The effect of operative duration on the outcome of colon cancer procedures

Although early reports on laparoscopy-assisted colectomy (LAC) in patients with colon cancer suggested that it reduces perioperative morbidity, its influence on long-term results is unknown. Our study aimed to compare efficacy of LAC and open colectomy (OC) for treatment of non-metastatic colon cancer in terms of tumour recurrence and survival. From November, 1993, to July, 1998, all patients with adenocarcinoma of the colon were assessed for entry in this randomised trial. Adjuvant therapy and postoperative follow-up were the same in both groups. The main endpoint was cancer-related survival. Data were analysed according to the intention-to-treat principle. 219 patients took part in the study (111 LAC group, 108 OC group). Patients in the LAC group recovered faster than those in the OC group, with shorter peristalsis-detection (p=0.001) and oral-intake times (p=0.001), and shorter hospital stays (p=0.005). Morbidity was lower in the LAC group (p=0.001), although LAC did not influence perioperative mortality. Probability of cancer-related survival was higher in the LAC group (p=0.02). The Cox model showed that LAC was independently associated with reduced risk of tumour relapse (hazard ratio 0.39, 95% CI 0.19-0.82), death from any cause (0.48, 0.23-1.01), and death from a cancer-related cause (0.38, 0.16-0.91) compared with OC. This superiority of LAC was due to differences in patients with stage III tumours (p=0.04, p=0.02, and p=0.006, respectively). LAC is more effective than OC for treatment of colon cancer in terms of morbidity, hospital stay, tumour recurrence, and cancer-related survival.

Anastomotic leak (AL) represents a dreaded complication following colorectal surgery, with a prevalence of 1-19 per cent. There remains a lack of consensus regarding factors that may predispose to AL and the relative risks associated with them. The objective was to perform a systematic review of the literature, focusing on the role of preoperative, intraoperative and postoperative factors in the development of colorectal ALs.

In medical research, continuous variables are often converted into categorical variables by grouping values into two or more categories. We consider in detail issues pertaining to creating just two groups, a common approach in clinical research. We argue that the simplicity achieved is gained at a cost; dichotomization may create rather than avoid problems, notably a considerable loss of power and residual confounding. In addition, the use of a data-derived 'optimal' cutpoint leads to serious bias. We illustrate the impact of dichotomization of continuous predictor variables using as a detailed case study a randomized trial in primary biliary cirrhosis. Dichotomization of continuous data is unnecessary for statistical analysis and in particular should not be applied to explanatory variables in regression models. Copyright 2005 John Wiley & Sons, Ltd.