Risk factors for self‐injurious behaviour in custody: Problems of definition and prediction

Abstract

This article reviews the international literature of the last two decades on self-injurious behaviour in prisons and jails and introduces the risk factors associated with this behaviour. Studies from a variety of countries investigated different samples (e.g. in jails or prisons; female or male inmates). We only chose those studies using a control group of inmates without self-injurious behaviour. The findings on potential risk factors for self-injurious behaviour are largely contradictory because of the differences in sample selection and dependent variables (deliberate self-harm without suicidal intent vs. suicide attempts). We also discuss some methodological problems in predicting self-injurious behaviour. Defining the subjectThe results of studies on risk factors for self-injurious behaviour can vary greatly depending on the type of self-injurious behaviour being examined (Borrill et al., 2003;Daigle, 2004;Kerkhof & Bernasco, 1990). Thus, the selection and operational definition of the dependent variable(s) are two of the most basic and decisive steps in designing such a study. In this regard, studies have investigated a wide range of phenomena. Ivanoff (1989) called the definitions idiosyncratic. Malone et al. (1995) found inconsistencies in the definition of suicide attempts not only in the cases they examined, but also in the literature. The definitions included behaviours that ranged from suicidal ideation (i.e. no acting-out behaviour in the real sense; Daigle, 2004) to self-harming behaviour with various degrees of lethality and motives, to acts that might be called suicide attempts because the inmates had relatively strong desires to die or had inflicted considerable harm on themselves. In conceptualizing self-injurious behaviour in custody, some authors, on the one hand, use varying terms such as self-harm, suicide attempt, or parasuicide interchangeably

The predictive validity of the Depression Hopelessness Suicide screening form for self-harm among prisoners

“…The DHS critical items reflect well supported risk factors for self-harm (Dixon- Gordon et al, 2012;Fazel et al, 2008;Lohner & Konrad, 2007), but many of these factors lacked incremental predictive validity after accounting for those factors associated with the greatest risk (i.e. more extensive and/or recent histories of self-harm).…”

“…The 12 critical items assess both historical and current self-injurious and suicidal thoughts and behaviours, a self-reported diagnosis of depression, and whether the inmate has "close friends or family members who have killed themselves". These are among the risk factors with the most consistent support according to recent reviews (Dixon- Gordon et al, 2012;Fazel, Cartwright, Norman-Nott, & Hawton, 2008;Lohner & Konrad, 2007). The user manual provides for discretion of the test administrator to develop a protocol to determine the appropriate response to offenders endorsing one or more of the critical items based on available resources and needs (Mills & Kroner, 2003).…”

Reasons for Non-Suicidal Self-Harm in Adult Male Offenders With and Without Borderline Personality Traits

“…In the review below, we use the terms self-harm, but revert to NSSI when reviewing studies or theories that have focused specifically on non-suicidal self-injury (for discussion of the approaches to defining and conceptualising self-harm see Lohner and Konrad, 2007).…”

Psicopatía y conducta suicida en una muestra de delincuentes con trastorno mental

“…The data in the present work support that such relation could also be applicable to people who suffer a psychiatric pathology, independently of the suicide risk that might be related to such condition. The contrast of these data to other resulting from other studies 9 could be due to methodological matters as the definition of antisocial personality that has been used, the evaluation tools, the information sources (interviews or surveys rather than documental analysis, for instance) or how the suicidal behaviour operates, as some authors have already pointed out 14 . This work presents diverse limitations, mainly, the reduced size of the analyzed sample (which can increase the probabilities of statistical TypeII mistakes, meaning, sustaining a hypothesis which is indeed false) and the heterogeneity of the included diagnoses.…”