目的：比較急性精神住院病患在精神衛生法公佈施行（ 1990 年 12 月 7 日）前 後各六年的死亡率並探討危險因子。方法：將一所精神專科醫院於 1985 年至 1996 年急性 精神住院病患（ N=8847 ），以我國精神衛生法的公佈施行分為獨立的前後二組（ 1985-90 年， 簡稱前組； 1991-96 年，簡稱後組）， 以記錄連結方式， 分別追蹤至 1990 年底與 1996 年底， 而以標準化死亡比（ standardized mortality ratio，簡稱 SMR ）為指標， 比較兩組之差異。 結果：前後組各有 139 及 272 位個案死亡。依診斷別，後組的 SMR 均 有下降的趨勢，尤其是精神分裂症（前組與後組之 SMR 分別為 5.25 與 3.39 ）。 若再分 為自然及非自然死因，各個診斷別在自然死亡比下降，但在非自然死亡方面，情感性精神疾 患及物質使用疾患則反而增加。 診斷別中， 前後組均以物質使用疾患的 SMR （分別為 12.33 與 9.26 ）最高，且在相對危險性上，亦均大於其他診斷，顯示物質使用疾患對急性 精神住院患者亦為高危險群。結論：後期急性精神住院病患的自然死亡有減少趨勢，但相對 地，非自然死亡無下降趨勢，此與西方工業化國家去機構化的過程類似。雖後期的死亡率仍 偏高，但未來應強調非自然死亡的預防，特別是物質使用疾患。

　　　　　Objective: This study investigated the difference of mortality in acute psychiatric inpatients before and after Mental Health Act, which became law on December 7, 1990. It also explored the risk factors for mortality in these patients. Methods: All subjects admitted to acute wards in psychiatric hospital from 1985 to 1996 (N=8847) were included in the analysis. These patients were divided into subgroups based on pre- and post-Mental Health Act eras (1985-90 vs. 1991-96). Using record linkage analysis, records on the previous subgroup were followed up to December 1990 while records from the later subgroup were followed until December 1996. The standardized mortality ratio (SMR) was used as an index for comparison. Results: There are 139 and 272 expired subjects in the pre- and post-Mental Health Act subgroups, respectively. All of the SMRs for different diagnoses declined in the later subgroup. This decline was marked in the schizophrenia subgroup (previous vs later: 5.25 vs. 3.39). The SMRs for natural death declined after enactment. However, the incidence of unnatural death SMRs in affective disorder and substance use disorders increased in the post-Mental Health Act era. SMRs for substance use disorder (SUD) were the highest among all diagnostic categories in both pre- and post-enactment subgroups (12.33 vs. 9.26). The relative risk in SUD was also highest, showing that subjects with SUD are the highest risk group among acute psychiatric inpatients. Conclusion: The results of this study show that the mortality rate caused by natural death declined after enactment of the Mental Health Act, while the unnatural death rate did not decline. This trend is similar to reports on the effects of the deinstutionalized process in Western industralized countries. The study found an excessive mortality rate among acute psychiatric inpatients in the post-Mental Health Act era, suggesting that the prevention of unnatural death should be a major area of emphasis in the treatment of psychiatric inpatients, especially in patients with SUD.