A total of 2740 patients fulfilled the inclusion criteria of which 2464 had definitive surgical procedure in form of segmental colon resection (89.9%). Stoma was constructed in 276 patients (10%) while SEMS was used only in eight patients (Figure 1). Mean age was 72 years (median 74 with IQR 65-81). More than half of the cohort were females (1530/2740, 55.8%). Most of the tumours were located in Caecum then ascending colon, hepatic flexure and least in the transverse colon (Table 1).
Multivariate analyses showed the following results:
Post-operative complications’ rates
Age, ASA and tumour location in transverse colon were the most significant risk factors associated with post-operative complications in the group of patients who had colectomy (Table 1).
A closer look at this cohort reveals that two of these factors influenced also the severity of complications: older patients (OR 1.032 CI 95 % [1.009-1.055] p= 0.005) and patients with high ASA score (OR 1.61 CI 95% [1.422-1.830] p<0.001) had higher rates of grade 3b and above post-operative complications as classified by Clavian-Demartines-Dindo classification (supplementary material to support table 1).
In addition to age and ASA score, patients with high alcohol consumption (more than 21 unit per week) had higher risk of post-operative surgical complications (OR 2.516 CI 95% [1.520-4.165] p<0.001) such as intra-abdominal septic complications OR 2.516 CI 95% [1.520- 4.165] p<0.001 and wound dehiscence OR 2.935 CI 95% [1.317- 6.543] p=0.009. On the other hand, the effect of high alcohol consumption on post-operative medical complications such as sepsis, cardio-pulmonary and thrombo-embolic complications was not significant as shown in the supplementary material to support table 2).
Post-operative medical complications were mainly associated with higher age (OR 1.034 CI 95% [1.023-1.045] p<0.001), higher ASA scores (OR 1.61 CI 95% [1.422-1.830], p<0.001), higher Charlson comorbidity index (CCI) (OR 1.533 CI 95% [1.072- 2.193] p=0.019) and laparotomy (OR 2.562 CI 95% [1.202-5.463] p=0.015). Women had less risk of post-operative medical complications compared to men (OR 0.797 CI 95% [0.647-0.982] p=0.033) (Table 2). Smokers had two times higher risk of Intra-abdominal septic complications (IASCs) compared to non-smokers (OR 1.930 CI 95% [1.047-3.558] p=0.035). Tumour location in transverse colon doubled the risk of IASCs OR 2.261 CI 95% [1.425241-3.5891] p=0.001 (supplementary material to support table 1).
Post-operative mortality
Women had less 30-day post-operative mortality (OR 0.618 CI 95% [0.461- 0.828] p=0.001) compared to men as shown in table 3. Tumour location in transverse colon (OR 1.866 CI 95% [1.283-2.712] p=0.001), tumour perforation (OR 2.275 CI 95% [1.429-3.619] p= 0.001) and metastasis (OR 1.658 CI 95% [1.207-2.274] p=0.002) increased the risk of 30-day post-operative mortality. Age and ASA score were still significant factors in 30-day post-operative mortality after adjusting for confounding factors as shown in table 3. This pattern was also noticed in 90-day post-operative mortality in which age, ASA score, CCI and tumour perforation were the most significant risk factors in addition to metastasis and BMI (supplementary material to support table 3).
Definitive surgical procedures versus de-functioning procedures
The characteristics of patients who had definitive surgical procedures i.e., resections were then compared to those who had only de-functioning procedures i.e., stoma construction or SEMS without oncological resection. Patients who had de-functioning procedures were mostly males (OR 1.485 CI 95% [1.117-1.974] p=0.007), with higher ASA scores (OR 0.7349 CI 95% [0.608-0.889] p= 0.001), higher CCI (OR 0.659 CI 95% [0.477- 0.910] p=0.011), on pre-operative chemotherapy (OR 0.137 CI 95% [0.0698-0.268] p<0.001) and/or had pre-operative tumour perforation (OR 0.132 CI 95% [0.095- 0.182] p<0.001) (Table 4). However, de-functioning procedure did not protect these patients from the high risk of post-operative complications nor post-operative 90-day mortality as shown in table 5.
Surgeon’s grade of specialization and access to abdominal cavity
Surgeon’s grade of specialization was not a significant factor to affect the post-operative outcome in right colon resections as shown in the tables 1,2 and 3. Neither tumour stage nor type of resection played a role in affecting post-operative outcome in emergency colon resections.
Open access to abdomen i.e., laparotomy was used in more than 90% of cases while laparoscopic approach was attempted in the rest of the cohort with a conversion rate of approximately 50%. The rate of laparoscopic surgery increased when we selected patients operated after 2010 up to 16.3%. At the same time, the rate of conversion increased from 129/240 (53.8%) to 113/190 (59.5%). However, access to abdominal cavity had no significant effect on post-operative outcome after adjusting for covariates. Factors that affected the choice of laparoscopic approach were ASA (OR 0.43 CI 95% 0.24-0.62] p<0.001) and tumour location (OR 0.02 CI 95% [0.006-0.033] p=0.043). Tumours in Caecum and ascending colon were most likely to be treated with laparoscopic access than tumours in transverse colon and hepatic flexure.
Comparison between fit to fight patients and patients with pre-operative compromised status
Patients with compromised status had higher risk of overall post-operative complications especially medical complication in form of sepsis and cardio-pulmonary complications. Patients with compromised status had also higher risk of 30 days post-operative mortality as well as 90 days post-operative mortality (OR 0.260 CI 95% [0.184- 0.368] p<0.001).
Age, ASA scores, CCI, open access to abdomen, cancer location in transverse colon were still the most significant factors contributing to overall post-operative complications (supplementary material to support table 4). After adjusting for confounders, age and open access to abdomen continued to be significant risk factors of overall post-operative complications (OR 1.534 CI 95% [1.222-1.928] p<0.001) and (OR 2.422 CI 95% [1.195- 4.909] p= 0.014) respectively.
The effect of time lapse on post-operative outcome
The rates of post-operative overall, medical and surgical complications were not affected by the year of intervention but 30-day and 90 days post-operative mortality were significantly reduced over the time of the study (OR 0.943 CI 95% [0.922-0.965], p<0.001) and OR 0.953 CI 95% [0.934- 0.972] p<0.001) respectively. Over the years covered by the study, there was a significant increase in the rates of stoma construction (OR 1.270 CI 95% [1.23-1.315], p<0.001), laparoscopic interventions (OR 1.235 CI 95% [1.174- 1.299] p<0.001) and the rate of conversion during laparoscopy (OR 1.284 CI 95% [1.219-1.351], p<0.001).