治療経過中の自傷行為予測に対する初診時問診の有用性 : 心療内科専門施設での検討  [in Japanese] The Utility of Symptoms and Past History at the First Visit to Predict Self-injury Behavior during Treatment : Research in a Psychosomatic Clinic  [in Japanese]

初診時の問診や質問紙が治療中の自傷行為の予測に有用かを検討した.対象と方法:2000〜2005年に初診した1,665名を対象とした.(男性605名,女性1,060名,年齢36.4±18.6歳).カルテ記載から精神科・心療内科受診歴(受診歴),希死念慮,自傷行為の既往(自傷既往),受診直前の自傷行為(直前自傷),治療経過中の自傷行為(治療中自傷)などを調べた.初診時にGHQ28を行った.結果:全対象の22.6%に受診歴,24.7%に希死念慮,5.1%に自傷既往,1.5%に直前自傷を認めた.初診以降も治療を継続した1,132名中,治療中自傷は4.3%にみられ,非自傷者より低年齢で,受診歴,希死念慮,自傷既往が多く,GHQ28の重症抑うつ尺度が高く,診断別では摂食障害,うつ病に多かった.多変量解析では年齢,希死念慮,自傷既往が治療中自傷予測に有意な変数だが,自傷者の正分類率は4%だった.結語:初診時の希死念慮と自傷行為の既往が治療中自傷の予測に最も有用だが,十分ではない.

Objectives: In this study, we examined whether clinical profiles and psychological tests can predict the act of self-injury during treatment. Subjects and Method: Subjects were 1665 patients (male 605, female 1060, age 36.4±18.6 years) who had consulted at our clinic between 2000 and 2005. All subjects completed GHQ at the first visit. We assessed the history of consultation at psychiatry or psychosomatic clinics, the suicide ideation, the history of self-injury behavior, the act of self-injury just before having the first session at our clinic and the act of self-injury during our treatment by checking on clinical records. Results: Twenty-three percent of all subjects had the history of treatment at other clinics, 24.7 percent had suicide ideation, 5.1 percent had a history of self-injury behavior, and 1.5 percent had that just before consulting at our clinic. The act of self-injury during treatment was seen in 4.3 percent of 1132 patients who received treatment in our clinic. Self-harmed individuals were younger, and had higher prevalence of the history of psychiatric or psychosomatic treatment, the suicide ideation, the history of self-injury behavior, and showed elevated value of severer depression subscale of GHQ compared with non-self-harmed ones. Patients with eating disorders and depression hurt themselves more often than those with other diagnosis. Multivariate analyses revealed that the age, the suicide ideation and the history of self-injury behavior were significant variables to predict the self-harmed behavior in treatment. However, the prediction of the self-harmed behavior on an individual level was difficult. Conclusion: It was suggested that the age, the suicide ideation and the history of self-injury behavior at the first medical examination were most useful predictors of the self-harmed behavior during treatment among parameters examined in the current study. But predictive power was insufficient.