Abstract Introduction Both under-and over-resuscitation are associated with worse outcomes in septic shock. In the FRESH trial, a protocol of stroke volume (SV)-guided fluid resuscitation improved outcomes. We implemented an adapted FRESH trial resuscitation protocol and sought to evaluate protocol adherence. Methods We implemented a SV-guided fluid protocol in Michigan Medicine ICUs in March 2024. For hypotensive patients, the protocol advises a dynamic assessment using a non-invasive bioreactance device (Starling) to measure SV change after a passive leg raise (PLR) or test fluid bolus (250 or 500 mL), followed by a fluid bolus for SV change ≥10%. We randomly sampled patients who had the bioreactance device ordered and/or used between 4/1/2024 to 3/31/2025. Patients were included in the study if they had a dynamic assessment performed. We evaluated patient characteristics and protocol adherence, defined as receiving fluid as directed by the protocol. Blood products were counted towards adherence for fluid-responsive patients but not for fluid-unresponsive patients. Using multivariable logistic regression, we assessed the association of protocol adherence with 30-day mortality, adjusted for the following variables measured at protocol initiation: age, sex, body mass index, history of heart failure, history of CKD, ICU type, SOFA score, respiratory support, vasopressor support, and fluid volume in prior 12 hours. Results Of 325 patients who had the bioreactance device ordered and/or used, we randomly sampled 207 (63.7%), of whom 168 (81.2%) received at least one dynamic assessment of fluid-responsiveness and were included in the study. Patients were a median 65 years (IQR 55, 72), 55.0% male, 56.0% with confirmed sepsis, and had high illness severity median SOFA score 7 (IQR 5,10); 63.1% receiving vasopressors at protocol initiation. Patients received a median of 1L (IQR: 250, 2500) in the 12 hours preceding protocol initiation. On first dynamic assessment, 80 (47.6%) patients were fluid-responsive. PLR was the most common assessment (91, 54.2%), followed by 250 mL fluid bolus (38, 22.6 %). 39 (23.2%) patients had more than one assessment of fluid-responsiveness. Treatment adhered to the fluid-resuscitation protocol in 107 (63.7%) patients (range 50% to 72% across ICUs). Only 2 factors were associated with protocol adherence: type of ICU and fluid-responsiveness. In adjusted analysis, protocol adherence was associated with lower odds of 30-day mortality (40.9% vs 59.2%, aOR 0.41, 95%CI: 0.17-0.99). Conclusions Resuscitation protocols that tailor fluid volume based on dynamic measures of fluid-responsiveness can improve patient outcomes. Efforts to promote adherence are likely needed to maximize benefit. This abstract is funded by: not applicable
Munroe et al. (Fri,) studied this question.