In Reply Drs Zhou and Tang comment that the survival curves crossed each other in our study,1 and thus they questioned whether the proportionality assumption was met. Although the 1-year survival curves in the Kaplan-Meier plot overlapped at some points, the lines were generally distinct and there was no actual line crossing except at the very end, when there were few patients left in follow-up. In a diagnostic graph of log(−log[survival]) vs log of survival time, there was visible crossing of lines only in the very beginning of the log-time axis. Thus, overall, the time proportionality assumption seemed to hold well. In addition, in a Cox regression model including the evaluation of proportional hazards using the ASSESS statement (SAS proc phreg), no evidence of proportional hazards assumption violation was observed (P > .05). Further testing applying a Cox regression model including evaluation of proportional hazards using an interaction test with log(time) did not reveal a significant interaction (P > .05). In all, no significant deviation from the proportional hazards assumption was noted in visual or statistical evaluation of survival curves or resulting hazards.
Bellingan G, Ranieri VM, Karvonen MK. Intravenous Interferon β-1a for Severe ARDS—Reply. JAMA. 2020;324(5):515–516. doi:10.1001/jama.2020.7992
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: