精神分裂症病人使用抗精神病藥物之中風風險探討—嵌入型病例對照研究

Abstract

研究背景：抗精神病藥物廣泛用於治療精神分裂症、憂鬱症與躁鬱症等精神疾病，以及其他原因導致的精神異常，例如失智症病人的精神異常及行為問題。近來有關於抗精神病藥物用於失智症病人及老年人會增加腦血管疾病風險的研究，陸續被發表。但目前相關研究並未觸及最為廣泛使用抗精神病藥物的精神分裂症病人族群。考慮到精神分裂症病人特性與失智症病人或一般老年人的不同，藉研究探討精神分裂症病人使用抗精神病藥物與中風風險的相關性，並找出影響最鉅的藥物種類及其他風險因子，以提供臨床醫療人員為精神分裂症病人選擇藥物的參考，至為重要。 研究目的：分析新診斷精神分裂症病人中，發生中風事件與未發生中風事件者，在使用抗精神病藥物及其他中風相關風險因子的差異，以了解在精神分裂症病人使用抗精神病藥物與中風風險的相關性；並找出開始使用抗精神病藥物後，中風風險最高的時間點，以及找出影響中風風險最為顯著之抗精神病藥物。 研究方法：本研究為一回溯性嵌入式病例對照研究，以臺灣健保資料庫2000年及2005年兩套百萬人承保抽樣歸人檔為資料來源，找出2001年至2009年間新診斷為精神分裂症的病人，並刪除診斷精神分裂症前有中風相關診斷者，為本研究之研究世代。自此研究世代中挑選出進入研究世代後有中風或暫時性腦缺血診斷者，為本研究之病例組，以首次出現中風診斷之日期為index date；並依性別及年齡，與研究世代中其他無中風診斷者，以1:2比例進行隨機配對，為本研究之對照組。將研究個體依最後使用抗精神病藥物日期（處方結束日）與index date之時間間隔，分為目前使用者（7日內）、最近曾使用者（8至30日內）以及過去曾使用者（大於30日）三組。並將目前使用者組依index date前處方抗精神病藥物之時間長度，分為0-15日、16-30日、31-90日與超過90日四組。另針對抗精神病藥物，分為單獨使用第一代抗精神病藥物、單獨使用第二代抗精神病藥物與合併使用第一代及第二代抗精神病藥物三組，並細分為各種不同成分的抗精神病藥物，分別進行統計分析，找出對中風風險影響最鉅的藥物種類。本研究所納入其他與中風風險相關的共病症包括高血壓、缺血性心臟病、心衰竭、心律不整、糖尿病、血脂異常、慢性肺阻塞肺病、失智、帕金森氏症以及癌症；併用藥物包括抗心律不整藥物、降血壓藥物、降血脂藥物、抗血小板藥物、warfarin、抗憂鬱藥物、鋰鹽、valproate、phenytoin、苯二氮平類藥物、非類固醇類消炎藥、避孕藥以及類固醇。針對病例組與對照組之基本性質分析，連續變項使用t-test，類別變項則使用Chi-square analysis；針對抗精神病藥物與中風風險的相關性，則使用條件式邏輯迴歸模型進行統計分析。 研究結果：新診斷精神分裂症且在進入研究世代前無中風相關診斷者共9715人，病例組386人，對照組772人。基本性質分析發現，病例組相較於對照組，在性別、年齡及平均觀察時間無統計顯著差異，但有中風相關之共病症及併用藥物的比例較高。經校正風險因子後發現，抗精神病藥物之目前使用者相較於未使用者（非抗精神病藥物使用者），中風風險顯著較高（OR: 1.94，95% CI: 1.11-3.39）。最近曾使用者與過去曾使用者之中風風險相較於未使用者則未有統計上顯著差異。由多變項條件式邏輯迴歸分析經stepwise篩選出最終模式發現，顯著增加中風風險的共病症有失智、慢性肺阻塞肺病以及高血壓，併用藥物則有抗血小板藥物、非類固醇類消炎藥、valproate及phenytoin。在使用抗精神病藥物之時間長短的分析，發現無論抗精神病藥物使用時間多長，目前使用者之中風風險均顯著高於未使用者，但以0-15日及16-30日組之中風風險較高，OR分別為13.04及13.66；31-90日與超過90日組之OR則較低，分別為3.19及3.87。經校正風險因子後在0-15日組及16-30日組之OR分別為9.41（95% CI: 3.08-28.71）及6.90（95% CI: 1.09-43.69），31-90日與超過90日組則未達統計顯著差異。在目前使用者之抗精神病藥物種類的分析，則發現使用第一代抗精神病藥物與併用第一代及第二代抗精神病藥物者，中風風險顯著高於未使用者，經校正風險因子後之OR分別為2.75（95% CI: 1.34-5.64）及2.37（95% CI: 1.20-4.68），而單獨使用第二代抗精神病藥物者相較於未使用者，無統計顯著差異。Index date前7日內有使用chlorpromazine、flupentixol、haloperidol及prochlorperazine者，發生中風事件之風險顯著較使用其他藥物者高，其中以prochlorperazine風險最高（OR: 9.83，95% CI: 1.49-65.02），chlorpromazine次之（OR: 9.26，95% CI: 1.54-55.69）；而使用risperidone者其中風風險顯著較低（OR: 0.40，95% CI: 0.20-0.82）。 結論：由本嵌入式病例對照研究可發現，在精神分裂症病人使用抗精神病藥物，可能會增加中風及暫時性腦缺血發生的風險，其他可能增加精神分裂症病人中風風險之因子包括病人本身患有需要使用抗血小板藥物的心血管相關疾病、失智、慢性肺阻塞肺病及高血壓，或是併用非類固醇類消炎藥、valproate及phenytoin。抗精神病藥物所造成之中風風險，在剛開始使用藥物前一個月（尤其前半個月）內最為顯著，但隨著抗精神病藥物使用時間增加而遞減，使用抗精神病藥物達一個月後，其中風風險與非抗精神病藥物使用者相同。第一代抗精神病藥物使用於精神分裂症病人，會增加中風風險，其中以phenothiazine類藥物，如prochlorperazine及chlorpromazine等，造成中風風險最高。第二代抗精神病藥物則不會增加中風風險，其中risperidone可能對中風風險有降低作用。

Background: Antipsychotics have been used in treatment of psychological diseases such as schizophrenia, depression and bipolar, as well as in dementia patients with psychosis and behavioural problems. Recent researches indicated an increasing risk of developing cerebrovascular adverse events due to the use of antipsychotics in patients with dementia and elderly users. The risk of antipsychotics on the occurrence of stroke in schizophrenic patients remains unknown. Therefore, we conducted this study to examine the relationship between antipsychotics and risk of stroke among schizophrenic patients, to provide a reference to clinical medical specialists when treating schizophrenic patients. Objectives: To identify whether the use of antipsychotics increase risk of stroke among newly diagnosed schizophrenic patients, onset of such adverse event, and find out the risk factors that affect the risk of stroke among schizophrenic patients. Methods: A retrospective nested case-control study was conducted by using the data from two sets of longitudinal health insurance database (LHID 2000 and LHID 2005). Newly diagnosed schizophrenic patients (ICD-9 code=295.xx) between 2001 and 2009 without diagnosis of stroke before first diagnosis of schizophrenia were identified to form the study cohort. The date of first diagnosis of schizophrenia was defined as entry date. Patients with diagnosis of stroke or transient ischemic attack (ICD-9 code=430.xx-435.xx) were defined as the cases, and the date of first diagnosis of stroke was taken as index date for the cases. For each case, two controls were randomly selected from the cohort, matched by age (±2 years) and sex, and were assigned an index date of the corresponding case. The study individuals were divided into three groups based on the use of antipsychotic agents before the index date. Current users were those who had prescription period within 7 days before index date, including index date, recent users were defined as patients with the last prescription period ended within 8 days till 30 days before index date, and past users were those with the prescription period not overlapping the last 30 days before index date. Non-users were defined as patients with no use of antipsychotics during the whole follow-up period. Based on the length of antipsychotic use, the current users were divided into four categories (0–15, 16–30, 31–90 and >90 days) to assess the relationship between length of antipsychotic use and risk of stroke. The type of antipsychotic prescriptions were analyzed by variables of single use of first generation antipsychotics, single use of second generation antipsychotics, and combination of first and second generation antipsychotics, as well as individual antipsychotics. Other potential confounders included co-morbidities and co-medication that related to the risk of stroke. For conducting the basic demographic analysis, we use t-test for continuous variables and Chi-square analysis for discrete variables. We used conditional logistic regressions to examine the relationship between antipsychotic use and risk of stroke. Results: There were 9715 individuals in newly diagnosed schizophrenic cohort, we identified 386 cases and 772 controls. There were no differences in age, sex and length of follow-up period between two groups, but cases tend to had higher prevalence of co-morbidities and co-medication than controls. After adjustment of potential confounders, current users had increased risk of stroke compared with non-users (adjusted OR: 1.94, 95% CI: 1.11-1.39). Other risk factors included dementia, COPD, hypertension, and concomitant use of anti-platelet drugs, NSAIDs, valproate and phenytoin. The risk of stroke had a temporal relationship with antipsychotic use, which concentrated in the first month of use, with the adjusted OR for the length of antipsychotics used less than 15 days and 16-30 days 9.41 (95% CI: 3.08-28.71) and 6.90 (95% CI: 1.09-43.69), respectively. The effect decreased with the length of antipsychotics used, that risk of stroke did not different from non-user when duration greater than 30 days. Among antipsychotics, risk of stroke increased when patients used first generation antipsychotics alone or combined with second generation antipsychotics, with adjusted OR 2.75 (95% CI: 1.34-5.64) and 2.37 (95% CI: 1.20-4.68), respectively. Uses of chlorpromazine (OR: 9.26, 95% CI: 1.54-55.69), flupentixol (OR: 3.39, 95% CI: 1.22-9.42), haloperidol (OR: 4.74, 95% CI: 1.45-15.47) and prochlorperazine (OR: 9.83, 95% CI: 1.49-65.02) had increased risk of stroke. Uses of risperidone had relative protective effect in stroke, with OR 0.40 (95% CI: 0.20-0.82). Conclusion: Use of antipsychotics is associated with risk of stroke and transient ischemic attack among schizophrenic patients, with the greatest risk in first month after started antipsychotics. Other risk factors that increase the risk of stroke in schizophrenic patients include diseases that take anti-platelet drugs as standard therapy, dementia, COPD, hypertension, and concomitant use of NSAIDs, valproate and phenytoin. First generation antipsychotics, especially chlorpromazine, flupentixol, haloperidol and prochlorperazine, have increased risk of stroke, with the greatest risk when using prochlorperazine and chlorpromazine. Risperidone seems to have protective effect against stroke, but need further studies to prove it.