Naslov
Extent of oncological resection in laparascopic vs. open colorectal surgery: meta-analysis

Autori
Korolija, Dragan ; Tadić, Stjepan ; Šimić, Diana

Izvornik
Langenbecks archives of surgery (1435-2443) 387 (2003), 9/10; 366-371

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, članak, znanstveni

Ključne riječi
colorectal surgery. cancer. laparoscopy. meta-analysis

Sažetak
Laparoscopy is not yet wide accepted in cancer colorectal surgery. This study investigated whether the extent of anatomical resection in laparoscopic colorectal surgery is equivalent to that in the open colorectal surgery. We examined reports on the results of laparoscopic or open colorectal procedures published between 1990 and 1999 and selected 35 with data on lymph node count and distal margin clearance (total of 3935 patients). There were 16 comparative studies, 6 open series, and 13 laparoscopic series of patients. Several meta-analytic models were used to evaluate the difference between open and laparoscopic surgery, including multiple-outcome random-effects models that account for correlation between multiple-outcomes. More lymph nodes were extracted laparoscopically (0.3-2.14 lymph nodes more) ; however, only for the fixed-effects single-outcome model this difference was statistically significant. The average distal margin clearance was 4.6 cm with the laparoscopic approach and 5.3 with the open approach Accounting for random variation among the studies, the estimate of this difference obtained from meta-analysis was 0.64-1.19 cm, favoring the open approach ; this difference was statistically significant. The estimated within-study correlation between lymph node count and distal margin clearance was 0.26 with the open and 0.19 with the laparoscopic procedure. The estimated within-study correlation between outcomes of laparoscopic and open procedures was 0.86 for the lymph node count and 0.96 for the distal margin clearance. This meta-analysis shows that laparoscopic approach is as adequate as the conventional approach. Estimates of difference between the two therapeutic options obtained from random-effects models were more precise than those from fixed-effects models.

Izvorni jezik
Engleski

Znanstvena područja
Kliničke medicinske znanosti

POVEZANOST RADA