| Once psychological crisis occurs,people often show a variety of behavioral signs.By systematic study of these signs,we would have an effective-tool for the early identification and warning of psychological crisis.At present,there is still a lack of unified definition on psychological crisis in the academic community,risk factors and crisis signs are often undifferentiated,crisis signs and symptoms are usually confused.The psychological crisis signs published by various research institutions and professional websites are inconsistent;most of the identified signs are difficult to determine their sensitivity and specificity.In China,the crisis early warning is mainly characterized by static warning system,which is based on self-rating screening scales,ignoring the role of crisis gatekeepers.Crisis signs research has not yet received attention,localized study is extremely lacking.The main purposes of the current study were to develop a crisis signs indicators system and early warning strategies suitable for the cultural and cognitive characteristics of Chinese people,and to test the efficacy of the crisis gatekeepers’ behavioral observation dynamic warning system.The current study focused on two types of common crisis:suicidal/self-harming crisis and violent crisis,which might cause significant negative consequences.The current study developed a new tool for the identification and early warning of psychological crisis---Psychological Crisis Signs Checklist(PCSC)through two rounds of Delphi expert advice.We modified PCSC into a 30-item scale through two rounds of administration.We also found that PCSC has good psychometric characteristics in terms of internal reliability,inter-rater reliability,as well as content validity,discriminant validity and criterion-related validity.The current study developed three independent crisis warning strategies to alert yellow and red crisis respectively:total cutoff score warning strategy,single-item warning strategy,and multi-item combination warning strategy.Yellow crisis indicates less serious suicidal/self-injury or violent risks,red crisis indicates a high degree of suicidal/self-injury or violent risks.Total cutoff score warning strategy is stipulated as:when the total PCSC score is≥9<14,yellow crisis alarms;when the total cutoff score is≥14,red crisis alarms.We also identified 3 items with high specificity as red crisis single-item warning indicators,7 items with high specificity as yellow crisis single-item warning indicators.By creatively using supervised machine learning modeling and support vector machine as the prediction model,we identified the following multi-item combination warning strategies:the three-item combination could predict yellow crisis with satisfactory accuracy,and the five-item combination could predict red crisis with satisfactory accuracy.We tested the efficacy of the three warning strategies in a sample recruited from multi-centers(N=217)and found that the three early warning strategies could complementarily predict crisis with high warning sensitivity and accuracy.By using the three warning strategies simultaneously,a flexible,dynamic and graded crisis early warning system could be established.The three-warning-strategy-combination also solved the predicament existing in previous studies that the sensitivity and specificity could hardly be balanced by solely using the total cutoff score to screen high-risk population.PCSC provides crisis gatekeepers a scientific,operational monitoring tool,which would greatly improve the early warning effects of crisis gatekeepers.This tool could be used in colleges and universities,enterprises,prisons and other institutions,to train the crisis gatekeepers as an important part of the crisis early warning system.To improve the overall crisis warning efficacy,the current behavioral observation dynamic warning strategy should be used simultaneously and complementarily with the traditional static warning system,which is based on the self-rating screening scales. |
PDF Full Text Request