Background: Management of obstructing colorectal cancers has been a challenging clinical problem for surgeons. Little 
is known of the reasons for the dismal survival. The aim of this study is to clarify significant risk factors related to the 
development of obstruction and to identify determinants of survival. 
Patients and Methods: Over a 5-year interval from April 1997 through April 2002, 63 consecutive patients with colorectal carcinoma were surgically treated. Patients undergoing surgery for malignant colorectal obstruction (n= 28) were compared with those undergoing elective surgical treatment (n= 35). Case notes were prospectively collected for information on demographic, clinical, operative, and pathologic variables; as well as follow-up for the detection of local recurrence or distant metastasis. A univariate and multivariate Cox proportional hazards model were then constructed to compare both groups and to examine the effects of these variables on survival. 
Results: The most common site of obstruction was the sigmoid colon (32.1%). Curative resection was possible in 45 patients [14 obstructed (50%) and 31 non-obstructed (88.6%), P= 0.0007]. Anastomatic leakage occurred in 5 [3 obstructed (14.3%) and 2 non-obstructed (5.7%), P= 0.56] and operative deaths occurred in 9 [6 obstructed (21.4%) and 3 non-obstructed (8.4%), P= 0.68]. Multivariate analysis of factors related to an obstructing tumour were patient's age and Dukes' stage. Median followup period was 44.5 months. Local recurrence occurred after curative resection in 9 patients [4 obstructed (28.5%) and 5 nonobstructed (16.1%), P= 0.42] and metastatic disease in 11 [5 obstructed (35.7%) and 6 non-obstructed (19.3%), P= 0.28]. The overall 5-year survival rates were 52% [30% obstructed and 68% non-obstructed, P=0.001]. Multivariate analysis showed that tumour perforation, histologic grade, curative resection, tumour location and tumour fixity were the significant determinants of survival. 
Conclusion: Multivariate analysis of risk factors predicting the development of obstruction were: patient's age and Dukes'stage. Whereas, multivariate analysis of significant variables related to survival were: tumour perforation, histologic grade, curative resection, tumour location and tumour fixity. Large bowel obstruction was a factor which did not influence the prognosis significantly.