Noradrenaline dose cutoffs to characterise the severity of cardiovascular failure: Data-based development and external validation.

Abstract

Background: The vasopressor dose needed is a common measure to assess the severity of cardiovascular failure, but there is no consensus on the ranges of vasopressor doses determining different levels of cardiovascular support. We aimed to identify cutoffs for determining low, intermediate and high doses of noradrenaline (norepinephrine), the primary vasopressor used in intensive care, based on association with hospital mortality.

Methods: We conducted a binational registry study to determine cutoffs between low, intermediate and high noradrenaline doses. We required the cutoffs to be statistically rational and practical (rounded to the first decimal and easy to remember), and to result in increasing mortality with increasing doses. The highest noradrenaline dose in the first 24 h after intensive care unit (ICU) admission was used. The cutoffs were developed using data from 8079 ICU patients treated in the ICU at Kuopio University Hospital, Finland, between 2013 and 2019. Subsequently, the cutoffs were validated in the eICU database, including 39,007 ICU admissions to 29 ICUs in the United States of America in 2014-2015. The log-rank statistic, with the Contal and O'Quigley method, was used to determine the cutoffs resulting in the most significant split between the noradrenaline dose groups with regard to hospital mortality.

Results: The two most prominent peaks in the log-rank statistic corresponded to noradrenaline doses 0.20 and 0.44 μg/kg/min. Accordingly, we determined three dose ranges: low (<0.2 μg/kg/min), intermediate (0.2-0.4 μg/kg/min) and high (>0.4 μg/kg/min). Mortality increased, whereas the number of patients decreased consistently with increasing noradrenaline doses in both cohorts. In the development cohort, hospital mortality was 6.5% in the group without noradrenaline administered and 14.0%, 26.4% and 40.2%, respectively, in the low-dose, intermediate-dose and high-dose groups. Compared to patients who received no noradrenaline, the hazard ratio for in-hospital death was 1.4 for the low-dose group, 4.0 for the intermediate-dose group and 7.5 for the high-dose group in the validation cohort (p < .001).

Conclusions: The highest noradrenaline dose is a useful measure for quantifying circulatory failure. Cutoffs 0.2 and 0.4 μg/kg/min seem to be suitable for defining low, intermediate and high doses.