1
|
Wang KZ, Bani-Fatemi A, Adanty C, Harripaul R, Griffiths J, Kolla N, Gerretsen P, Graff A, De Luca V. Prediction of physical violence in schizophrenia with machine learning algorithms. Psychiatry Res 2020; 289:112960. [PMID: 32361562 DOI: 10.1016/j.psychres.2020.112960] [Citation(s) in RCA: 13] [Impact Index Per Article: 3.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 10/22/2019] [Revised: 03/17/2020] [Accepted: 03/27/2020] [Indexed: 10/24/2022]
Abstract
Patients with schizophrenia have been shown to have an increased risk for physical violence. While certain features have been identified as risk factors, it has been difficult to integrate these variables to identify violent patients. The present study thus attempts to develop a clinically-relevant predictive tool. In a population of 275 schizophrenia patients, we identified 103 participants as violent and 172 as non-violent through electronic medical documentation, and conducted cross-sectional assessments to identify demographic, clinical, and sociocultural variables. Using these predictors, we utilized seven machine learning classification algorithms to predict for past instances of physical violence. Our classification algorithms predicted with significant accuracy compared to random discrimination alone, and had varying degrees of predictive power, as described by various performance measures. We determined that the random forest model performed marginally better than other algorithms, with an accuracy of 62% and an area under the receiver operator characteristic curve (AUROC) of 0.63. To summarize, machine learning classification algorithms are becoming increasingly valuable, though, optimization of these models is needed to better complement diagnostic decisions regarding early interventional measures to predict instances of physical violence.
Collapse
Affiliation(s)
- Kevin Z Wang
- Group for Suicide Studies, Centre for Addiction and Mental Health, 250 College St, M5T1R8, Toronto, Canada
| | - Ali Bani-Fatemi
- Group for Suicide Studies, Centre for Addiction and Mental Health, 250 College St, M5T1R8, Toronto, Canada
| | - Christopher Adanty
- Group for Suicide Studies, Centre for Addiction and Mental Health, 250 College St, M5T1R8, Toronto, Canada
| | - Ricardo Harripaul
- Group for Suicide Studies, Centre for Addiction and Mental Health, 250 College St, M5T1R8, Toronto, Canada
| | - John Griffiths
- Group for Suicide Studies, Centre for Addiction and Mental Health, 250 College St, M5T1R8, Toronto, Canada
| | - Nathan Kolla
- Group for Suicide Studies, Centre for Addiction and Mental Health, 250 College St, M5T1R8, Toronto, Canada
| | - Philip Gerretsen
- Group for Suicide Studies, Centre for Addiction and Mental Health, 250 College St, M5T1R8, Toronto, Canada
| | - Ariel Graff
- Group for Suicide Studies, Centre for Addiction and Mental Health, 250 College St, M5T1R8, Toronto, Canada
| | - Vincenzo De Luca
- Group for Suicide Studies, Centre for Addiction and Mental Health, 250 College St, M5T1R8, Toronto, Canada.
| |
Collapse
|
2
|
Measurement of agitation and aggression in adult and aged neuropsychiatric patients: review of definitions and frequently used measurement scales. CNS Spectr 2017; 22:407-414. [PMID: 28179043 DOI: 10.1017/s1092852917000050] [Citation(s) in RCA: 17] [Impact Index Per Article: 2.4] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 11/07/2022]
Abstract
Agitation and aggression in adult psychiatric patients with psychoses and in persons with dementia increase the burden of disease and frequently cause hospitalization. The implementation of currently available management strategies and the development of new ones is hindered by inconsistent terminology that confuses agitation with aggression. This confusion is maintained by many rating scales that fail to distinguish between these two syndromes. We review the frequently used rating scales with a particular focus on their ability to separate agitation from aggression. Agitation and aggression are two different syndromes. For example, reactive aggression is often precipitated by rejection of care and may not be associated with agitation per se. We propose, in treatment studies of behavioral symptoms of dementia and challenging behaviors in psychoses, that outcomes should be evaluated separately for agitation and aggression. This is important for investigation of drug effectiveness since the medication may be effective against one syndrome but not the other. Separate assessments of agitation and aggression should be a general principle of trial design with particular salience for registration studies of medications proposed for approval by the U.S. Food and Drug Administration and other regulatory bodies.
Collapse
|
3
|
Association between schizophrenia and violence among Chinese female offenders. Sci Rep 2017; 7:818. [PMID: 28400569 PMCID: PMC5429758 DOI: 10.1038/s41598-017-00975-2] [Citation(s) in RCA: 7] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/14/2016] [Accepted: 03/20/2017] [Indexed: 01/11/2023] Open
Abstract
Little is known about the association between schizophrenia and violence in women in China. This study aimed to examine the association between schizophrenia and violence in Chinese female offenders. Fifty-two schizophrenia patients were identified from the female offenders who received forensic psychiatric assessments in 2011 in Hunan province, China. Using a propensity score matching method, 104 matched controls without psychiatric disorders were selected from female criminals in Hunan province. Violent offences and homicides were verified and recorded. The percentages of violent offences and homicides were significantly higher in female offenders with schizophrenia than in controls (78.8% vs. 30.8%, P < 0.001; 44.2% vs. 18.3%, P = 0.001, respectively). Multivariate logistic regression analyses revealed that diagnosis of schizophrenia, younger age at first offence, living in rural area and a lower education level were independently and positively associated with violent offences, while having a diagnosis of schizophrenia and lower education level were associated with homicides. There appears to be an independent and positive association between schizophrenia and violent offence in Chinese female offenders. Effective preventive approaches on violence in female schizophrenia patients are warranted.
Collapse
|
4
|
Abstract
This study aimed to evaluate the effects of aripiprazole lauroxil on hostility and aggressive behavior in patients with schizophrenia. Patients aged 18-70 years with a diagnosis of schizophrenia and currently experiencing an acute exacerbation or relapse were randomized to intramuscular (IM) aripiprazole lauroxil 441 mg (n=207), 882 mg (n=208), or placebo (n=207) for 12 weeks. In post-hoc analyses, hostility and aggression were assessed by the Positive and Negative Syndrome Scale (PANSS) Hostility item (P7) and a specific antihostility effect was assessed by adjusting for positive symptoms of schizophrenia, somnolence, and akathisia. The PANSS excited component score [P4 (Excitement), P7 (Hostility), G4 (Tension), G8 (Uncooperativeness), and G14 (Poor impulse control)], and the Personal and Social Performance scale disturbing and aggressive behavior domain were also assessed. Of the 147 patients who received aripiprazole lauroxil 882 mg and with a baseline PANSS Hostility item P7 more than 1, there was a significant (P<0.05) improvement versus placebo on the PANSS Hostility item P7 score by mixed-model repeated-measures at the end of the study, which remained significant when PANSS-positive symptoms and somnolence or akathisia were included as additional covariates. The proportion with PANSS Hostility item P7 more than 1 at endpoint was significantly (P<0.05) lower with aripiprazole lauroxil versus placebo (53.6, 46.1, and 66.3% for 441, 882 mg, and placebo). A significant (P<0.05) improvement was found with aripiprazole lauroxil versus placebo for change from baseline in the PANSS excited component score. The proportion of patients with aggressive behavior on the Personal and Social Performance scale was significantly (P<0.05) lower for aripiprazole lauroxil: 30.0% for 441 mg versus 44.1% for placebo (P=0.006) and 22.2% for 881 mg (P<0.001 versus placebo). Treatment with aripiprazole lauroxil resulted in decreases in agitation and hostility in patients with schizophrenia and this antihostility effect appears to be independent of a general antipsychotic effect.
Collapse
|