The Factor Structure and Validity of the Psychopathy Checklist-Short Version When Used With Autistic Psychiatric Inpatients.

Not preregistered, and data, analysis code and research materials are not available. The paper reports 282 participants, 257 PCL:SV total scores and 26 missing PCL:SV data, so the analysed count is not fully clear. Some findings changed after missing-data sensitivity analysis, including Factor 2 and Mental Health Act section. Aggressive or problematic behaviour was recorded as frequency and presence or absence, not severity, so high-frequency low-intensity autism-related behaviours may have been mixed with more serious offending behaviours. Autism symptom severity was not measured, and the authors note that autism or intellectual disability behaviours may contribute to PCL:SV scores, especially Factor 2. Forty-nine percent had an intellectual disability, and intellectual disability was associated with fewer current offences, violent offences and forensic background in several models. Clinicians knew participants' histories and records when completing the PCL:SV, and no inter-rater reliability checks were undertaken. The sample was autistic psychiatric inpatients in England and Wales, with 251 males, 30 females and one transgender person, and results do not generalise to autistic prisoners, community autistic people or other countries. The PCL:SV was poor at discriminating aggressive or problematic behaviour, so it should not be treated as a standalone behaviour prediction tool.

Funded by the National Institute for Health and Care Research. Authors declared no conflicts of interest.

Do not use the PCL:SV as a standalone tool to predict aggressive or problematic behaviour in autistic inpatients. It had poor ability to discriminate these behaviours, and some behaviours may reflect autism or intellectual disability rather than psychopathy.