台灣民眾自殺行為及自殺前之醫療利用分析

Analysis of People Suicidal Behavior and Their Medical Care Utilization before Suicide in Taiwan

陳明澤(Ming-Tse Chen)

指導教授 ： 邱政元

中山醫學大學/健康管理學院/醫療產業科技管理學系碩士班/碩士(2016年)

摘要

目的
自殺是全世界重要的公共衛生議題，台灣的自殺問題在亞洲排名第三， 僅次於南韓與日本，本研究主要目的在描述台灣歷年來自殺的情況，並分析台灣歷年來自殺者在自殺前一年內的醫療利用情形。

方法
本研究使用2010年全民保險研究資料庫的百萬歸人檔，自2002/01/01到2013/12/31止，研究對象為台灣地區自殺個案之個人屬性檔、門診住院檔…等資料。架構一使用資料庫「住院醫療費用清單明細檔(DD檔)」與外因分類欄位篩選出首次自殺個案共1677人為自殺類型描述性統計樣本; 架構二結合2002-2013之門診及住院檔案中 ICD9-CM代碼資訊，加入另外兩種自殺判定方式，樣本數共3196人。
在基本屬性方面利用T檢定及ANOVA分析，比較自殺個案中年齡、性別、投保金額及區域與平均門診、急診、住院天數…等醫療利用行為差異性；並利用卡方檢定初步分析醫療利用行為與自變項之顯著性；最後使用邏輯斯迴歸分析自殺個案於自殺前一年醫療利用情況。

結果
描述性統計結果顯示於2002年起至2013年止，自殺總人數達1677人，以藥物自殺比例最高，佔61.90%、其次為切穿工具自殺，佔18.60%、一氧化碳自殺則位居第三，佔9.18%，複合性自殺方式中以藥物自殺結合一氧化碳手法自殺為最大宗，占複雜手段自殺人數的48.72%，區域別分層統計自殺人數以中部其他地區自殺比例最高(0.24%)，於投保金額分層下之自殺比例，金額介於2萬到4萬者為最高自殺危險群，不同年齡層下自殺比例最高者為25到44歲者，年齡越大更傾向使用藥物自殺。

推論性統計結果顯示，曾於精神科就診、年齡25-44歲與年齡大於65歲 (v.s 未滿25歲)、男性、CCI共病指數大於2分者，高門診次數利用的機率較大; 曾於精神科就診、區域為其他地區( v.s台北地區)、年齡未滿25歲、男性、CCI共病指數大於2分者，高急診次數利用的機率較大; 曾於精神科就診、區域為其他地區(v.s台北地區)、年齡大於65歲(v.s 未滿25歲)、CCI大於2分(v.s CCI 0分)者，高住院次數利用的機率較大; 曾於精神科就診、年齡大於65歲(v.s 未滿25歲)、CCI共病指數較高者，高住院天數的機率較大; 曾於精神科就診、年齡大於65歲(v.s 未滿25歲)、CCI共病指數越大者，高醫療花費的機率較大。
總結來說，自殺企圖者在自殺前一年的醫療利用中，會去就診精神科的比例是有顯著意義的高，年齡大於65歲者的自殺者自殺前一年的門診次數、住院次數、住院天數與醫療花費均比未滿25歲多，而未滿25歲的這個年齡層，在急診次數均比其他年齡多，顯見這兩個自殺年齡層在自殺前一年醫療資源利用的差異性，而慢性病多者(CCI>2分) 的自殺者自殺前一年的門診次數、住院次數、住院天數與醫療花費均較多。

結論
自殺前一年的醫療利用，以精神科患者、年紀大於65歲 (但急診利用是年紀小較高) 、慢性病多者使用較高; 以自殺防治的角度來看，政府應針對自殺類型、性別、年齡、慢性病患等不同族群做各別的自殺防治策略。

關鍵字

自殺 ； 全民健保資料庫 ； 醫療利用

並列摘要

Objective
Suicide is an important issue of public health around the world. The problem of suicide in Taiwan is ranked third in Asia, second only to South Korea and Japan. The main purpose of this study is to describe the situation regarding suicide in Taiwan over the past years and analyze the medical care utilization of the suicidal people in the year before they committed suicide.

Methods
This study used one million people subsumed files of the National Insurance
Research Database in 2010. From 2002/01/01 to 2013/12/31, the object of this study
is to use individual attribute files and outpatient service files of suicide cases in Taiwan. First construction use information bank “in hospital medical expense detailed list detailed files (the DD files)” screened with the external factor classification fence position for the first time committing suicide. The cases altogether 1677 people are included for descriptive statistics sample; second construction unifies in 2002-2013 outpatient service and in hospital file the ICD9-CM code information, joins other two suicide determination way, the sample number altogether 3196 people.
In basic terms, using T-test and ANOVA analysis to compare the suicide cases by age, sex, the insured amount and region between average outpatient numbers, emergency numbers, hospital days ... and other medical utilization behavioral differences; and using Chi-square test to preliminary analyze medical service utilization between the variable significance; finally using the logistic regression to analyze the medical utilization of the suicidal people in the year before suicide.
Results
Descriptive statistics shows that of the total number of 1677 people who committed suicide from 2002 to 2013, the highest proportion is pharmaceutical suicide, accounting for 61.90%, followed by the cutting tool suicide, accounting for 18.60 %, and carbon monoxide suicide which is ranked third, accounting for 9.18%. The most in compound suicide way is pharmaceutical suicide combined with carbon monoxide, accounting for 48.72%. In regional stratified proportional amount, suicides in the middle portion and any other area has the highest suicide rate ( 0.24% ), in insurance stratified proportional amount, ranging from 20,000 to 40,000 is the highest suicide risk group. Between different ages group the highest proportion of suicides is 25-44 years of age, older people inclined to use drugs to commit suicide.
Inferential statistics showed that, had been received psychiatric treatment, aged 25-44 years old and older than 65 years (v.s younger than 25 years old ), male, CCI comorbidity index more than 2 points, the probability of more outpatient utilization is higher; had been recived psychiatric treatment, the rest of the region (v.s Taipei ), under the age of 25 years old, male, CCI comorbidity index more than 2 points, the probability of more emergency department utilization is higher; had been received psychiatric treatment, the rest of the region(v.s Taipei ), older than 65 years (v.s younger than 25 years old ), CCI comorbidity index more than 2 points (v.s CCI 0 points ), the probability of more hospitalizations is higher; had been received psychiatric treatment, older than 65 years (v.s younger than 25 years old ), CCI comorbidity index higher , the probability of more hospital days is higher ; had been received psychiatric treatment, older than 65 years (v.s younger than 25 years old ), CCI comorbidity index larger, the probilitay of more medical costs is higher.
In conclusion, the medical utilization of suicide attempts before the suicide in the previous year, the proportion of receiving psychiatric treatment is highly significant, the medical utilization of age older than 65 years in the previous year, the outpatient numbers, the number of hospitalizations, hospital days and medical costs are more than age under 25 years old. Age under 25 years has more emergency numbers than other age groups. This result shows the two age groups are different in medical utilization before suicide. The number of hospitalizations, hospital days, health care costs of people who has more chronic diseases (CCI indext more than 2 points) in previous year before suicide, are higher.

Conclusion
The medical utilization within the year before suicide, are higher in the psychiatric patients, age over 65 years old (but younger age has more emergency numbers), and chronic illness; from the view of preventing and controlling suicide, the government should aim at the suicide type, the sex, the age, the chronic illness to make the different suicide preventing and controlling strategies.

並列關鍵字

Suicide ； National Health Insurance Research Database ； Medical service utilization

參考文獻

李明濱、戴傳文、廖士程、江弘基 (2006)：自殺防治策略推動現況與展望。護

李明濱、廖士程 (2006)：自殺原因與防治策略。台灣醫學，10(3): 367-374

李明濱(2007) 精神疾病與自殺防治

李政洋、陳虹汶、李玉嬋、唐婉如、方俊凱(2011): 癌症照顧的自殺防治。 內科

林帛賢、李明濱、陳宜明、廖士程 癌症病患之自殺防治 (2013) 台灣醫學 ;17:163-70