Abstract：
BACKGROUND: Validity of the laparoscopic approach in pancreatic head lesion remains debatable. This study aims to compare the safety and effectiveness of laparoscopic pancreatoduodenectomy (LPD) and open pancreatoduodenectomy (OPD) and investigate the source of heterogeneity from surgeons\' and patients\' perspectives.
METHODS: We searched PubMed, Cochrane, Embase, and Web of Science for studies published before February 1, 2021. Of 6578 articles, 81 were full-text reviewed. The primary outcome was mortality. Three independent reviewers screened and extracted the data and resolved disagreements by consensus. Studies were evaluated for quality using ROB2.0 and ROBINS-I. According to different study designs, sensitivity and meta-regression analyses were conducted to explore the heterogeneity source. This meta-analyses was also conducted to explore the learning curve\'s heterogeneity. This study was registered with PROSPERO, CRD42021234579.
RESULTS: We analyzed 34 studies involving 46,729 patients (4705 LPD and 42,024 OPD). LPD was associated with lower (P = 0.025) in unmatched studies (P = 0.017). No differences in mortality existed in randomized controlled trials (P = 0.854) and matched studies (P = 0.726). Sensitivity analysis found no significant difference in mortality in elderly patients, patients with pancreatic cancer, and in high- and low-volume hospitals (all P > 0.05). In studies at the early period of LPD (<40 cases), higher mortality (P < 0.001) was found (all P < 0.05).LPD showed non-inferiority in length of stay, complications, and survival outcomes in all analyses.
CONCLUSIONS: In high-volume centers with adequate surgical experience, LPD in selected patients appears to be a valid alternative to LPD with comparable mortality, LOS, complications, and survival outcomes.
METHODS: We searched PubMed, Cochrane, Embase, and Web of Science for studies published before February 1, 2021. Of 6578 articles, 81 were full-text reviewed. The primary outcome was mortality. Three independent reviewers screened and extracted the data and resolved disagreements by consensus. Studies were evaluated for quality using ROB2.0 and ROBINS-I. According to different study designs, sensitivity and meta-regression analyses were conducted to explore the heterogeneity source. This meta-analyses was also conducted to explore the learning curve\'s heterogeneity. This study was registered with PROSPERO, CRD42021234579.
RESULTS: We analyzed 34 studies involving 46,729 patients (4705 LPD and 42,024 OPD). LPD was associated with lower (P = 0.025) in unmatched studies (P = 0.017). No differences in mortality existed in randomized controlled trials (P = 0.854) and matched studies (P = 0.726). Sensitivity analysis found no significant difference in mortality in elderly patients, patients with pancreatic cancer, and in high- and low-volume hospitals (all P > 0.05). In studies at the early period of LPD (<40 cases), higher mortality (P < 0.001) was found (all P < 0.05).LPD showed non-inferiority in length of stay, complications, and survival outcomes in all analyses.
CONCLUSIONS: In high-volume centers with adequate surgical experience, LPD in selected patients appears to be a valid alternative to LPD with comparable mortality, LOS, complications, and survival outcomes.
摘要：
背景：胰头病变腹腔镜入路的有效性尚有争议。本研究旨在比较腹腔镜胰十二指肠切除术（LPD）和开腹胰十二指肠切除术（OPD）的安全性和有效性，并从外科医生和患者的角度探讨异质性的来源。
方法：我们搜索了PubMed，科克伦,Embase,和WebofScience在2021年2月1日之前发表的研究。在6578篇文章中，81篇全文回顾。主要结果是死亡率。三名独立审稿人筛选并提取了数据，并以协商一致方式解决了分歧。使用ROB2.0和ROBINS-I对研究进行了质量评估。根据不同的研究设计，进行敏感性分析和meta回归分析以探索异质性来源.这项荟萃分析也是为了探索学习曲线的异质性。这项研究在PROSPERO注册，CRD42021234579。
结果：我们分析了34项研究，涉及46,729例患者（4705例LPD和42,024例OPD）。在非匹配研究(P=0.017)中,LPD与低(P=0.025)相关。随机对照试验（P=0.854）和配对研究（P=0.726）的死亡率无差异。敏感性分析发现老年患者死亡率无显著差异,胰腺癌患者，以及高容量和低容量医院（均P>0.05）。在LPD早期的研究中（<40例），发现较高的死亡率（P<0.001）（所有P<0.05）。LPD在住院时间上显示非劣，并发症,和所有分析中的生存结果。
结论：在具有足够手术经验的高容量中心，选定患者的LPD似乎是LPD的有效替代方案，死亡率相当，LOS,并发症,和生存结果。
方法：我们搜索了PubMed，科克伦,Embase,和WebofScience在2021年2月1日之前发表的研究。在6578篇文章中，81篇全文回顾。主要结果是死亡率。三名独立审稿人筛选并提取了数据，并以协商一致方式解决了分歧。使用ROB2.0和ROBINS-I对研究进行了质量评估。根据不同的研究设计，进行敏感性分析和meta回归分析以探索异质性来源.这项荟萃分析也是为了探索学习曲线的异质性。这项研究在PROSPERO注册，CRD42021234579。
结果：我们分析了34项研究，涉及46,729例患者（4705例LPD和42,024例OPD）。在非匹配研究(P=0.017)中,LPD与低(P=0.025)相关。随机对照试验（P=0.854）和配对研究（P=0.726）的死亡率无差异。敏感性分析发现老年患者死亡率无显著差异,胰腺癌患者，以及高容量和低容量医院（均P>0.05）。在LPD早期的研究中（<40例），发现较高的死亡率（P<0.001）（所有P<0.05）。LPD在住院时间上显示非劣，并发症,和所有分析中的生存结果。
结论：在具有足够手术经验的高容量中心，选定患者的LPD似乎是LPD的有效替代方案，死亡率相当，LOS,并发症,和生存结果。
