In Reply McCaw et al propose using restricted mean survival time (RMST)1 as an alternative to the Cox proportional hazards models we used to examine the influence of body composition on survival in patients with breast cancer.2 While we agree that RMST presents an alternative method of interpreting data with less restrictions about model fit, there are several reasons why we chose the approach used. First, our article is an epidemiological study examining body composition risk factors for survival, not a randomized clinical trial. Thus, it is unlikely that a clinician would ever inform a patient with low muscle or excess adiposity that she had 6 fewer months to live compared with a patient with normal weight or normal muscle, especially because it is unknown if losing adiposity or gaining muscle will change a patient’s survival. Mean differences generated from RMST may be more appropriate for patients who need to make informed decisions about treatment. Our study is meant to generate new hypotheses for potential future clinical trials and to help clinicians understand the underlying mechanisms regarding the role of body composition to survival. Thus, presenting relative risks for this purpose is informative. Additionally, other studies3,4 on this topic have used statistical approaches similar to the approach we used so that results can easily be compared.
Caan BJ, Quesenberry CP, Lee C. Body Composition and Overall Survival in Patients With Nonmetastatic Breast Cancer—Reply. JAMA Oncol. 2019;5(1):115–116. doi:10.1001/jamaoncol.2018.5299
Coronavirus Resource Center
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: