评估腹腔镜脾切除术（LS）治疗免疫性血小板减少性紫癜（ITP）不同阶段的手术效果，探讨LS的学习曲线问题。

回顾性分析2003年5月至2010年3月期间同一主刀医生完成的105例LS治疗ITP的临床资料。按入院顺序分为A、B、C 3组，每组35例，比较3组手术时间、术中出血量、术后48 h总引流量、并发症发生率及术后住院时间等效果指标。

所有患者无需中转传统手术。线性回归分析显示手术例数与手术时间、术中出血量均呈线性关系（相关系数分别为-0.408和-0.234，P值分别为<0.001和0.016）。3组手术时间分别为（125.0±33.5）min、（111.8±26.3）min和（100.1±25.7）min（P=0.002），术中出血量分别为（95.7±166.0）ml、（64.3±100.8）ml和（38.3±34.3）ml（P=0.007）。两两比较，A组与C组手术时间和术中出血量差异有显著性（P值分别为0.001和0.002）。3组术后48小时总引流量、并发症发生率及术后住院时间差异无统计学意义。

腹腔镜脾切除术治疗免疫性血小板减少性紫癜是安全可行的。随着手术例数增加，手术时间和术中出血量逐渐减少。学习曲线约为35例，可达到较熟练程度。

To evaluate the outcomes of laparoscopic splenectomy (LS) at different stages in patients with immune thrombocytopenic purpura(ITP), and to define the learning curve of LS.

The clinical data of 105 cases of LS performed for ITP by same surgeon between May 2003 and March 2010 were analyzed retrospectively. The cases were divided into group A, B and C with 35 cases in each group according to the sequence of hospitalization. The surgical outcomes of three group, including operative time, estimated intraoperative blood loss, estimated 48-hour volumes of postoperative drainage, major morbidity, and postoperative hospitalization time, were compared subsequently.

No convertion to traditional operation was needed in all the patients. There were linear relationships between operative number and operative time or estimated intraoperative blood loss (correlation coefficients were -0.408 and -0.234, respectively, P<0.001 and 0.016, respectively). Operative times of three groups were (125.0±33.5) min, (111.8±26.3) min, and (100.1±25.7) min, respectively (P=0.002). Estimated intraoperative blood losses were (95.7±166.0) ml, (64.3±100.8) ml, and (38.3±34.3) ml, respectively(P=0.007). When comparing between two groups, statistical diferences of operative times and estimated intraoperative blood losses were found between goup A and C (P values were 0.001 and 0.002, respectively). There were no significant diferences of estimated 48-hour volumes of postoperative drainage, major morbidity, and postoperative hospitalization time among three groups.

LS for ITP is safe and feasible. Operative time and estimated intraoperative blood loss decrease significantly with the increase of operative number. The learning curve is about 35 cases to reach proficiency.