Tripathi Dhiraj, Graham Catriona, Hayes Peter C
Department of Hepatology, Royal Infirmary of Edinburgh, Wellcome Trust Clinical Research Facility, Western General Hospital, Edinburgh, UK.
Eur J Gastroenterol Hepatol. 2007 Oct;19(10):835-45. doi: 10.1097/MEG.0b013e3282748f07.
BACKGROUND/AIMS: Variceal band ligation (VBL) can reduce the rate of the first variceal by 45-52% compared with beta-blockers (BBs). We performed an updated meta-analysis of nine randomized controlled trials published as full papers, comparing VBL with BB for primary prevention.
Relative risk (RR) was computed using a random effects model. Sensitivity analysis was performed using a fixed effects model. Publication bias was also assessed using funnel plots and the rank correlation test.
In total, 734 patients were studied (356, VBL; 378, BB). The pooled RR favoured VBL for first variceal bleed [0.63; 95% confidence interval (CI), 0.43-0.92] with number needed to treat being 13 (95% CI, 7-33), and for adverse events resulting in treatment withdrawal (0.24; 95% CI, 0.12-0.47) with the corresponding number needed to treat being 10 (95% CI, 6-25). Banding-related bleeding occurred in six patients (fatal in two). No difference was seen in bleeding-related deaths (RR, 0.71; 95% CI, 0.38-1.32), or overall mortality (RR, 1.09; 95% CI, 0.86-1.38). No significant heterogeneity or publication bias was present, and outcomes remained robust after sensitivity analyses.
VBL was superior to BB in preventing the first variceal bleed, with fewer adverse events resulting in treatment discontinuation. Careful attention to technique and patient selection are important to minimize iatrogenic complications with VBL. VBL has a role in patients with poor drug compliance, or tolerance, and in those who bleed on BB therapy.
背景/目的：与β受体阻滞剂（BBs）相比，曲张静脉套扎术（VBL）可使首次静脉曲张出血率降低45%-52%。我们对发表为全文的9项随机对照试验进行了更新的荟萃分析，比较VBL与BB用于一级预防的效果。
采用随机效应模型计算相对危险度（RR）。使用固定效应模型进行敏感性分析。还通过漏斗图和秩相关检验评估发表偏倚。
总共研究了734例患者（356例接受VBL；378例接受BB）。汇总RR显示VBL在预防首次静脉曲张出血方面更具优势[0.63；95%置信区间（CI），0.43-0.92]，需治疗人数为13（95%CI，7-33），在导致治疗中断的不良事件方面（0.24；95%CI，0.12-0.47），相应需治疗人数为10（95%CI，6-25）。6例患者发生套扎相关出血（2例死亡）。在出血相关死亡（RR，0.71；95%CI，0.38-1.32）或总死亡率（RR，1.09；95%CI，0.86-1.38）方面未见差异。不存在显著的异质性或发表偏倚，敏感性分析后结果依然可靠。
VBL在预防首次静脉曲张出血方面优于BB，导致治疗中断的不良事件更少。仔细关注技术和患者选择对于将VBL的医源性并发症降至最低很重要。VBL在药物依从性差或耐受性差的患者以及接受BB治疗时出血的患者中具有作用。
