Effects of Norepinephrine and Vasopressin Discontinuation Order in the Recovery Phase of Septic Shock: A Systematic Review and Individual Patient Data Meta‐Analysis

Objective The impact of vasopressin and norepinephrine discontinuation order in the recovery phase of septic shock remains controversial. This systematic review and patient‐level meta‐analysis were performed to determine the impact of vasopressin and norepinephrine discontinuation order on clinically significant outcomes in the recovery phase of septic shock. Methods Cumulative Index to Nursing and Allied Health Literature, Embase, PubMed, and ClinicalTrials.gov were searched from inception through November 2018 for studies comparing outcomes after the discontinuation of vasopressin or norepinephrine in septic shock. Individual patient‐level data were obtained from included studies and combined using a two‐stage meta‐analysis. Results Six studies of low or moderate risk of bias with 957 patients were included. Clinically significant hypotension occurred more frequently when vasopressin was discontinued first compared to norepinephrine (60.7% versus 43.3%, respectively). First discontinuation of norepinephrine compared to vasopressin had lower pooled odds of developing clinically significant hypotension (odds ratio [OR] 0.22, 95% confidence interval [CI] 0.07–0.68, I2 87%). No differences were detected in short‐term mortality (OR 1.12, 95% CI 0.67–1.86, I2 45%), intensive care unit length of stay (mean difference 0.15 day, 95% CI −1.58 to 1.88, I2 21%), or hospital length of stay (mean difference 1.65 days, 95% CI −0.47 to 3.76, I2 0%). Conclusions Discontinuation of norepinephrine prior to vasopressin during the recovery phase of septic shock resulted in less clinically significant hypotension but no difference in mortality or lengths of stay. Larger, prospective studies evaluating the impact of relative vasopressin deficiency and norepinephrine and vasopressin discontinuation order and timing on patient‐centered outcomes are needed.