Summary Background Rates of suicide among people in prison are elevated compared with people of similar age and sex who are living in the community. Improving assessments and interventions to reduce suicide risk requires updated evidence on risk factors. We aimed to examine risk factors associated with suicide in prisoners. Methods We did an updated systematic review and meta-analysis of risk factors for suicide among people in prison. We searched five biblographic databases for articles published between Jan 1, 2006, and Aug 13, 2020, and one database for articles published between Jan 1, 1973, and Aug 13, 2020. Eligible studies reported risk factors in individuals who died by suicide while in prison and in controls from the general prison population. Two reviewers independently extracted data for each study using a standardised form. We calculated random-effects pooled odds ratios (ORs) for the association of suicide with demographical, clinical, criminological, and institutional risk factors, and investigated heterogeneity using subgroup and meta-regression analyses. This systematic review is registered with PROSPERO, CRD42020137979. Findings We identified 8041 records through our searches, and used 77 eligible studies from 27 countries, including 35 351 suicides, in the main analysis. The strongest clinical factors associated with suicide were suicidal ideation during the current period in prison (OR 15·2, 95% CI 8·5–27·0), a history of attempted suicide (OR 8·2, 4·4–15·3), and current psychiatric diagnosis (OR 6·4, 3·6–11·1). Institutional factors associated with suicide included occupation of a single cell (OR 6·8, 2·3–19·8) and having no social visits (OR 1·9, 1·5–2·4). Criminological factors included remand status (OR 3·6, 3·1–4·1), serving a life sentence (OR 2·4, 1·3–4·6), and being convicted of a violent offence, in particular homicide (OR 3·1, 2·2–4·2). Interpretation Several modifiable risk factors, such as psychiatric diagnosis, suicidal ideation during the current period in prison, and single-cell occupancy, are associated with suicide among people in prison. Preventive interventions should target these risk factors and include improved access to evidence-based mental health care. Understanding other factors associated with suicide might improve risk stratification and resource allocation in prison services. Funding Wellcome Trust, National Institute for Health Research Applied Research Collaboration Oxford and Thames Valley.

IMPORTANCEThe endocannabinoid system (ECS) is a lipid-based endogenous signaling system. Its relevance to psychosis is through the association between cannabis use and the onset and course of illness and through the antipsychotic properties of cannabidiol, a potential ECS enhancer.OBJECTIVE To conduct a systematic review and meta-analysis of the blood and cerebrospinal fluid (CSF) measures of the ECS in psychotic disorders.DATA SOURCES Web of Science and PubMed were searched from inception through June 13, 2018. The articles identified were reviewed, as were citations to previous publications and the reference lists of retrieved articles.STUDY SELECTION Original articles were included that reported blood or CSF measures of ECS activity in patients with psychotic illnesses and in healthy controls.DATA EXTRACTION AND SYNTHESIS PRISMA guidelines, independent extraction by multiple observers, and random-effects meta-analysis were used. Heterogeneity was assessed with the I 2 index. Sensitivity analyses tested the robustness of the results. MAIN OUTCOMES AND MEASURESThe clinical relevance of ECS modifications in psychotic disorders was investigated by (1) a quantitative synthesis of the differences in blood and CSF markers of the ECS between patients and healthy controls, and (2) a qualitative synthesis of the association of these markers with symptom severity, stage of illness, and response to treatment. RESULTS A total of 18 studies were included. Three individual meta-analyses were performed to identify the differences in ECS markers between people with schizophrenia and healthy controls. Five studies, including 226 patients and 385 controls, reported significantly higher concentrations of anandamide in the CSF of patients than controls (standardized mean difference [SMD], 0.97; 95% CI, 0.67-1.26; P < .001; I 2 = 54.8%). In 9 studies, with 344 patients and 411 controls, significantly higher anandamide levels in blood were found in patients, compared with controls (SMD, 0.55; 95% CI, 0.05-1.04; P = .03; I 2 = 89.6%). In 3 studies, involving 88 patients and 179 controls, a significantly higher expression of type 1 cannabinoid receptors on peripheral immune cells was reported in patients compared with controls (SMD, 0.57; 95% CI, 0.31-0.84; P < .001; I 2 = 0%). Higher ECS tone was found at an early stage of illness in individuals who were antipsychotic naïve or free, and it had an inverse association with symptom severity and was normalized after successful treatment. Moderate to high level of heterogeneity in methods was found between studies. CONCLUSIONS AND RELEVANCETesting clinically relevant markers of the ECS in the blood and CSF of people with psychotic illness appears possible, and these markers provide useful biomarkers for the psychotic disorder; however, not all studies accounted for important variables, such as cannabis use.

Background: Oxford Mental Illness and Suicide tool (OxMIS) is a brief, scalable, freely available, structured risk assessment tool to assess suicide risk in patients with severe mental illness (schizophrenia-spectrum disorders or bipolar disorder). OxMIS requires further external validation, but a lack of large-scale cohorts with relevant variables makes this challenging. Electronic health records provide possible data sources for external validation of risk prediction tools. However, they contain large amounts of information within free-text that is not readily extractable. In this study, we examined the feasibility of identifying suicide predictors needed to validate OxMIS in routinely collected electronic health records. Methods: In study 1, we manually reviewed electronic health records of 57 patients with severe mental illness to calculate OxMIS risk scores. In study 2, we examined the feasibility of using natural language processing to scale up this process. We used anonymized freetext documents from the Clinical Record Interactive Search database to train a named entity recognition model, a machine learning technique which recognizes concepts in freetext. The model identified eight concepts relevant for suicide risk assessment: medication (antidepressant/antipsychotic treatment), violence, education, self-harm, benefits receipt, drug/alcohol use disorder, suicide, and psychiatric admission. We assessed model performance in terms of precision (similar to positive predictive value), recall (similar to sensitivity) and F1 statistic (an overall performance measure). Results: In study 1, we estimated suicide risk for all patients using the OxMIS calculator, giving a range of 12 month risk estimates from 0.1-3.4%. For 13 out of 17 predictors, there was no missing information in electronic health records. For the remaining 4 predictors missingness ranged from 7-26%; to account for these missing variables, it was possible for OxMIS to estimate suicide risk using a range of scores. In study 2, the named entity recognition model had an overall precision of 0.77, recall of 0.90 and F1 score of 0.83. The concept with the best precision and recall was medication (precision

Background There has been a rapid growth in the publication of new prediction models relevant to child and adolescent mental health. However, before their implementation into clinical services, it is necessary to appraise the quality of their methods and reporting. We conducted a systematic review of new prediction models in child and adolescent mental health, and examined their development and validation. Method We searched five databases for studies developing or validating multivariable prediction models for individuals aged 18 years old or younger from 1 January 2018 to 18 February 2021. Quality of reporting was assessed using the Transparent Reporting of a multivariable prediction models for Individual Prognosis Or Diagnosis checklist, and quality of methodology using items based on expert guidance and the PROBAST tool. Results We identified 100 eligible studies: 41 developing a new prediction model, 48 validating an existing model and 11 that included both development and validation. Most publications (k = 75) reported a model discrimination measure, while 26 investigations reported calibration. Of 52 new prediction models, six (12%) were for suicidal outcomes, 18 (35%) for future diagnosis, five (10%) for child maltreatment. Other outcomes included violence, crime, and functional outcomes. Eleven new models (21%) were developed for use in high‐risk populations. Of development studies, around a third were sufficiently statistically powered (k = 16%, 31%), while this was lower for validation investigations (k = 12, 25%). In terms of performance, the discrimination (as measured by the C‐statistic) for new models ranged from 0.57 for a tool predicting ADHD diagnosis in an external validation sample to 0.99 for a machine learning model predicting foster care permanency. Conclusions Although some tools have recently been developed for child and adolescent mental health for prognosis and child maltreatment, none can be currently recommended for clinical practice due to a combination of methodological limitations and poor model performance. New work needs to use ensure sufficient sample sizes, representative samples, and testing of model calibration.