Background: The most important determinant of survival in patients with colon cancer is the presence or absence of regional lymph node metastases. This factor is consistently associated with long-term and disease-specific survival. Cumulative summation of differences (CUSUM) charts can help to discriminate abnormalities that cannot be explained by the general variability of a process. We used CUSUM charts to analyse the quality of nodal analysis in colon cancer and to use a population-registry cancer database to estimate the optimal number of lymph nodes for adequate prognostic analysis.
Methods: This was a multicentre population-registry cancer study from January 2004 to December 2007. We used these data to produce the different CUSUM curves, focusing on the main variables. To calculate survival, we used the Kaplan-Meier method.
Results: In this study, we examined 548 patients. The CUSUM curves were calculated for overall mortality, specific mortality, and recurrence according to (1) the number of lymph nodes analysed and affected and (2) compared the ratio of the number of lymph nodes affected to the number analysed. Finally, the lymph node ratio was compared to the overall survival CUSUM curve.
Discussion: This CUSUM control chart analysis reinforces the unquestionable importance of analysing at least 12 lymph nodes in patients with colon cancer in order to accurately estimate their prognosis. However, our findings indicate that the analysis of at least 20 lymph nodes is a more appropriate cutoff point for accomplishing the demanding objective of diagnosing a high-quality prognosis in colon cancer patients.
Keywords: Colon cancer; Lymph nodes; Quality control; Survival.