RESUMEN
PURPOSE: This study aimed to investigate in a multicenter cohort study the radicality of colorectal cancer resections, to assess the oncosurgical quality of colorectal specimens, and to compare the performance between centers. METHODS: One German and nine Swiss hospitals agreed to prospectively register all patients with primary colorectal cancer resected between September 2001 and June 2005. The median number of eligible patients with one primary tumor included per center was 95 (range 12-204). RESULTS: The following variations of median values or percentages between centers were found: length of bowel specimen 20-39 cm (25.8 cm), maximum height of mesocolon 6.5-12.5 cm (9.0 cm), number of examined lymph nodes 9-24 (16), distance to nearer bowel resection margin in colon cancer 4.8-12 cm (7 cm), and in rectal cancer 2-3 cm (2.5 cm), central ligation of major artery 40-97 % (71 %), blood loss 200-500 ml (300 ml), need for perioperative blood transfusion 5-40 % (19 %), tumor opened during mobilization 0-11 % (5 %), T4-tumors not en-bloc resected 0-33 % (4 %), inadvertent perforation of mesocolon/mesorectum 0-8 % (4 %), no-touch isolation technique 36-86 % (67 %), abdominoperineal resection for rectal cancer 0-30 % (17 %), rectal cancer specimen with circumferential margin ≤1 mm 0-19 % (10 %), in-hospital mortality 0-6 % (2 %), anastomotic leak or intra-abdominal abscess 0-17 % (7 %), re-operation 0-17 % (8 %). CONCLUSION: In colorectal cancer, surgery considerable variations between different centers were found with regard to radicality and oncosurgical quality, suggesting a potential for targeted improvement of surgical technique.