以統計方法探討引起自殺或再自殺之可能危險因素

Using statistical method to explore the possible risk factors for suicide and its recurrence

林盈佳(Ying-Chia Lin)

指導教授 ： 張玉坤

淡江大學/理學院/數學學系碩士班/碩士(2005年)

https://doi.org/10.6846/TKU.2005.00562

摘要

近十年來台灣地區自殺率居高不下，自殺問題逐漸受到重視，相關的議題包括在流行病學、精神病學、心理學、社會學、經濟學…等方面皆有研究。
　　本研究試圖探討引起自殺與再自殺之可能危險因素，採用四組台灣醫學中心資料，並以統計方法中的Logistic Regression Model之涉險率探討引起再自殺的可能危險因素，再以存活分析中Kaplan-Meier Estimate方法之累積發生率以及Cox's Proportion Hazards Model之瞬間發生率，探討引起第一次自殺或第一次因自殺而就診之年齡的可能影響因素。
　　研究結果顯示，病患再自殺方面，年齡層愈低愈危險，對於女性以及未婚者皆是事件再發率較高的族群，進而深入探討在第一次自殺或是第一次因自殺而就診之年齡的可能影響因素方面，具顯著性的影響因素有教育程度以及婚姻狀況，而國中(含)以下明顯比其他教育程度者有較低的發生率。

關鍵字

自殺 ； 重複事件 ； 存活分析 ； 羅吉斯迴歸 ； Kaplan-Meier方法 ； 卡氏迴歸模型

並列摘要

The suicide rate in Taiwan has remained highly in the last decade. Gradually, the suicide related problems have been taken seriously. Researchers from many different fields have been devoting themselves to explore this topic, e.g. epidemiology, psychiatry, psychology, sociology, economics...etc.
　The aim of this study is to explore the possible risk factors for suicide and its recurrence. Accumulating four data sets from four medical centers in Taiwan, we applied logistic regression to explore the possible relationships between the odds of recurrent event of suicide and some potential prognostic factors. The comparisons of cumulative occurrence rates of suicide were done by the Kaplan-Meier method. Cox's Proportion Hazards Models were used to compare the duration from date of birth to the time of first suicide or time to hospitalization due to suicide after adjusting the effects of some potential confounding variables.
　We found that the risk of recurrent rate was higher for younger person. Other significant risk factors were gender (Female was higher than male) and marriage status (single was higher than married). After adjusting the effect of age, the hazard rates of first time suicide and/or first time hospitalization due to suicide were statistically significant among different education levels and/or marital statuses. Especially, person with education level lower than or equal to junior high has lower occurrence rate compare to other education level person.

並列關鍵字

Suicide ； Recurrent event ； Survival analysis ； Logistic regression analysis ； Kaplan-Meier Method ； Cox regression model

參考文獻

[1] 行政院衛生署統計資訊網

[1] Yiing-Jenq Chou, Nicole Huang et al. (2003).Suicides after the 1999 Taiwan earthquake. International journal of Epidemiology;32:1007-1014.

[2] Elisa,T.Lee. (2003).Statistical Methods for Survival Data Analysis. John Wiley & Sons, lnc.

[3] Gumnell, D. and Middleton, N. (2003). National Suicide Rates as an Indicator of the Effect of Suicide Premature Mortality. The Lancet, 362:961-62.

[4] Levi, Fabio et al. (2003). Global Suicide Rates. European Journal of Public Health, 13(2): 97-98.