What is the Prevalence of Self-harming and Suicidal Behaviour in Under 18s with ASD, With or Without an Intellectual Disability?

Abstract

A systematic literature review was undertaken to ascertain the prevalence of self-harm and suicidal behaviour in children and young people under 18 years old with a diagnosis of autism spectrum disorder (ASD), with or without an intellectual disability. There was variation in the reported prevalence rates but results suggested that rates of both self-harm and suicidal behaviour may be elevated in ASD compared to the general population. This is in keeping with literature relating to autistic adults but in contrast to conclusions of a previous systematic review. This review highlights the need for further research to explore the experience of self-harm and suicidal behaviour in autistic children and young people.

Appendices

Appendices

Appendix 1: Quality Assessment Summary Grid

Appendix 2: Scoring Grid Criteria

Domain	Criteria	Score/classification	Scoring criteria
Aims	Are the aims, hypotheses and objectives clearly described?	3—well covered	The aims and hypotheses of the study are clearly described and state that they are determining the prevalence of self-harming/suicidal behaviour
		2—adequately covered	Aims and hypotheses are stated but somewhat unclear. Minimal or no reference to the prevalence of self-harming/suicidal behaviour
		1—poorly covered/not addressed	The aims of the study are unclear or not stated
	Is determining the prevalence of self-harming/suicidal behaviour the primary aim?	3—well covered	Yes and this is stated explicitly in the aims of the paper
		2—adequately covered	This is a secondary aim of the paper
		1—poorly covered/not addressed	This is not stated as one of the aims of the paper but can be extracted from the data described
	Is a description or definition of self-harming or suicidal behaviours provided?	3—well covered	Clear description of self-harming or suicidal behaviours is outlined. Description is in line with the definition given in this review
		2—adequately covered	Limited description of self-harming or suicidal behaviours outlined. Description may not align with definition given in this review
		1—poorly covered/not addressed	Minimal description of self-harming or suicidal behaviours included
Sample	Does the recruitment methodology attempt to gain a sample that is representative of the study population?	3—well covered	A population based, non-clinical sample was targeted (e.g. all eligible participants in a geographical area) and differences between participants and non-participants were controlled for
		2—adequately covered	A highly selective recruitment method was used (e.g. selectively referred patients to secondary or tertiary care)
		1—poorly covered/not addressed	Recruitment method was inadequately explained or not reported
	Are sample characteristics of the participants described?	3—well covered	Demographic details of the children recruited such as age, gender and/or levels of impairment are clearly stated
		2—adequately covered	Some demographic details of the children recruited are reported but additional information could be provided
		1—poorly covered/not addressed	Minimal or no demographic details are provided
	Are study inclusion and exclusion criteria described?	3—well covered	Inclusion/ exclusion criteria are explicitly reported
		2—adequately covered	Inclusion/ exclusion criteria are provided but additional information would improve clarity
		1—poorly covered/not addressed	Inclusion/exclusion criteria are not explicitly reported
	Was there a control or comparison group?	3—well covered	Yes, control or comparison group included and well described. Control/comparison group were recruited in the same way and evidence is provided of their equivalence on possible confounds
		2—adequately covered	A control group is included but limited effort has gone into ensuring that this is the most appropriate group and that non-specific effects are controlled for
		1—poorly covered/not addressed	No control or comparison group included in study
Diagnostic confirmation	How was a diagnosis of ASD confirmed for inclusion into the study?	3—well covered	Diagnoses have been confirmed for this study by use of a ‘gold-standard’ diagnostic tool (i.e. ADOS or ADI-R)
		2—adequately covered	Diagnoses have been confirmed for this study, but not by use of a gold standard tool
		1—poorly covered/not addressed	Paper reports that diagnoses were confirmed but does not provide detail how
	How was a diagnosis of Intellectual disability confirmed?	3—well covered	Diagnosis has been confirmed for this study by use of a ‘gold-standard’ diagnostic tool (i.e. WISC or Stanford Binet)
		2—adequately covered	Diagnosis has been confirmed for this study, but not by use of a gold standard tool
		1—poorly covered/not addressed	Paper reports that diagnoses were confirmed but does not provide detail how
Outcome measures	Were self-harming and suicidal behaviours measured using standardised, valid and reliable measures for the population?	3—well covered	Standardised measures are used and evidence of those which validated for use in children and young people with ASD ± an intellectual disability
		2—adequately covered	Standardised measure used, although evidence of validity and reliability not provided in this study, but reference to cited studies provide evidence of acceptable properties
		1—poorly covered/not addressed	Non standardised or validated measures are used, limited reported evidence of validity or reliability provided
	Are self-harming and suicidal behaviours reported by participants, by parents/carers, by both or by review of notes?	3—well covered	Behaviours reported by participants
		2—adequately covered	Behaviours reported by parents/carers
		1—poorly covered/not addressed	Behaviours reported by reviewing notes/medical records
Data analysis	Are the statistical methods selected appropriate (i.e. suitable for study design, sample size and data type)?	3—well covered	The analytic strategy is appropriate to the design and data
		2—adequately covered	The analytic strategy is appropriate but has some limitations (e.g. other analytical strategy would have been more powerful or some assumptions have been violated) or limited statistical analysis was carried out
		1—poorly covered/not addressed	Inappropriate statistical tests were used or insufficient information is provided to judge the appropriateness of the analysis
	Is the sample size sufficient?	3—well covered	Sample size appears large and appropriate for prevalence in population based study
		2—adequately covered	Sample size appears adequate. Some discussion about size of sample
		1—poorly covered/not addressed	No justification is given for sample size and sample size appears small
	Have measures to handle missing data been described and selected appropriately?	3—well covered	A clear description as to how missing data has been handled is outlined. The approach selected is robust and appropriate
		2—adequately covered	A description of how missing data has been handled is outlined however the method chosen is based upon factors such as simplicity/convenience (e.g. Listwise deletion, last observation carried forward)
		1—poorly covered/not addressed	Analysis does not consider the handling of missing data in any form
Interpretation	Do the conclusions follow adequately from the results and address the aims?	3—well covered	Main findings are clearly described and follow appropriately from the results and analyses. They address the initial stated aims
		2—adequately covered	Some limitations in the clarity of description of main findings and their relation to results. May not address stated aims fully
		1—poorly covered/not addressed	Lack of appropriate description of findings and/or findings are over/ understated and do not follow clearly from results. Do not appear to address initial aims
	Is clinical significance of results considered?	3—well covered	In depth consideration is given to the clinical significance of results or outcome measures have clear inherent clinical meaning (e.g. meeting diagnostic criteria or not, clinical severity etc.)
		2—adequately covered	Some limited consideration of clinical significance in discussion or results or measures provide some indication of clinical significance
		1—poorly covered/not addressed	Clinical significance of findings not addressed
	Are limitations acknowledged?	3—well covered	Clear acknowledgement of main limitations of the study and consideration given to the impact of these on interpretation
		2—adequately covered	Some limitations are acknowledged but not all, or no consideration given to the impact of limitations on interpretation
		1—poorly covered/not addressed	Acknowledgement of limitations largely absent
Overall quality		Good	Majority of domains defined as well-covered
		Adequate	Majority of domains are described as adequately covered with some rated as well-covered and/or poorly-covered
		Poor	Few if any domains defined as well or adequately covered. Majority are described as poorly covered