inadequate; rather, it indicated that the cutoff of 24 could best differentiate the patient survival in the data set.On the contrary, the study by Kuo et al 3 performed multiple multivariable analyses to determine the lymph node count cutoff, based on the survival decrease in successive groups compared with the count associated with the best survival. It was unclear how 5 lymph nodes were chosen for nodal grouping and categorization. The authors eventually picked 16 or more as the cutoff, where all groups under this value were significantly different from the reference group. The method used here did not account for multiple comparisons. In addition, the reason for choosing the reference group based on the lowest hazard was unclear, it implied that patients in the reference group had the best outcome and therefore perhaps the lymph node count of the reference group should be used as the threshold value.We do agree with the comments made by Dr Judson stating that, when using lymph node yield as a quality metric, it is crucial to employ the appropriate analysis to determine the actual threshold. It is perhaps more important to recognize that the association observed here and in other studies does not imply causation. Although attempts can be made to generate a practice guideline for lymph node dissection, other clinical and patient factors should also remain the key drivers for estimating patient prognosis. 4