasions [5]. Dawe et al. [3] estimated that 13.2% of children are at risk of exposure to short-term risky drinking in Australian households by at least one adult, and in the United Kingdom an estimated 30% (or million) children live with at least one binge drinking parent [6]. These researchers acknowledge that this drinking is not necessarily linked with harm, yet evidence, largely from case control studies, shows that children of heavy drinkers are at greater risk of abuse and neglect [7,8]. Knowledge about the role of alcohol in child abuse has been based primarily on cases known to authorities [9,10]. Recent Australian data indicate that alcohol is a risk factor in an estimated cases of substantiated child abuse and cases of domestic violence [11]. On the basis of this case-load, a connection between problematic drinking and child maltreatment in the general population has been assumed, despite evidence that cases in the general population with such problems may differ from clinical cases [10,12]. Very few studies of the general population report on the prevalence of child maltreatment, let alone whether or not carer intoxication or problematic drinking was involved. In Italy in 1998, self-reported high and average, compared with

2 Alcohol-related harm to children 1083 moderate, parental alcohol consumption, was associated with severe but not minor physical violence towards children [13]. Straus et al. noted that 2.3% of parents in the United States [12] self-reported being so drunk or high in the last year they had a problem taking care of their child. A third study, examining the life-time history of a range of social problems in a large Californian sample [8], found that 23% of adults reported the presence of a heavy drinking parent during childhood. Among the 26% who reported this, or use of street drugs, by one or more parents (heavy drinking and substance using parents were analysed together, although only 5% reported street drug use), 22% reported psychological abuse, 19% reported physical abuse and 34% reported sexual abuse as a child. In Australia and elsewhere, knowledge about the relationship between heavy drinking and child rearing in the general population has been limited to superficial information on the extent of drinking by parents. Dawe et al. [3] summarize these data, noting the deficiencies. In this study, we seek to move beyond these limits, to determine (i) the prevalence of alcohol-related harm to children (ARHC) in the population because of someone else s drinking (not including the respondent s drinking) and (ii) whether rates vary by relationship and sociodemographic factors. METHODS Data The Alcohol s Harm To Others (AHTO) survey involved a national random sample of 2649 Australians aged 18 years or older who completed a computer-assisted telephone interview (CATI) in November and December 2008 [14]. Eligibility was restricted to people from private households able to be interviewed in English, with the respondent selected from within the household by the next-birthday method. The cooperation rate was 49.7% (the proportion of responders among the eligible people actually contacted) and the response rate was 35.2%, based on the standards of the American Association of Public Opinion Research [15], including non-contacts estimated to be eligible in the denominator. The sample was generally representative of the national population in income type and level [16], although men and young people, less formally educated and overseas-born Australians were under-represented. Data were weighted inversely by sample selection probability in order to reproduce the age, sex and geographic composition of the Australian adult population in the 2006 census, with the weighted total number set equal to the unweighted sample size. See Wilkinson et al. for more detail on the AHTO survey methods [16]. For the purpose of this paper a parental subset of the study population was identified, consisting of all respondents who indicated that they were living with children 17 years of age or younger, and all others with parental responsibility for children 17 years of age or younger. ARHC variables The parental subset of respondents were asked five specific questions concerning harm that a child or children of theirs had experienced in the past year as the result of someone else s drinking (see Table 1, asterisked questions). These questions, which do not include harm caused by the respondent s own drinking, were simplified versions of the official definitions of types of primary harm, excluding sexual harm, commonly investigated by Australian child protective services [17]. The respondents were also asked a summative question: how much has Table 1 The percentage of respondents (families) reporting ARHC by maltreatment type and level of effect (n = 1142, %). Total (n) Total (%) 95% CI (%) Because of someone else s drinking how many times in the last 12 months: *Were children left in an unsupervised or unsafe situation? 40 3 (2,5) *Were children yelled at, criticized or verbally abused? 97 9 (7,11) *Were children physically hurt? 16 1 (1,2) *Did children witness serious violence in the home? 34 3 (2,4) *Was a protection agency or family services called? (0.1, 0.8) % Families reporting one or more of above a 135 a 12 (10,14) How much has the drinking of other people negatively affected your children/the children you are responsible for? b A lot 40 3 (2,4) A little (12,16) Specifically affected in any way or affected a lot or a little (19,24) a Total ns and %s for rows do not add to 135 and 12% as respondents may have reported that their children experienced more than one type of abuse; b 12 people were excluded from the denominator because they did not report any level of effect (i.e. did not answer a lot, a little or not at all). CI: confidence interval.

3 1084 Anne-Marie Laslett et al. the drinking of other people negatively affected (this child/these children) in the last 12 months?. The response options were a lot, a little or not at all. ARHC due to the drinking of others, the outcome or dependent variable used was coded as 1 if a positive response to one or more of the five specific questionnaire items above was reported, or a lot or a little harm was reported, and a 0 otherwise. Calculating the (unweighted) percentage of children affected in the past year because of others drinking More than one child may have been affected in the household and/or family. Respondents were asked whether they either lived with children or had some parent-like responsibility for children and how many children fell into these categories. A conservative approach to calculating the prevalence of alcohol-related harm to children because of the drinking of others was used whereby it was assumed that each respondent was only referring to one child who was harmed. The least conservative measure assumed that the respondent was referring to all children for whom the respondent had responsibility. 1 Independent variables Respondents who reported any of the specific items were asked what was the relationship to the children? of the person or individuals whose drinking adversely affected them. The response categories for these identified individuals were: caregiver (parent, step-parent or and/or guardian); sibling; another relative; family friend or person with whom the child comes into contact, such as a sports coach, teacher, priest or someone else. Demographic data on age, gender, educational background, work status and postcode of the respondent were collected and categorized as described in Table 3. A socio-economic status variable was constructed using a Socio-Economic Index for Areas (SEIFA) of disadvantage [18] measure with participant postcodes grouped into quintiles according to SEIFA score. Household family structure was coded as two carers with children, single carer with children and other. Analysis Data analyses were undertaken using STATA version 11 [19]. Respondents with children in or outside the household were the unit of analysis. Raw numbers and weighted percentages with confidence intervals were calculated for annual prevalence estimates on alcoholrelated harm variables. Weighted bivariate and multivariate logistic regression analyses were undertaken to examine associations between the independent variables and ARHC. Bivariate relationships significant at the P < 0.1 level were added to the adjusted model; significant variables were retained. RESULTS Sample characteristics There were 1142 respondents in the parental subset who had parental responsibility for 2457 children (of whom 2004 lived with respondents). On average, each respondent was associated with 2.1 children and lived with 1.8 children. Table 1 shows that respondents reported most commonly that children had been yelled at, criticized or otherwise verbally abused because of someone else s drinking (9%). Fewer respondents reported that children were left unsupervised or in unsafe situations because of others drinking (3%), that children witnessed domestic violence (3%) or that children were physically hurt (1%). Very few reported that family services were called (<1%). In total, 12% of respondents reported that children had been affected in one or more of these specified ways. The total number of types of ARHC reported (excluding the five calls to protection agencies or family services) was 187. On average, respondents who reported any of the specific harms to children due to someone else s drinking reported more than one type of harm (i.e. 1.4 harms: 187 types of harm were reported by 135 people). Children who were reported to have experienced any specific harm most commonly experienced three incidents (mode), and on average experienced 12.4 incidents of harm in the past 12 months. Approximately one in six (17%) respondents reported that a child/children had been affected a lot (3%) or a little (14%) by someone else s drinking. ARHC of any type (reporting either a specific form of ARHC or any negative effect) was identified by 22% of respondents. Just under half of the respondents (123, or 48% of those who identified a lot or a little negative effect) reported that a child in their family was affected adversely by others drinking, but not in ways that were covered by the five concrete questionnaire items. A total of 135 respondents reported on 195 specific alcohol-related harms, attributed to 138 people. Respondents were asked only about whom they held responsible for the specific ARHC. Table 2 depicts the relationship between the child/children experiencing alcohol-related harm and the identified people implicated. Approximately 50% of specified ARHC were forms of verbal abuse, 17% involved exposure to domestic violence, 21% involved 1 Children under the age of 18 were assumed to be children of someone within the household. A small number of these may have been the partner of an adult for example, a 17-year-old partner of a 19-year-old.

4 Alcohol-related harm to children 1085 Table 2 The number of identified specific alcohol-related harms to children by relationship of affecting drinker, and percentage of harms attributed to each relationship (percentages shown in italic type). Relationship (n respondents) Parent (68) Siblings (12) Other relative (19) Family friend (25) Other (14) Total (138) Because of someone else s drinking how many times in the last 12 months: n % n % n % n % n % n % Were children yelled at, criticized or verbally abused? (n yes/% of cases by relationship) Did children witness serious violence in the home? Were children left in an unsupervised or unsafe situation? Were children physically hurt? Were child protection/family services called? Total number of identified harms Percentage of total harms attributed to people in that relationship to the child (row %) lack of supervision and 8% involved physical injury. Similar distributions of harms were noted for all relationship types. The most common harm reported from siblings was verbal abuse (56%), followed by physical harm and poor supervision (both 17%), although ARHC attributed to siblings drinking were slightly more likely to involve physical injury than harms from respondents overall. ARHC from other relatives were more likely to involve an incident of serious violence in the home, and harms involving children being left unsupervised were somewhat more likely to be attributed to a family friend s drinking. The people most likely to have been responsible for the largest percentage of specified ARHC (see bottom row of Table 2) were parents, step-parents or guardians (52%), followed by siblings and other relatives (21%) and family friends (16%). This distribution of harm across relationship types for each of the individual types of abuse (percentages not shown) was similar to the overall pattern: in approximately 50% of verbal abuse, exposure to domestic violence and lack of supervision incidents reported, parents/carers were held responsible. These data were also examined using the number of cases reporting harm from each particular relationship source (i.e. as a percentage of column respondent totals and not harm totals). These percentages are based upon small numbers and need to be treated with caution. Of the 68 parents responsible for ARHC, the majority (n = 52) were identified because the harm involved verbal abuse (76%). However, 31% of these parents reportedly left children unsupervised, 25% were reportedly involved in serious domestic violence that was witnessed by the child and 10% were reported to have physically hurt their children because of the person s drinking. Types of ARHC reported by the other relationship types held responsible were distributed similarly, except that other relatives were held responsible for proportionally more children witnessing serious violence (37%); and siblings were held responsible for more alcohol-related physical harms (25%). For all relationships, verbal abuse was the most commonly identified form of harm. The prevalence of children affected in the past year because of others drinking The 2457 children identified as being in the care of the parental subset were comprised of 1397 children aged 0 12 years, 607 children aged years with whom the respondent lived and 453 children aged 0 17 years with whom the respondent did not live but for whom they had parental responsibility. In total, 258 respondents identified that one or more children had been affected in one or more of the specified ways or affected negatively in some way. Thus, the conservative estimate was that 11% of children were affected. In total, 604 children were associated with the 258 respondents reporting some form of alcohol-related harm to the children; if all these children were harmed, the prevalence rate would be 25% of all children. Thus, the true prevalence rate lies somewhere between 11 and 25% of children being harmed in any way as the result of the drinking of people other than the respondent. Analysing relationships between ARHC and family socio-demographic factors Table 3 presents information on the percentage of respondents, across a range of socio-demographic characteristics, who reported that children had been affected because of others drinking in any way. Bivariate analysis revealed that respondents in single-carer households, compared to two-carer households, were more than twice as likely to report that children had been harmed because of others drinking. Weekly drinking respondents were

5 1086 Anne-Marie Laslett et al. Table 3 Percentage and odds ratios of alcohol related harm to children (ARHC) by selected socio-demographic characteristics (n = 1142). Independent variable (n a ) % reporting specific ARHC % reporting any ARHC Unadjusted odds ratios (CI) Adjusted odds ratios (CI) Gender of respondent Male (417) Ref Female (725) (0.76,1.50) Age of respondent (years) (168) Ref (920) (0.81, 2.04) (54) (0.26, 1.40) Household family structure Two carers and children (828) Ref Single carer (198) *** (1.79, 3.99) 2.66 (1.63, 4.36)* Other (116) b (1.79,3.99) 0.92 (0.48, 1.76) Education of respondent <Secondary (240) Ref Secondary (452) (0.74, 1.75) Post-secondary (436) (0.52, 1.26) Family income type Paid work (809) Ref Other c (332) (0.49, 1.02) Income Less than $ (224) Ref $ (189) * (0.30, 0.88) 0.66 (0.38, 1.16) $ or more (519) * (0.39, 0.90) 0.77 (0.47, 1.25) SEIFA score of neighbourhood 1 Low status (157) Ref 2 (185) (0.47, 1.57) 3 (240) (0.80, 2.39) 4 (259) (0.60, 1.91) 5 High status (297) (0.50, 1.48) Respondent s drinking status Not in previous year (512) Ref Monthly (504) (0.79, 1.60) 1.21 (0.83, 1.78) Weekly (105) * (1.00, 2.78) 2.01 (1.13, 3.57) a Total observations for socio-demographics do not add to 1142 due to missing responses (typically less than 1%). b Includes sole person and couple-only households who reported parental responsibility for, but not living with, children. c Includes respondents who were students, unemployed, employed in home duties, volunteers and others. *P < 0.05, ***P < CI: confidence interval; SEIFA: Socio-Economic Indexes for Areas. more likely to report ARHC compared to non-drinkers, while middle- and higher-income respondents were less likely to report ARHC than respondents on low incomes. None of the other socio-demographic variables were associated significantly with ARHC. The multivariate model indicates that once family type and respondent drinking status were adjusted for, income was no longer associated significantly with ARHC. DISCUSSION We estimate that 22% of Australian families have one or more children who have experienced ARHC as a result of other s drinking in some way. Almost 12% of families reported one or more of the specific harms asked about. Here, the majority were reported to have been affected by family members, including 61% by immediate family members and 12% by other relatives. One-quarter were reported to have been affected by the drinking of family friends, friends, neighbours, teachers, coaches, religious leaders or others. That children in this survey who were reported to have experienced any specific harm most commonly experienced more than one type of harm, and on average experienced 12.4 incidents (most commonly three) in the past 12 months, suggest that these incidents are not simply one-off problems. The most commonly reported form of specified ARHC identified in this study was verbal abuse. Verbal abuse can be cast as emotional abuse [17,20]; however, emotional abuse is usually a chronic pattern of behaviours such as yelling, denigrating, humiliating, isolating or terrorizing the child to such an extent that the behaviour is likely

6 Alcohol-related harm to children 1087 to cause the child significant emotional harm [20]. Nevertheless, it is notable that incidents of verbal abuse because of others drinking have been shown in this analysis to be prevalent across a range of relationships. More detailed data on how children are affected by verbal abuse because of others drinking is needed. Reports of potentially more severe outcomes because of others drinking were less common. For example, just 3% of families with children were reportedly exposed to domestic violence because of others drinking in the last year. Similarly, Manning et al. [6] found that 2% of children in Scotland witnessed a violent incident to an adult partner after the perpetrator had been drinking. However, these small percentages translate into a large number of affected families: 3% equates to families Australia-wide. These percentages are much higher than the number of families identified within Australian child protection systems. For example, in Victoria the rate of substantiated cases involving alcohol in 2005, for all abuse categories together, was 0.7% of the child population [11]. Our findings regarding exposure to serious violence in the home, because of others drinking, are similar to those on alcohol-related domestic violence from general population personal safety surveys conducted in Australia [21]. Our results suggest that the drinking of parents and other adults involved with the children have consequences for children. Child protection data suggest that in the majority of neglect cases biological parents are held responsible [22,23], and often mothers more than fathers, because they are more often the primary carers [24,25], although in physical abuse cases personal safety surveys suggest that in Australia fathers are more likely to be responsible [26]. In child protection it is acknowledged that the drinking of either the maltreating or the protective parent could contribute to increased maltreatment of children. If a maltreating parent becomes intoxicated they may be less inhibited and more likely to cause harm by saying or doing things they would not otherwise. If the protective parent drinks heavily they may be less able to protect the child from a partner s or other person s actions [26]. In this study, the respondent s drinking pattern was associated significantly with whether children were more likely to be affected negatively by others, at least in part reflecting that heavier drinkers often associate with other heavy drinkers. The study does not address specifically whether these drinkers are less able to protect their children from others drinking if they are also drinking, an issue which future studies should investigate. In the present general-population survey frame, ARHC was more common in the low-income group, but this difference disappeared when family type and respondent drinking status were taken into account. This contrasts with the picture from the Child Protection case data, which predominantly identify victims who are often financially disadvantaged. Because the majority of parenting occurs in private spaces that state agencies have little ability to observe or control, those families more exposed to a range of state agencies may be more likely to be notified to state welfare organizations [12,27] indeed, the majority of child abuse cases in Victoria and Australia are identified in socially and economically disadvantaged and disempowered families. How much this is because the more severe forms of child abuse and neglect occur less often in more socioeconomically advantaged homes and how much because these homes are less likely to come under state surveillance should be studied further. Study limitations The response rate of the survey, although comparable with many current Australian surveys [28 30], is suboptimal and a common concern [31], although findings from different surveys with lower and higher response rates have been similar across a number of different health measures [32]. This places limits on the confidence with which extrapolations to the general population can be made. However, as this study was under-representative of young men and young women (18 34 years) who drink more and report higher rates of harm from others drinking than older men and women, it is likely that this study under-represents the alcoholrelated harms experienced by children within households with younger parents. Furthermore, as this study did not include data on whether or not children have been affected by the respondents own drinking, this may contribute further to possible under-reporting. Future studies might consider including own or others drinking in the questionnaire, although careful wording and rapport will be necessary to ensure that data are not compromised by respondents who do not self-report or recognize the harms that their own drinking causes. The difference identified in rates of reported harms in a survey that used both methods would be illuminating. There may be some response bias when respondents attribute drinking behaviours and consequences to someone else in relation to children in their own family, although the direction and extent of bias is difficult to estimate. For example, partners might either minimize or exaggerate the effects of domestic violence on their children. Gilbert et al. [10] reported comparable responses from children and parents. The wording of the questions wherein one or more children were affected means that there cannot be exact calculation of prevalence figures of alcohol-related harms experienced by children. Conservative estimates were used in this paper.

7 1088 Anne-Marie Laslett et al. Alcohol-related harm to children has been broadly defined in this paper, and may be only loosely connected with substantiated child abuse and neglect; and, as stated, the level of severity and alcohol involvement respondents report is subjective. The numbers are small to examine patterns in more severe or specific types of child abuse, and it would be desirable to repeat the study in larger future samples. These findings form a sound base for calculation of power estimates for larger samples. Future qualitative studies should ask open-ended questions about the nature of this harm. Despite these difficulties, surveys such as this provide a unique window on child maltreatment in the general population. The alternatives offer different perspectives: self-reporting by perpetrators provides a second viewpoint but can be fraught, and questioning children below the age of consent, let alone very young children, is often impractical and raises ethical issues. Surveys that ask about life-time experiences (e.g. Cawson [33]) are helpful, but subject to recall biases. Child protection system data may be used to triangulate findings, but are subject to their own operational limitations and by their nature are not representative of the general population. The study s strength is that it provides a critical first estimate of ARHC in the general population. CONCLUSION ARHC was reported in 22% of families in the year under study. Specific alcohol-related harm to children was reported in 12% of families. This is concerning, although arguably not surprising, given that alcohol s effects on behaviour are well known and levels of heavy episodic drinking in Australia are high. Children in families from a wide range of social backgrounds experienced harm because of others drinking, suggesting that alcohol policies with wide application may be indicated, although a special focus on single-carer households may be warranted. Declarations of interest None. Acknowledgements This paper is based on data from the national survey of the range and magnitude of alcohol s harm to others (AHTO) survey commissioned by the Foundation for Alcohol Research and Education (FARE). A.M.L. is supported by a postgraduate doctoral scholarship from The Sidney Myer Fund and the Australian Rechabite Foundation; P.D. is supported by an ARC Future Fellowship (FT ). We would like to acknowledge the participants who completed the questionnaire. References 1. Bittman M., Wajcman J. The rush hour: the character of leisure time and gender equity. Soc Forces 2000; 79: Australian Institute of Health and Welfare (AIHW). National Drug Strategy Household Survey 2007: First Results. Drug Statistics Series no. 20. Cat. no. PHE 98. Canberra: AIHW; Dawe S., Frye S., Best D., Moss D., Atkinson J., Evans C. et al. Drug Use in the Family: Impacts and Implications for Children. Canberra: Australian National Council on Drugs; NSW Department of Community Services. Research to practice notes: parental alcohol misuse and the impact on children. New South Wales Department of Government Services Cook E. Is it six o clock yet? Guardian, 5 November Manning V., Best D. W., Faulkner N., Titherington E. New estimates of the number of children living with substance misusing parents: results from UK national household surveys. BMC Public Health 2009; 9: Rossow I. Suicide, violence and child abuse: a review of the impact of alcohol consumption on social problems. Contemp Drug Probl 2000; 27: Felitti M. D. Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults: the adverse childhood experiences (ACE) study. Am J Prev Med 1998; 14: Straus M. A., Stewart J. H. Corporal punishment by American parents: national data on prevalence, chronicity, severity, and duration, in relation to child and family characteristics. Clin Child Fam Psychol Rev 1999; 2: Gilbert R., Widom C. S., Browne K., Fergusson D., Webb E., Janson S. Burden and consequences of child maltreatment in high-income countries. Lancet 2009; 373: Laslett A.-M., Catalano P., Chikritzhs T., Dale C., Doran C., Ferris J. et al. The Range and Magnitude of Alcohol s Harm to Others. Canberra: Alcohol Education and Rehabilitation Foundation; Straus M., Hamby S., Finkelhor D., Moore D. W., Runyan D. Identification of child maltreatment with the parent child conflict tactics scales development and psychometric data for a national sample of American parents. Child Abuse Negl 1998; 22: Bardi M., Borgognini-Tarli S. M. A survey on parent child conflict resolution: intrafamily violence in Italy. Child Abuse Negl 2001; 25: Laslett A., Room R., Ferris J., Wilkinson C., Livingston M., Mugavin J. Surveying the range and magnitude of alcohol s harm to others in Australia. Addiction 2011; 106: American Association of Public Opinion Research (AAPOR). Standard Definitions: Final Disposition of Case Codes and Outcome Rates for Surveys. Ann Arbor, MI: AAPOR; Wilkinson C., Laslett A.-M., Ferris J., Livingston M., Mugavin J., Room R. The range and magnitude of alcohol s harm to others: study design, data collection procedures and measurement. Fitzroy: AER Centre for Alcohol Policy Research, Turning Point Alcohol and Drug Centre; Australian Institute of Health and Welfare (AIHW). Child Protection Australia CWS 31. Canberra: AIHW; Australian Bureau of Statistics. Information Paper: An Introduction to Socio-Economic Indexes for Areas (SEIFA). Cat.no Canberra: Australian Bureau of Statistics; 2006.

8 Alcohol-related harm to children StataCorp. STATA Statistical Software: Release 11. College Station, TX: StataCorp LP; Victorian Department of Human Services. Abuse and harm legal and practice definitions. Date of Advice: 23 April Advice no: 1008 Protecting Victoria s Children, Child Protection Practice Manual is an on line application available via the Department of Human Services intranet for Department of Human Services Child Protection practitioners and managers. Melbourne, Australia: Victorian Department of Human Services; Mouzos J., Makkai T. Women s experiences of male violence: findings from the Australian component of the international violence against women survey. Canberra: Australian Institute of Criminology; Australian Institute of Health and Welfare. Child Protection Australia Canberra: Australian Institute of Health and Welfare; Sedlak A., Mettenburg J., Basena M., Petta I., McPherson K., Green A., Li S. Fourth National Incidence Study of Child Abuse and Neglect (NIS-4): report to Congress. Rockville, MD: Westat, Inc., for the United States Department of Health and Human Services (DHHS), Administration for Children and Families (ACF), Office of Planning, Research and Evaluation (OPRE) and the Children s Bureau; May-Chahal C., Cawson P. Measuring child maltreatment in the United Kingdom: a study of the prevalence of child abuse and neglect. Child Abuse Negl 2005; 29: Lamont A. Who abuses children? National Child Protection Clearinghouse: NCPC Issues. 2011; February: Australian Bureau of Statistics (ABS). Personal Safety Survey, Australia, 2005 (reissue). Cat. no Canberra: ABS; Baachi C. L. Women, Policy and Politics: The Construction of Policy Problems. London: Sage Publications; Brown W. J., Bryson L., Byles J. E., Dobson A. J., Lee C., Mishra G. et al. Women s Health Australia: recruitment for a national longitudinal cohort study. Women Health 1999; 28: Roxburgh A., Hall W. D., Degenhardt L., McLaren J., Black E., Copeland J. et al. The epidemiology of cannabis use and cannabis-related harm in Australia Addiction 2010; 105: Purdie D. M., Dunne M. P., Boyle F. M., Cook M. D., Najman J. M. Health and demographic characteristics of respondents in an Australian national sexuality survey: comparison with population norms [Theory and methods]. J Epidemiol Commun Health 2002; 56: Curtin R., Presser S., Singer E. R. Changes in telephone survey nonresponse over the past quarter century. Public Opin Q 2005; 69: Serraglio A., Carson N., Ansari Z. Comparison of health estimates between Victorian Population Health Surveys and National Health Surveys. Aust NZ J Public Health 2003; 27: Cawson P., Wattam C., Brooker S., Kelly G. Child maltreatment in the United Kingdom: a study of the prevalence of abuse and neglect. London: National Society for Prevention of Cruelty to Children; 2000.