Comments on the paper "c-erbB-2 expression and nuclear pleomorphism in canine mammary tumors"

Response to the Comments of K. Metze and R.L. Adam on the paper "c-erbB-2 expression and nuclear pleomorphism in canine mammary tumors"

“…In our paper (12), we did not find any correlation between survival and c-erbB-2 expression or nuclear pleomorphism (P > 0.05), which is commented on by Metze and Adam (1). Really, the statistical method (Spearman rank correlation) is not the most adequate to evaluate the survival in this case as commented by Metze and Adam.…”

Is There a Reliable Minimum Number of Lymph Nodes for T1 and T2 Colon Cancer?

“…We think that prognostic factor studies with surprising results should always be checked for pitfalls [2][3][4][5]. Dichotomizing a continuous or scaled variable causes several important problems.…”

“…Dichotomizing a continuous or scaled variable causes several important problems. The definition of a cutpoint implies a threshold effect with loss of information, thus lowering efficiency, reliability, and test power, and increasing type I and type II errors [4][5][6][7][8][9][10]. The so-called optimal cutpoint method by maximizing a measure of difference, as used by Maggard et al [1], may lead to spuriously significant results by overestimating the difference in the outcome variable between the groups.…”