bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
BMJ Paediatrics Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.
When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.
The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.
BMJ Paediatrics Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay- per-view fees (http://bmjpaedsopen.bmj.com).
If you have any questions on BMJ Paediatrics Open’s open peer review process please email [email protected] http://bmjpaedsopen.bmj.com/ on September 28, 2021 by guest. Protected copyright. BMJ Paediatrics Open bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
Confidential: For Review Only
Child homicide perpetrators worldwide – a systematic review
Journal: BMJ Paediatrics Open
Manuscript ID bmjpo-2017-000112
Article Type: Original article
Date Submitted by the Author: 29-May-2017
Complete List of Authors: Stöckl, Heidi; Global Health and Developmen Dekel, Bianca; South African Medical Research Council, Cape Town Gehring, Alison; London School of Hygiene and Tropical Medicine Faculty of Public Health and Policy, Global Health and Development Watts, Charlotte; London School of Hygiene and Tropical Medicine Faculty of Public Health and Policy, Global Health and Development Abrahams, Naeemah; South African Medical Research Council, Cape Town
Keywords: Child Abuse, Forensic Medicine, Injury Prevention, Mortality, Statistics http://bmjpaedsopen.bmj.com/
on September 28, 2021 by guest. Protected copyright.
https://mc.manuscriptcentral.com/bmjpo Page 1 of 27 BMJ Paediatrics Open bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 Child homicide perpetrators worldwide – a systematic review 3 Heidi Stöckl1*, Bianca Dekel2, Alison Morris-Gehring1, Charlotte Watts1 and Naeemah Abrahams2 4 5 1 London School of Hygiene and Tropical Medicine, London, UK 6 7 2 Medical Research Council, Cape Town, South Africa 8 Confidential: For Review Only 9 10 * 11 Corresponding author: Heidi Stöckl, 12 13 Department of Global Health and Development 14 London School of Hygiene and Tropical Medicine 15 16 15-17 Tavistock Place 17 18 London WC1H 9SE, UK 19 20 Phone: 0044 2079272506 21 22 Email: [email protected] 23
24 25 Word count manuscript: 2381 words 26 27 28 29 30
31 http://bmjpaedsopen.bmj.com/ 32 33 34 35 36 37 38 39 40 41 42 on September 28, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjpo BMJ Paediatrics Open Page 2 of 27 bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 Abstract 3 4 Objective: 5 6 7 This study aims to describe child homicide perpetrators and estimate their global and regional 8 Confidential: For Review Only 9 proportion to inform prevention strategies to reduce child homicide mortality worldwide. 10 11 Design: 12 13 14 A systematic review of 18 databases led to the inclusion of 126 studies after double screening. All 15 16 included studies reported a number or proportion of child homicides perpetrators. 169 countries and 17 18 homicide experts were surveyed in addition. The median proportion for each perpetrator category 19 20 was calculated by region and overall, and by age groups and sex. 21 22 23 Results: 24 25 Data were obtained for 44 countries. Overall, parents committed 56.47% (IQR 23.72-69.57) of child 26 27 homicides, 58.35% (0.00-66.67) of female and 46.84% (14.13-63.82) of male child homicides. 28 29 Acquaintances committed 12.55% (5.88-31.27) of child homicides. Almost a tenth (9.18% (IQR 0.00- 30
31 http://bmjpaedsopen.bmj.com/ 21.85) of child homicides had missing information on the perpetrator. The largest proportion of 32 33 34 parental homicides of children was found in High Income Countries (64.24%; 44.74-71.83) and East 35 36 Asia and Pacific Region (61.66%; 46.67-78.57). Parents committed the majority (77.82% (61.54- 37 38 100.00)) of homicides of children under the age of one. For adolescents, acquaintances were the 39 40 main group of homicide perpetrators (36.86%, 6.62-51.75). There is a notable lack of studies from 41 on September 28, 2021 by guest. Protected copyright. 42 low and middle income countries and children above the age of one. 43 44 45 Conclusion: 46 47 Contrary to common perceptions, children face the highest risk of homicide by parents and someone 48 49 they know. Increased investment into the compilation of routine data on child homicide, and the 50 51 perpetrators of this homicide is imperative for understanding and ultimately reducing child homicide 52 53 54 mortality worldwide. 55 56 Prospero registration number: CRD42015030125 57 58 59 60 https://mc.manuscriptcentral.com/bmjpo Page 3 of 27 BMJ Paediatrics Open bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 3 Key messages 4 5 6 What is known about the subject 7 8 Confidential:• Child homicide is the most extreme For form of violence Review against children and a Onlytragic event with 9 10 serious effects on families and the community. 11 12 13 • Approximately 95000 child are murdered each year globally (UNICEF 2014), and being 14 15 murdered in childhood is strongly associated with age, gender and geography. 16 17 • In many settings, homicide data is incomplete, and information on the victim-perpetrator- 18 19 relationship is lacking, seriously limiting the evidence base needed for devising effective 20 21 prevention strategies. 22 23 24 25 What this study adds 26 27 • Contrary to common perceptions, children face the highest risk of homicide by parents. 28 29 30 • The limited availability of data on victim perpetrator relationships of child homicides in many
31 http://bmjpaedsopen.bmj.com/ 32 regions, even within countries with well-established homicide monitoring systems is 33 34 concerning. 35 36 • Increased investment into the compilation of routine data, and the perpetrators of child 37 38 homicide is imperative for understanding and ultimately reducing child homicide mortality 39 40 worldwide. 41 42 on September 28, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjpo BMJ Paediatrics Open Page 4 of 27 bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 Introduction 3 4 Child homicide is the most extreme form of violence against children and a tragic event with serious 5 6 effects on families and the community. Approximately 95000 child are murdered each year globally1, 7 8 Confidential:and the risk of being murdered in childhood For is strongly associatedReview with age, gender Only and geography. 9 10 The distribution of child homicides by age is U-shaped, and skewed towards adolescence. Children 11 12 aged 15 to 19 constitute 57% of the global child homicides, followed by children under the age of five 13
14 1 15 (20%) . The majority of child homicide victims (90%) live in low and middle income countries, with 16 1 17 the highest child homicide rates found in Latin America . The risk of homicide is particularly high for 18 19 boys, who account for 70 per cent of all child homicides, and for whom homicide rates dramatically 20 21 increase in late adolescence1. Evidence beyond the mere prevalence of child homicide, the sex and 22 23 age of the victims however remains sparse. 24 25 In general, homicides are likely to be underreported, particularly in countries with poor monitoring 26 27 28 systems and suffering from conflict. In many settings, homicide data is often incomplete, and 29 30 information on the victim-perpetrator-relationship is lacking, seriously limiting the evidence base
31 http://bmjpaedsopen.bmj.com/ 32 needed for devising effective prevention strategies. Existing studies on child homicide perpetrators 33 34 often focus on a particular age group or type of homicide, for example, filicide or neonaticides; are 35 36 restricted to high income countries; or rely on small sample sizes or outdated data2-5. This systematic 37 38 review presents a synthesis of existing global data to describe child homicide perpetrators and to 39 40 41 estimate their global and regional proportion. 42 on September 28, 2021 by guest. Protected copyright. 43
44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjpo Page 5 of 27 BMJ Paediatrics Open bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 Methods 3 4 Our systematic review was conducted in accordance with PRISMA guidelines 6 and registered at 5 6 7 PROSPERO (Registration number: CRD42015030125) on the 9th December 2015. To obtain 8 Confidential: For Review Only 9 representative data on the victim-offender relationship of child homicides, we searched the following 10 11 databases and electronic resources from first record to the 25th April 2017: Medline, Global Health, 12 13 Embase, PsycINFO, Social Policy, Popline, Pubmed, Web of Science, LILACS, Medcarib, ADOLEC, 14 15 Cumulative Index to Nursing and Allied Health Literature [CINAHL], Biosis Citation index, KCI-Korean 16 17 Journal Database, SciELO citation Index, Western Pacific Region Index Medicus (WPRIM), Index 18 19 20 Medicus for the WHO Eastern Mediterranean Region (IMEMR) and International Bibliography of 21 22 Social Sciences IBSS. Controlled vocabulary terms specific to each database were used. For each 23 24 database, terms included those designed to capture “child homicide” and, where possible, terms to 25 26 capture perpetrators and to exclude non-relevant article types, such as commentaries. The search 27 28 strategy is displayed in Box 1. 29 30
31 Studies were eligible for inclusion if they stated a number or proportion of children murdered by http://bmjpaedsopen.bmj.com/ 32 33 distinctly stated perpetrator(s). The definition of homicide followed the definitions used in the 34 35 individual papers or official statistics. The sample could be based on a country, province, or town and 36 37 38 be derived from national databases, national representative studies, police, court, mortuary, or 39 40 prison data. Studies were excluded if they only reported data collected from newspaper reports and 41 42 did not differentiate between attempted and completed homicides. Estimates that combined child on September 28, 2021 by guest. Protected copyright. 43 44 homicides with adult homicides were excluded. The citations of included articles were also searched. 45 46 Two authors (HS and AMG) screened the 6096 abstracts and the resulting 563 full texts 47 48 independently and resolved any disagreements by discussion. 123 studies were finally included (See 49 50 51 Flow Chart in Figure 1). HS extracted the data and BD verified each data extraction point for 52 53 accuracy. 54 55 56 In addition to the systematic review, we surveyed country statistics offices, ministries of justice, 57 58 home offices, or police headquarters of 169 WHO-listed countries to further identify country-level 59 60 https://mc.manuscriptcentral.com/bmjpo BMJ Paediatrics Open Page 6 of 27 bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 data for child homicide perpetrators and victim-perpetrator relationships. We received responses 3 4 from 90 countries. Additionally, we made contact with homicide research experts to ask for 5 6 unpublished data on child homicide perpetrators in their studies. In total, this lead to an additional 7 8 Confidential:inclusion of data for 24 countries. For Review Only 9 10 11 The analysis consisted of two main steps, the selection of one estimate per country year and the 12 13 calculation of the proportion of child homicides committed by different perpetrator categories, for 14 15 different age groups and by gender separately. 16 17 18 As several countries had more than one estimate available, an algorithm was used to develop a single 19 20 estimate per country for each of the types of homicide and age group or gender analysed, to avoid 21 22 potential double counting, and to ensure to use the best quality estimates available. Where possible, 23 24 25 comparable, non-overlapping data was combined. Otherwise we chose an estimate according to the 26 27 following hierarchical order: (1) nationally representative estimates were preferred to provincial 28 29 estimates, which in turn were prioritised over estimates representing a single town or mortuary, (2) 30
31 we chose estimates with more detailed information on the victim-perpetrator relationship, estimates http://bmjpaedsopen.bmj.com/ 32 33 that captured children of the age group, (3) studies that were more recent and (4) estimates directly 34 35 from statistics offices was preferred and (5) we preferred estimates that covered more years to 36 37 38 estimates based on a few years. 39 40 Approximately two thirds of the studies only reported combined estimates for up to 20 years on the 41 42 on September 28, 2021 by guest. Protected copyright. 43 proportions of child homicide perpetrators, starting with 1970 onwards. The included estimates 44 45 yielded 51 different victim-perpetrator categories. To facilitate comparisons across countries, we 46 47 grouped the perpetrators of child homicide into distinct, broad categories: parents, other family 48 49 members, acquaintances (meaning someone they knew), strangers (capturing people they had no 50 51 prior contact with), and unknown (not documented perpetrators). As countries used different age 52 53 categories, countries that only considered children up to age 12 are also included, if no age category 54 55 56 up to 18 was available. 57 58 59 60 https://mc.manuscriptcentral.com/bmjpo Page 7 of 27 BMJ Paediatrics Open bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 To calculate the percentage of child homicide perpetrators by country, we divided the total number 3 4 of homicides in each perpetrator category by the total number of child homicides in that country. To 5 6 establish the global and regional estimates, the relevant available national percentages were added 7 8 Confidential:and divided by the number of countries with For available data. Review For the regional estimates, Only we grouped 9 10 countries as per the six WHO regions (Africa, Americas, Eastern Mediterranean, Low and middle 11 12 income Europe, Southeast Asia and Western Pacific). A seventh region was made of the high income 13 14 15 countries from all regions. 16 17 After conducting the primary descriptive analysis, secondary analyses investigated parental 18 19 20 homicides separately, to explore how many children murdered by parents were murdered by their 21 22 mother or father. Additional analyses investigated the child homicide perpetrators of children below 23 24 the age of one, neonaticides, adolescents and by gender. 25 26 27 Data were analysed with Stata version 12. Traditional meta-analysis techniques could not be used 28 29 because nearly all studies were representative of the whole population and not restricted to 30
31 population samples. Details on individual studies are reported in the supplement. Because the http://bmjpaedsopen.bmj.com/ 32 33 percentages were skewed, we report the median percentage and the IQR. 34 35 36 37 38 39 40 41 42 on September 28, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjpo BMJ Paediatrics Open Page 8 of 27 bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 Results 3 4 Overall, 9431 abstracts and the 565 full texts were screened and 126 studies were included into the 5 6 7 study (see Flow chart). In addition to information received from statistical offices and experts, 1031 8 Confidential: For Review Only 9 estimates on perpetrators of child homicides across 44 countries informed our estimates, 22 10 11 countries in the high income countries, six countries in the Western Pacific region, five in the low and 12 13 middle income European Region, four countries in the African Region and the Americas and three in 14 15 the Mediterranean Region. No data was found for the South Asia region. Information on perpetrators 16 17 of all child homicides of children under 18 was available for 35 countries, compiling data for 73242 18 19 20 homicides. Data on parental homicides, distinguishing between mothers and fathers, was available 21 22 for 32 countries, for children under one including neonaticides for 20 countries and for adolescents 23 24 for 10 countries only. Few data was available for the Americas, the Eastern Mediterranean and the 25 26 African Region. Most data outside high-income countries came from cities or individual mortuaries 27 28 and were therefore based on small numbers of child homicides. 29 30
31 Across all 35 countries with data on homicide perpetrators for all children under the age of 18, we http://bmjpaedsopen.bmj.com/ 32 33 found that more than half (Median percentage 56.47, IQR 23.71-69.57) were murdered by a parent, 34 35 2.96 percent (IQR 0.00-7.14) by another family member, 12.55 percent (5.88-31.27) by an 36 37 38 acquaintance, e.g. a neighbor or friends, 2.14 percent (0.00-11.14) by a stranger and perpetrator 39 40 remained unknown for 9.18 percent (0.00-21.85). The percentage of girls murdered by their parents 41 42 was 58.35 (IQR 0.00-66.67) versus 46.84 for boys (IQR 14.13-63.82), while the median percentage for on September 28, 2021 by guest. Protected copyright. 43 44 boys murdered by strangers was 9.22 (IQR 4.22-38.58) versus 6.48 (IQR 3.58-6.67) for girls. The total 45 46 number of child homicides reported and the prevalence by perpetrator categories across region is 47 48 reported in Table 1. The largest proportion of parental homicides among all child homicides was 49 50 51 found in high-income countries (64.24%; 44.74-71.83) and the East Asia and Pacific Region (61.66%; 52 53 46.67-78.57). 54 55 56 Data from 33 countries distinguishing the perpetrators of parental homicides of children under the 57 58 age of 18 showed that mothers committed just over half of all parental homicides (Median 54.7%, 59 60 https://mc.manuscriptcentral.com/bmjpo Page 9 of 27 BMJ Paediatrics Open bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 IQR 36.7-68.8); in high income countries the median percentage was 44.44 percent (IQR 36.66- 3 4 66.67), in the East Asia and Pacific region 64.59 percent (IQR 59.01 – 69.33), in the Americas 15.36 5 6 percent (IQR 13.33-17.39), in Africa 88.60 percent (IQR 71.11-100.00), in low and middle income 7 8 Confidential:Europe 60.42 percent (IQR 45.83-75.00) and For in the Mediterranean Review Region 7.38 percent Only (IQR 0.00- 9 10 14.77). Only 12 countries had available data on how many parental homicides were committed by 11 12 stepparents, with 7.15 percent (IQR: 5.73-14.5) of parental homicides committed by stepparents. Out 13 14 15 of them, nine studies reported a break-down, with the median percentage of stepmothers 16 17 committing homicides among parental homicides being 1.03 percent (IQR: 065-3.33) and for 18 19 stepfathers 7.4 percent (IQR: 3.13-17.39). 20 21 22 Only 14 countries have data on homicide perpetrators of children under one year of age, excluding 23 24 studies investigating neonaticides only. These show that parents are the most common perpetrators 25 26 of homicides of children under 1 year (77.82, IQR: 61.54-100), followed by unknown perpetrators 27 28 with 7.52 percent (IQR: 0.00-24.04). The 12 countries with detailed data on parent as perpetrators 29 30 showed mothers’ commit the majority of parental homicides of children under one (71.72% (IQR:
31 http://bmjpaedsopen.bmj.com/ 32 33 50.18-75.66). Seven studies investigated perpetrators of neonaticides (defined as the murder of an 34 35 infant within 24 hours of giving birth with the exception of a study in France and Denmark which 36 37 extended it to one months after birth), without considering the mother as the default perpetrator. 38 39 They found that parents committed 100% of neonaticides (IQR 62.00-100.00). If studies only 40 41 considering parents were examined 13 countries had data and the vast majority of neonaticides are 42 on September 28, 2021 by guest. Protected copyright. 43 committed by mothers (100%, IQR 92.9-100). Fathers committed very few neonaticides (0.00%, IQR: 44 45 46 0.00-6.67). 47 48 Data on homicide perpetrators of adolescents, aged 10 to 17 years, was only available for 10 49 50 51 countries, with only three presenting data for the whole age group of 10 to 18. The most common 52 53 perpetrators among adolescent homicides were acquaintances (36.9%, IQR 6.62-51.75), followed by 54 55 family members (17.48%, IQR 4.27-33.3) and strangers (6.62%, IQR 0.00-25.00). 9.63% (IQR 1.04- 56 57 32.84) of perpetrators of adolescents were unknown. This trend was supported by one study each in 58 59 60 https://mc.manuscriptcentral.com/bmjpo BMJ Paediatrics Open Page 10 of 27 bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 Latin America or Africa, the two areas with the highest prevalence of male adolescent homicides 3 4 worldwide. 5 6 7 8 Confidential: For Review Only 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
31 http://bmjpaedsopen.bmj.com/ 32 33 34 35 36 37 38 39 40 41 42 on September 28, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjpo Page 11 of 27 BMJ Paediatrics Open bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 Discussion 3 4 Contrary to common perceptions, children face the highest risk of homicide by parents. The limited 5 6 7 availability of data on victim perpetrator relationships of child homicides in many regions, even 8 Confidential: For Review Only 9 within countries with well-established homicide monitoring systems is concerning. The poor state of 10 11 data is of particular concern in regions that have among the highest child homicide rates, and the 12 13 highest rate of homicides among adolescents. Where countries collect data on the perpetrator, often 14 15 this is not disaggregated by sex and age. This is an important omission, as studies suggest that child 16 17 abuse related homicides have a distinct gender pattern7 8. Lack of disaggregation by age and varying 18 19 20 age ranges also prevented a meaningful analysis of perpetrators of murdered children aged between 21 22 one to ten years. 23 24 There are many reasons for the identified data gaps. Homicide data is generally collected by the 25 26 police, mortuaries or court statistics and there is a lack of collaboration between those institutions. A 27 28 separate homicide analysis for homicides involving victims under the age of 18 is rarely conducted or 29 30 published in most official statistical reports. With increased investment into administrative data
31 http://bmjpaedsopen.bmj.com/ 32 33 systems, including systems to facilitate improved linkage of different data sources, it would be 34 35 feasible to address these issues. Internationally agreed standards of documentation of age and 36 37 victim-perpetrator categories would further facilitate cross-country comparisons. Future studies on 38 39 child homicide should also not be restricted by default to a certain type of victim-perpetrator 40 41 relationship, e.g. filicide or homicides by mothers only as our systematic review found studies 42 on September 28, 2021 by guest. Protected copyright. 43 showing that other family members might also be involved in neonaticides. 44 45 46 The regional differences found in this study might represent real differences in patterns of child 47 48 homicide 9 or be a product of differences in the existence, completeness and quality of data for 49 50 51 homicides among countries and regions. Data on the victim-perpetrator relationship is often only 52 53 available if the information is reported. Some murders will never be resolved, which is strongly 54 55 correlated with the quality of police investigations. In the case of child homicides, in particular, it is 56 57 estimated that a significant number of child homicides are not recognised as homicides, and 58 59 60 https://mc.manuscriptcentral.com/bmjpo BMJ Paediatrics Open Page 12 of 27 bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 therefore do not appear in official statistics. Studies on neonaticides, in particular, show that nearly 3 4 all of them follow concealed pregnancies and lone births 10 and that there often are difficulties in 5 6 distinguishing death due to natural disorders from non-natural causes, especially in cases labelled 7 8 Confidential:sudden infant death syndrome (SIDS) 11. There For are also otherReview limitations to this study. Only Averaging the 9 10 proportion of perpetrators of child homicides across years to generate one estimate by country could 11 12 have led to an overestimation or underestimation of certain types of perpetrator 10 12. Vast 13 14 15 differences in the definitions of perpetrator categories only allowed crude comparisons across 16 17 countries. The restricted availability also meant that the analysis could not be done for every age 18 19 group or by gender. 20 21 22 Our study is the first to provide global evidence on the perpetrators of child homicide. Our findings 23 24 highlight the gaps in data on child homicide that need to be addressed. Contrary to common 25 26 perceptions that risk lies outside the household, the findings underscore that parents commit a high 27 28 proportion of child homicides. The reasons for each are likely to be complex, and relate to broader 29 30 family, community and cultural contexts 13 14. In all contexts, the murder of children needs to be
31 http://bmjpaedsopen.bmj.com/ 32 33 better documented, prevented and addressed. 34 35 Acknowledgement 36 37 38 None of the authors has a conflict of interest. This paper was prepared under the Know Violence in 39 40 Childhood: Global Learning Initiative (http://www.knowviolenceinchildhood.org/). The authors 41 42 on September 28, 2021 by guest. Protected copyright. 43 acknowledge the support and internal review provided by the Initiative and its funders. 44 45 Author’s contributions 46 47 48 HS and NA conceived the study. HS, BD and AM searched the literature and screened the abstracts 49 50 and full-texts. HS and BD extracted the data. HS, NA and CW wrote the final draft. All authors 51 52 approved the final manuscript. 53 54 55 Figure legend 56 57 58 Figure 1: Flow chart of the systematic review 59 60 https://mc.manuscriptcentral.com/bmjpo Page 13 of 27 BMJ Paediatrics Open bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 Box 1: Detailed search terms 3 4 5 (“perpetrators” or “perpetrator” or “offender” or “offenders” or “aggressor” or “aggressors” or 6 “father” or “fathers” or “mother” or “mothers” or “parent” or “parents” or “brother” or 7 “brothers” or “sister” or “sisters” or “aunt” or “aunts” or “uncle” or “uncles” or “grandmother” or 8 Confidential:“grandmothers” or “grandfather” or “grandfathers” For or “grandparent”Review or “grandparents” Only or 9 “stranger” or “strangers” or “acquaintance” or “acquaintances” or “friends” or “friend” or 10 “partner” or “partners” or “victim-offender relationship” or “victim offender relationship” or 11 “victim-offender relationships” or “victim offender relationships” or “ex-partners” or “ex-partner” 12 or “husband” or “husbands” or “wife” or “wives” or “couple” or “couples” or “boyfriend” or 13 “girlfriend” or “spouse” or “spouses” or “lover” or “spousal” or “boyfriends” or “girlfriends” or 14 “relative” or “relatives” or “family member” or “family members” or “maternal” or “parental”) 15 AND (“siblicide” or “neonaticide” or “Infanticide” or “fratricide” or “victim-perpetrator 16 relationship” or “victim perpetrator relationship” or “sororicide” or “family homicide” or 17 “familicide” or “filicide” or “fratricides” or “family murder” or “family homicide suicide” or 18 19 “familicy” or “family suicide” or “fatal child abuse” or “infant baby dumping” or “child abuse 20 murders” or “child abuse fatalities” or “child abuse fatality” or “parental homicide” or “parental 21 homicides” or “parents who kill” or “mothers who kill” or “fathers who kill” or “grandparents who 22 kill” or “uncles who kill” or “siblings who kill” or “aunts who kill” or “brothers who kill” or “sisters 23 who kill” or “family murder” or “family killing” or “adolescent homicide” or “adolescence 24 homicide” or “adolescent murder” or “adolescence murder” or “adolescent killing” or 25 “adolescence killing” or “child murder” or “child homicide” or “child killing” or “children murder” 26 or “children killing” or “children homicide” or “sibling murder” or “sibling killing” or “sibling 27 homicide” or “infant murder” or “infant killing” or “infant homicide” or “toddler killing” or “toddler 28 murder” or “toddler homicide” or “young adult murder” or “young adult killing” or “young adult 29 homicide” or “sister murder” or “sister killing” or “sister homicide” or “brother murder” or 30 “brother homicide” or “brother killing” or “boy murder” or “boy killing” or “boy homicide” or “girl
31 http://bmjpaedsopen.bmj.com/ 32 murder” or “girl killing” or “girl homicide” or “child abandonment” or “infant abandonment” or 33 “toddler abandonment” or “girl abandonment” or “baby abandonment” or "homicide death rates 34 in childhood" or “unnatural sudden infant death” or “unnatural sudden infant deaths” or 35 “childhood homicide” or “childhood homicides” or “child killers” or “fatal maltreatment”) 36 37 38 39 40 41 42 on September 28, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjpo BMJ Paediatrics Open Page 14 of 27 bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 Table 1: Distribution of child homicide perpetrator proportions by perpetrator and region 3 4 Total homicides Parents Family no parents Acquaintances Stranger Unknown 5 All children <18 75496 59.73 2.96 12.55 2.14 9.18 (n=35) (25.00-69.57) (0.00-7.14) (5.88-31.27) (0.00-11.14) (0.00-21.85) 6 High Income 73242 64.24 2.39 10.32 2.03 8.68 7 countries (n=19) (56.20-71.83) (0.00-6.95) (5.36-13.69) (0.00-10.57) (0.00-17.65) 8 Confidential:East Asia and Pacific 559 61.66 For5.93 Review14.29 8.29 Only 1.58 9 (n=5) (46.67-78.57) (3.28-7.14) (10.00-26.94) (0.00-13.12) (0.00-11.92) 10 Americas (n=3) 121 27.63 21.71 44.41 6.25 31.11 11 (25.00-30.26) (0.00-43.42) (26.32-62.50) (0.00-12.50) (0.00-43.42) LMIC Europe (n=3) 71 16.67 0.00 14.29 8.70 15.22 12 (14.29-69.57) (0.00-4.35) (0.00-44.44) (0.00-27.78) (11.11-71.43) 13 Africa 1042 23.57 3.89 19.06 8.17 11.49 14 (n=3) (8.33-38.80) (0.00-7.76) (0.00-38.11) (3.83-12.50) (0.00-79.17) 15 Mediterranean 461 17.75 25.55 69.67 0.44 4.38 16 (n=2) (10.50-25.00) (1.09-50.00) (64.33-75.00) (0.00-0.88) (0.00-8.75) 17 <1 (n=13) 5529 77.82 0.00 0.00 0.00 7.32 (61.54-100.00) (0.00-12.97) (0.00-0.00) (0.00-0.00) (0.00-24.04) 18 High Income 5131 77.85 1.79 0.00 0.00 (0.00- 3.57 (0.00- 19 countries (n=9) (61.54-100.00) (0.00-7.34) (0.00-0.00) 0.00) 13.43) 20 East Asia and Pacific 40 96.34 0.00 0.00 0.00 3.95 21 (n=2) (92.68-100.00) (0.00-0.00) (0.00-0.00) (0.00-0.00) (0.00-7.90) 22 Americas (n=0) - - - - - 23 LMIC Europe (n=0) - - - - - Africa (n=2) 358 40.67 1.48 0.59 0.00 57.26 24 (9.52-71.81) (0.00-2.97) (0.00-1.87) (0.00-0.00) (24.04-90.48) 25 Mediterranean (n=0) - - - - - 26 Adolescents (n=10) 34599 11.00 1.54 36.86 6.62 9.63 27 (0.00-33.33) (0.00-6.14) (6.62-51.75) (0.00-25.00) (1.04-32.84) 28 High Income 34280 18.98 2.06 34.73 14.48 17.36 29 countries (n=4) (0.00-40.00) (0.00-7.87) (0.00-37.58) (0.00-20.83) (5.55 -37.90) East Asia and Pacific 50 3.13 0.00 80.21 14.58 1.04 30 (n=2) (0.00-6.25) (0.00-0.00) (60.42-100.00) (0.00-29.17) (0.00-2.08)
31 Americas (n=1) 68 0.00 4.41 13.24 0.00 13.24 http://bmjpaedsopen.bmj.com/ 32 (0.00-0.00) (4.11-4.41) (13.24-13.24) (0.00-0.00) (13.24-13.24) 33 LMIC Europe (n=0) - - - - - 34 Africa (n=1) 195 0.00 3.08 43.08 41.54 12.31 35 (0.00-0.00) (3.08-3.08) (43.08-43.08) (41.54-41.54) (12.31-12.31) Mediterranean (n=2) 6 16.67 16.67 33.33 0.00 33.33 36 (0.00-33.33) (0.00-33.33) (0.00-66.67) (0.00-0.00) (0.00-66.67) 37 Girls (n=9) 21800 58.35 5.27 15.52 6.48 8.82 38 (0.00-66.67) (0.00-6.48) (0.00-24.44) (3.58-6.67) (0.00-9.82) 39 Boys (n=8) 44714 46.84 5.83 11.18 9.23 11.39 40 (14.13-63.82) (3.13-7.98) (3.64-28.31) (4.22-38.58) (4.76-16.84) Data are number of homicides or median (IQR). n=number of countries with existing data. LMIC= Low and middle income countries. *The high-income 41 countries (classified by the World Bank) included Australia, Austria, Canada, Chile, Denmark, England and Wales, Estonia, Finland, France, Germany, 42 Hungary, Iceland, Isle of Man, Japan, Korea Rep., Netherlands, New Zealand, Norway, Portugal, Scotland, Sweden, Switzerland, the UK and the USA on September 28, 2021 by guest. Protected copyright.
43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjpo Page 15 of 27 BMJ Paediatrics Open bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 Reference 3 4 1. Unicef. Hidden in plain sight: A statistical analysis of violence against children. New York: United 5 6 Nations International Children's Emergency Fund 2014. 7 2. Resnick PJ. Child murder by parents: a psychiatric review of filicide. American journal of psychiatry 8 Confidential:1969;126(3):325-34. For Review Only 9 3. Heide KM. Youth homicide: A review of the literature and a blueprint for action. International 10 journal of offender therapy and comparative criminology 2003;47(1):6-36. 11 4. Overpeck MD, Brenner RA, Trumble AC, et al. Risk factors for infant homicide in the United States. 12 New England Journal of Medicine 1998;339(17):1211-16. 13 5. Friedman SH, Resnick PJ. Neonaticide: Phenomenology and considerations for prevention. 14 International journal of law and psychiatry 2009;32(1):43-47. 15 6. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta- 16 analyses: the PRISMA statement. Annals of internal medicine 2009;151(4):264-69. 17 7. Mathews S, Abrahams N, Jewkes R, et al. The epidemiology of child homicides in South Africa. 18 19 Bulletin of the World Health Organization 2013;91(8):562-68. 20 8. Jha P, Kumar R, Vasa P, et al. Low male-to-female sex ratio of children born in India: national 21 survey of 1· 1 million households. The Lancet 2006;367(9506):211-18. 22 9. United Nations Office on Drugs and Crime. Global study on homicide 2013: trends, contexts, data. 23 Vienna: UNODC 2013. 24 10. Abrahams N, Mathews S, Martin LJ, et al. Gender Differences in Homicide of Neonates, Infants, 25 and Children under 5 y in South Africa: Results from the Cross-Sectional 2009 National Child 26 Homicide Study. Plos Medicine 2016;13(4) 27 11. Kinney HC, Thach BT. The sudden infant death syndrome. New England Journal of Medicine 28 2009;361(8):795-805. 29 12. Greenfield LA. Child Victimizers: Violent Offenders and Their Victims. Washington: U.S. 30 Department of Justice. Office of Justice Programs 1996.
31 http://bmjpaedsopen.bmj.com/ 32 13. Sidebotham P, Bailey S, Belderson P, et al. Fatal child maltreatment in England, 2005–2009. Child 33 Abuse & Neglect 2011;35(4):299-306. 34 14. Adinkrah M. Child Homicide Victimization in Contemporary Fiji. International Journal of 35 Comparative Criminology 2001;1(1):23-39. 36
37 38 39 40 41 42 on September 28, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
https://mc.manuscriptcentral.com/bmjpo BMJ Paediatrics Open Page 16 of 27 bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
1 2 3 4 5 6 7 8 Confidential: For Review Only 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
31 http://bmjpaedsopen.bmj.com/ 32 33 34 35 36 37 38 39 40 41 42 on September 28, 2021 by guest. Protected copyright. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 https://mc.manuscriptcentral.com/bmjpo bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from
http://bmjpaedsopen.bmj.com/ BMJ Paediatrics Open Supplementary Materials Supplementary for https://mc.manuscriptcentral.com/bmjpo Corresponding author. Email: Email: author. Corresponding on September 28, 2021 by guest. Protected copyright. * , Bianca Dekel, Alison Morris-Gehring, Charlotte Watts and Naeemah Abrahams Abrahams Naeemah and Watts Charlotte Morris-Gehring, Alison Dekel, , Bianca * Child homicide Child perpetrators worldwide systematic a review–
Heidi Stöckl Heidi
Confidential: For Review Only
Page 17 of 27 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 bmjpo: first published as 10.1136/bmjpo-2017-000112 on 11 August 2017. Downloaded from − − − − − − − − − − − − − − − − − − − − Page 18 of 27 % Stepfathers%
− − − − − − − − − − − − − − − − − − − − 3,13 3,13 6,67 20,00 0,00 17,39 0,87 7,37 3,33 8,33 0,00 1,85 1,03 1,03 0,64 5,88 % Stepmothers%
− − − − − − − − − − − − − − − − − − % Stepparents%
0,00 35,00 42,71 5,21 25,00 6,25 60,00 26,67 65,22 17,39 34,45 36,36 69,23 45,30 19,05 26,00 25,00 11,67 45,05 8,24 47,37 72,22 61,90 33,33 15,98 25,00 6,25 50,00 1,85 40,21 2,06 63,31 79,31 32,65 36,66 6,45 58,75 100,00 % fathers fathers %
% % mothers
2 100 3 66,67 1 0,00 20 65,00 93 45,83 34 75,00 15 13,33 23 17,39 11 63,64 13 30,77 18 66,67 50 72,00 60 63,33 76 51,32 18 27,78 21 38,10 16 68,75 52 44,44 97 57,73 29 20,69 98 67,35 80 41,25 119 65,55 181 54,70 219 84,02 278 36,69 990 36,66 12915 45,85 http://bmjpaedsopen.bmj.com/ Total homicides Totalby parents homicides
BMJ Paediatrics Open <1 <16 <18 <15 <15 <14 <18 <18 <18 <20 <15 <18 <18 <18 <18 <18 <18 <18 <18 <17 <16 <18 <15 <18 <18 <12 <15 Not stated Age group Age https://mc.manuscriptcentral.com/bmjpo
on September 28, 2021 by guest. Protected copyright. 2003 2007 1995 2000 1991 2005 2000 2005 2003 2011 2001 2010 1991 2000 1985 1991 2006 2013 1994 1998 2007 2011 1982 1994 1989 1998 1976 2007 2005 2014 1985 2009 1990 1999 1989 2016 1960 2005 1985 1994 1991 2008 1996 2015 2001 2005 1997 2004 2000 2005 1998 2009 1974 2005 2004 2007 Years covered
al Local Local Local Local Local Local Local Local Local Local Local National National National National National National National National National National National National National National National National Nation Data type
9 29
Confidential: For Review Only
16
18
28
4
26