目的 利用網狀Meta分析比較不同中藥注射液治療卒中后認知障礙（PSCI）的有效性和安全性。方法 檢索中國學術期刊全文數據庫（CNKI）、維普生物醫(yī)學數據庫（VIP）、萬方數據庫（Wanfang）、PubMed、Web of Science、Cochrane Library數據庫建庫至2024年3月中藥注射液治療PSCI的隨機對照試驗（RCT）文獻，根據納入與排除標準篩選文獻，采用Cochrane偏倚風險評估工具評價文獻質量，采用Stata 16.0軟件進行統(tǒng)計分析。結果 最終納入文獻28篇，涉及患者2 942例，治療方法包括西醫(yī)常規(guī)和10種中藥注射液：丹參多酚酸鹽注射液（DSDF）、復方丹參注射液（FFDS）、銀杏達莫注射液（YXDM）、醒腦靜注射液（XNJ）、銀杏葉注射液（YXY）、舒血寧注射液（SXN）、血塞通注射液（XST）、疏血通注射液（SXT）、丹紅注射液（DH）、天麻素注射液（TMS）。網狀Meta分析結果顯示，在臨床總有效率方面，累積概率排序為：YXY＋西醫(yī)常規(guī)治療組（73.1%）＞SXN＋西醫(yī)常規(guī)治療組（62.9%）＞TMS＋西醫(yī)常規(guī)治療組（61.8%）＞YXDM＋西醫(yī)常規(guī)治療組（59.4%）＞DSDF＋西醫(yī)常規(guī)治療組（58.7%）＞XNJ＋西醫(yī)常規(guī)治療組（58.2%）＞SXT＋西醫(yī)常規(guī)治療組（39.3%）＞DH＋西醫(yī)常規(guī)治療組（36.0%）＞西醫(yī)常規(guī)治療組（0.6%）；在簡易智力狀態(tài)檢查量表（MMSE）評分方面，累積概率排序為：TMS＋西醫(yī)常規(guī)治療組（80.2%）＞YXY＋西醫(yī)常規(guī)治療組（70.4%）＞YXDM＋西醫(yī)常規(guī)治療組（60.7%）＞XNJ＋西醫(yī)常規(guī)治療組（58.8%）＞XST＋西醫(yī)常規(guī)治療組（51.8%）＞SXN＋西醫(yī)常規(guī)治療組（50.2%）＞DSDF＋西醫(yī)常規(guī)治療組（48.2%）>FFDS＋西醫(yī)常規(guī)治療組（43.5%）＞SXT＋西醫(yī)常規(guī)治療組（41.6%）＞DH＋西醫(yī)常規(guī)治療組（34.6%）＞西醫(yī)常規(guī)治療組（10.0%）；在蒙特利爾認知評估量表（MoCA）方面，累積概率排序為：SXT＋西醫(yī)常規(guī)治療組（86.0%）＞YXY＋西醫(yī)常規(guī)治療組（83.4%）＞SXN＋西醫(yī)常規(guī)治療組（74.9%）＞TMS＋西醫(yī)常規(guī)治療組（65.1%）＞XNJ＋西醫(yī)常規(guī)治療組（53.9%）＞FFDS＋西醫(yī)常規(guī)治療組（51.0%）＞YXDM＋西醫(yī)常規(guī)治療組（49.1%）＞XST＋西醫(yī)常規(guī)治療組（42.0%）＞DH＋西醫(yī)常規(guī)治療組（23.2%）＞DSDF＋西醫(yī)常規(guī)治療組（18.5%）＞西醫(yī)常規(guī)治療組（2.9%）。在日常生活能力量表（ADL）評分方面，累積概率排序為：TMS＋西醫(yī)常規(guī)治療組（85.4%）＞YXDM＋西醫(yī)常規(guī)治療組（63.7%）＞DH＋西醫(yī)常規(guī)治療組（59.7%）＞SXT＋西醫(yī)常規(guī)治療組（59.5%）＞XNJ＋西醫(yī)常規(guī)治療組（59.3%）＞XST＋西醫(yī)常規(guī)治療組（52.3%）＞DSDF＋西醫(yī)常規(guī)治療組（46.4%）＞西醫(yī)常規(guī)治療組（14.9%）＞SXN＋西醫(yī)常規(guī)治療組（8.8%）。10項研究報道了不良反應，主要涉及消化系統(tǒng)。結論 中藥注射液聯(lián)合西醫(yī)常規(guī)治療可提高PSCI的臨床療效，MMSE評分、MoCA評分和ADL評分表明YXY和TMS優(yōu)勢顯著，然而，由于納入研究的質量受限且存在發(fā)表偏倚，當前結論仍需通過更為嚴謹的高質量研究加以驗證。此舉將有助于為制定中藥注射液干預PSCI的診療方案提供更為堅實的循證醫(yī)學依據。

Objective To compare the efficacy and safety of Chinese medicine injection in the treatment of post-stroke cognitive impairment (PSCI) by network Meta-analysis. Methods The literature about RCTs of Chinese medicine injection in the treatment of PSCI was retrieved from databases such as CNKI, VIP, Wanfang Data, PubMed, Web of Science, Cochrane Library from the establishment of databases to March 2024. The literature was screened according to inclusion and exclusion criteria, and the Cochrane risk assessment tool was used for literature quality assessment. Stata 16.0 was used for statistical analysis. Results A total of 28 articles were included, involving 2 942 patients. The measures included 10 Chinese medicines injections, covering Danshen Duofen Injection (DSDF), Fufang Danshen Injection (FFDS), Yinxing Damo Injection (YXDM), Xingnaojing Injection (XNJ), Yinxingye Injection (YXY), Shuxuening Injection (SXN), Xuesaitong Injection (XST), Shuxuetong Injection (SXT), Danhong Injection (DH), Tianmasu Injection (TMS). In terms of improving the rate of clinical total effective, results of network Meta analysis showed that the rank of cumulative probability was: YXY+control subject (CS) (73.1%) > SXN+CS (62.9%) > TMS+CS (61.8%) > YXDM+CS (59.4%) > DSDF+CS (58.7%) > XNJ+CS (58.2%) > SXT+CS (39.3%) > DH+CS (36.0%) > CS (0.6%). On the part of MMSE, results of network Meta analysis showed that the rank of cumulative probability was: TMS+CS (80.2%) > YXY+CS (70.4%) > YXDM+CS (60.7%) > XNJ+CS (58.8%) > XST+CS (51.8%) > SXN+CS (50.2%) > DSDF+CS (48.2%) > FFDS+CS (43.5%) > SXT+CS (41.6%) > DH+CS (34.6%) > CS (10.0%). On the part of MoCA, results of network Meta analysis showed that the rank of cumulative probability was: SXT+CS (86.0%) > YXY+CS (83.4%) > SXN+CS (74.9%) > TMS+CS (65.1%) > XNJ+CS (53.9%) > FFDS+CS (51.0%) > YXDM+CS (49.1%) > XST+CS (42.0%) > DH+CS (23.2%) > DSDF+CS (18.5%) > CS (2.9%). On the part of ADL, results of network Meta analysis showed that the rank of cumulative probability was: TMS+CS (85.4%) > YXDM+CS (63.7%) > DH+CS (59.7%) > SXT+CS (59.5%) > XNJ+CS (59.3%) > XST+CS (52.3%) > DSDF+CS (46.4%) > CS (14.9%) > SXN+CS (8.8%). Ten studies reported adverse reactions, mainly involving digestive system. Conclusion Chinese medicine injection combined with C can improve the clinical efficacy in the treatment of PSCI. The MMSE score, MoCA score, and ADL score show that YXY and TMS have significant advantages in the treatment of PSCI. However, due to the limited quality and publication bias of the included studies, the current conclusions still need to be verified by more rigorous and high-quality studies. This will help to provide a more solid evidence-based medical basis for the development of treatment protocols for PSCI with herbal injections.

國家自然科學基金青年科學基金項目（82104808）