[Skip to Navigation]
Views 122
Citations 0
Comment & Response
April 2018

Lymph Node Yield as Quality Metric for Clinically N0 Oral Cancer—Reply

Author Affiliations
  • 1Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, New York
  • 2Department of Biostatistics, Memorial Sloan Kettering Cancer Center, New York, New York
JAMA Otolaryngol Head Neck Surg. 2018;144(4):386-387. doi:10.1001/jamaoto.2017.3336

In Reply We thank Dr Judson for his comments on our article.1 Dr Judson claimed that we selected the median number as our lymph node threshold and suggested that the cutoff was chosen arbitrarily. However, this was incorrect. As described in the article,1 we used restricted cubic spline, in particular, maximally selected rank statistics to choose the cutoff value.2 Therefore, the threshold 24 was selected objectively, not arbitrarily or based on the median lymph node count. We also examined model discriminative ability using bootstrapping bias-adjusted concordance probability (c-index) and calibration against the Kaplan-Meier estimated survival. Internal validation was performed using bootstrapping with 200 replications. The study by no means implied that half of the neck dissections in the cohort were inadequate; rather, it indicated that the cutoff of 24 could best differentiate the patient survival in the data set.

Add or change institution