van Heek N Tjarda, Kuhlmann Koert F D, Scholten Rob J, de Castro Steve M M, Busch Olivier R C, van Gulik Thomas M, Obertop Huug, Gouma Dirk J
Department of Surgery, Academic Medical Center, Meibergdreef 9, 1100 DD Amsterdam, The Netherlands.
Ann Surg. 2005 Dec;242(6):781-8, discussion 788-90. doi: 10.1097/01.sla.0000188462.00249.36.
To evaluate the best available evidence on volume-outcome effect of pancreatic surgery by a systematic review of the existing data and to determine the impact of the ongoing plea for centralization in The Netherlands.
Centralization of pancreatic resection (PR) is still under debate. The reported impact of hospital volume on the mortality rate after PR varies. Since 1994, there has been a continuous plea for centralization of PR in The Netherlands, based on repetitive analysis of the volume-outcome effect.
A systematic search for studies comparing hospital mortality rates after PR between high- and low-volume hospitals was used. Studies were reviewed independently for design features, inclusion and exclusion criteria, cutoff values for high and low volume, and outcome. Primary outcome measure was hospital or 30-day mortality. Data were obtained from the Dutch nationwide registry on the outcome of PR from 1994 to 2004. Hospitals were divided into 4 volume categories based on the number of PRs performed per year. Interventions and their effect on mortality rates and centralization were analyzed.
Twelve observational studies with a total of 19,688 patients were included. The studies were too heterogeneous to allow a meta-analysis; therefore, a qualitative analysis was performed. The relative risk of dying in a high-volume hospital compared with a low-volume hospital was between 0.07 and 0.76, and was inversely proportional to the volume cutoff values arbitrarily defined. In 5 evaluations within a decade, hospital mortality rates were between 13.8% and 16.5% in hospitals with less than 5 PRs per year, whereas hospital mortality rates were between 0% and 3.5% in hospitals with more than 24 PRs per year. Despite the repetitive plea for centralization, no effect was seen. During 2001, 2002, and 2003, 454 of 792 (57.3%) patients underwent surgery in hospitals with a volume of less than 10 PRs per year, compared with 280 of 428 (65.4%) patients between 1994 and 1996.
The data on hospital volume and mortality after PR are too heterogeneous to perform a meta-analysis, but a systematic review shows convincing evidence of an inverse relation between hospital volume and mortality and enforces the plea for centralization. The 10-year lasting plea for centralization among the surgical community did not result in a reduction of the mortality rate after PR or change in the referral pattern in The Netherlands.
通过对现有数据进行系统评价，评估胰腺手术量-结局效应的最佳可用证据，并确定荷兰正在进行的集中化呼吁的影响。
胰腺切除术（PR）的集中化仍在争论中。报道的医院手术量对PR术后死亡率的影响各不相同。自1994年以来，基于对手术量-结局效应的反复分析，荷兰一直在持续呼吁PR的集中化。
采用系统检索比较高手术量和低手术量医院PR术后医院死亡率的研究。对研究的设计特点、纳入和排除标准、高手术量和低手术量的临界值以及结局进行独立审查。主要结局指标是医院或30天死亡率。数据来自荷兰1994年至2004年PR结局的全国性登记处。根据每年进行的PR数量，将医院分为4个手术量类别。分析干预措施及其对死亡率和集中化的影响。
纳入了12项观察性研究，共19688例患者。这些研究异质性太大，无法进行荟萃分析；因此，进行了定性分析。高手术量医院与低手术量医院相比的死亡相对风险在0.07至0.76之间，且与任意定义的手术量临界值成反比。在十年内的5次评估中，每年PR少于5例的医院的医院死亡率在13.8%至16.5%之间，而每年PR超过24例的医院的医院死亡率在0%至3.5%之间。尽管反复呼吁集中化，但未见到效果。在2001年、2002年和2003年，792例患者中有454例（57.3%）在每年手术量少于10例的医院接受手术，而在1994年至1996年期间，428例患者中有280例（65.4%）。
PR术后医院手术量和死亡率的数据异质性太大，无法进行荟萃分析，但系统评价显示了医院手术量与死亡率之间存在反向关系的令人信服的证据，并强化了集中化的呼吁。外科界长达10年的集中化呼吁并未导致PR术后死亡率降低或荷兰转诊模式的改变。
