The psychometric properties of the Life History of Aggression evaluated in patients from a psychiatric emergency setting

Abstract

Aggression is a major concern in psychiatric units even more so in emergency settings and is prevalent across many mental disorders. Since a prior history of aggression remains a crucial predictor of violence in hospitalized patients, appropriate tools assessing the history of violence are essential to implement management plans. The Life History of Aggression (LHA) is a widely-utilized instrument. Yet, its psychometric properties have not been thoroughly evaluated and it has neither been validated in a variety of psychiatric illnesses nor in psychiatric emergency settings. To fill this gap, the properties of a French version of the LHA administered to 1097 psychiatric patients in this setting were examined. The internal consistency, construct validity, test-retest reliability and convergent validity were analysed. Regarding the internal consistency, the Cronbach's alpha varied between 0.83 and 0.89. The confirmatory factor analysis showed that it had one underlying factor, though gender invariance was not observed. Amid 148 patients reassessed within 30 days, the test-retest reliability produced a strong correlation coefficient above 0.80. Lastly, as for convergent validity, the scale was significantly correlated with a validated drug misuse scale. Hence, the LHA retains very good psychometric properties even when evaluated in a psychiatric emergency setting, justifying its clinical pertinence.