CRT-Guided Resuscitation Improves Hierarchical Outcomes in Early Septic Shock
Key Takeaways
- CRT-PHR was associated with a favorable 28-day hierarchical outcome versus usual care in early septic shock.
- Early resuscitation under CRT-PHR involved less fluid, more dobutamine use, lower central venous pressure and lactate, and more frequent normalization of capillary refill time by 6 hours.
- Twenty-eight-day mortality was similar, prespecified subgroup analyses did not show effect modification, and suspected unexpected serious adverse reactions were few without an apparent between-group difference.
The investigator-generated, multicenter, unblinded randomized clinical trial was conducted in 86 ICUs across 19 countries. Adults were enrolled within 4 hours of meeting septic shock criteria, and 1467 of 1501 randomized patients entered the primary analysis. The analyzed groups included 720 assigned to CRT-PHR and 747 assigned to usual care. Eligibility required suspected or confirmed infection, hyperlactatemia, and norepinephrine after at least 1000 mL of intravenous fluid, focusing the trial on early septic shock resuscitation.
In the capillary refill time protocol, CRT was measured by firm pressure on the ventral distal phalanx with a glass microscope slide, and more than 3 seconds was abnormal. The algorithm then used pulse pressure and diastolic arterial pressure, gave fluid boluses only when fluid responsiveness was present, and titrated norepinephrine for low diastolic pressure. If CRT remained abnormal, clinicians moved to basic echocardiography, additional fluid challenges, a higher MAP test in chronic hypertension, and a low-dose dobutamine test. By 6 hours, the intervention group had received less resuscitation fluid, 595 vs 847 mL, and had more dobutamine use, 12.3% vs 5.3%. Central venous pressure was 9.1 vs 9.8 mm Hg, lactate was 3.2 vs 3.5 mmol/L, and CRT normalized more often in the intervention group.
The primary outcome ranked 28-day death first, followed by duration of vital support and then hospital stay, with treatment effects analyzed by a stratified win ratio. CRT-PHR produced 131,131 wins, or 48.9%, versus 112,787, or 42.1%, for usual care, yielding a win ratio of 1.16 with a 95% CI of 1.02 to 1.33 and P=.04. Investigators attributed the primary result mainly to shorter duration of vital support. Twenty-eight-day mortality was 26.5% vs 26.6%, with an HR of 0.99, a 95% CI of 0.81 to 1.21, and P=.91. Vital support-free days were higher at 16.5 vs 15.4, hospital stay did not clearly differ, and the separation was not in mortality alone.
The trial was unblinded, and subgroup and tertiary analyses were exploratory. Implementation included mandatory training and weekly auditing, and protocol deviations and violations occurred in 111 of 720 patients, 15%, and 44 of 720, 6%, respectively. Five suspected unexpected serious adverse reactions were recorded without an apparent between-group difference, and none was considered likely related to the protocol. One dobutamine test was stopped for hypotension that reversed after discontinuation, and some DAP augmentation, MAP tests, and dobutamine tests were skipped for preemptive safety reasons. The primary result favored CRT-PHR on the hierarchical composite, alongside early differences in process measures.