First steps towards a core outcome Set for measuring aggressive behavior in prisoners: a systematic review of current methods

Abstract

ABSTRACTA large body of literature has examined the cause and treatment of aggressive behavior in prisons, but heterogeneity in the outcome measures used has led to difficulties comparing outcomes across studies. This systematic review aims to identify a Core Outcome Set for measuring aggression in prisons. Studies were eligible for inclusion if they measured aggressive behavior in adult male prisoners, described original research, and were in English. A systematic search of Medline, PubMed, PsycInfo, CINAHL and Cochrane databases to 7 March 2022 was conducted. Risk of bias was assessed using the QUADAS-2 method. The titles and abstracts of 19,619 records were screened, with 641 reports identified for full-text examination. From these reports, 283 studies with a total of 1,402,931 prisoners were included. These studies were predominantly published from 2000 onwards, published in North America, used a cross-sectional design, and included a single method to measure aggressive behavior. The studies most commonly used self-report and official records of misconduct, with staff report and mixed methods less common. Official records of misconduct and the Buss-Perry Aggression Questionnaire were identified as possible candidates for inclusion into a Core Outcome Set, but caveats exist around the specificity and sensitivity of these measures.KEYWORDS: PrisonaggressionviolenceCore Outcome SetQUADAS-2 AcknowledgementsCHC would like to acknowledge the University of Wollongong’s writing retreat for Early Career Research program, and Professor Chris Cook for early feedback on this article.Disclosure statementNo potential conflict of interest was reported by the author(s).Registration and protocol amendmentsThe protocol for this review was prospectively registered in the International Prospective Register of Systematic Reviews (PROSPERO) (crd.york.ac.uk/PROSPERO, registration number: CRD42020185679). The protocol states that the risk of bias will be assessed using COSMIN with adaptations suggested by Jaspers et al. 2019 (doi.org/10.1016/j.burns.2018.05.015) but this method was not found to be suitable and the QUADAS-2 was used instead.Supplementary materialSupplemental data for this article can be accessed online at https://doi.org/10.1080/14789949.2023.2262446.Additional informationFundingCHC is employed in the NHMRC Partnership Grant (GNT1113396).