Editor’s Note: This is the next installment of a continuing series highlighting researchers sponsored by the Emergency Medicine Foundation (EMF) and illustrating the impact EMF-funded research is having on emergency medicine.

Study Title: Agreement of different ultrasound measures of fluid responsiveness with each other and with clinical judgment in critically ill patients

Authors: Sara Crager, MD, Ricky Amii, MD, Caleb Canders, MD, Daniel Weingrow, DO, Stephanie Tseeng, MD, and Alan Chiem, MD

Researcher Information: Dr. Crager received her MD from Yale School of Medicine. She completed residency training in emergency medicine in 2015 at the Olive View-UCLA Medical Center and is currently a critical care fellow in the Department of Anesthesia at Stanford Hospital. She is interested in studying the use of multi-point ultrasound assessments to evaluate volume status in critically ill patients. Ultimately, she hopes to develop protocols for ultrasound assessment of volume status in undifferentiated shock patients.

Dr. Amii is an assistant clinical professor at UCLA-Olive View Medical Center.
Dr. Canders is an ultrasound fellow at UCLA-Olive View Medical Center.
Dr. Weingrow is an assistant Clinical professor at UCLA-Ronald Regan Medical Center.
Dr. Tseeng is an assistant clinical professor, UCLA-Ronald Regan Medical Center.
Dr. Chiem is an assistant clinical professor, UCLA-Olive View Medical Center.

Study Background: The ability to accurately assess fluid responsiveness (FRes) is central to guiding fluid management in critically ill patients. Evidence is accumulating that both inadequate and excessive fluid resuscitation are associated with increased morbidity and mortality, and there is a growing body of literature underscoring the importance of achieving this balance early in the course of treatment. It is becoming increasingly important to have access to point-of-care tools that facilitate tailoring fluid management to the needs of individual patients. As the utility of central venous pressure is being increasingly called into question, ultrasound is becoming the major modality used by emergency physicians to assess FRes, with multiple different ultrasound approaches currently in use. While left ventricular outflow tract velocity time integral (LVOTVTI) is the most validated of these, it’s difficult to perform correctly, and significant controversy remains as to which ultrasound modality is best for emergency physicians to use and even whether clinical judgment by itself is just as effective for accurately predicting FRes. In this study, we assessed the agreement of various ultrasound approaches to FRes prediction with one another and with the clinical judgment of experienced physicians.

Even when performed by expert sonographers, the variety of ultrasound measurements currently in use to assess FRes show only poor to fair agreement with one another.

Study Design: We evaluated 29 patients admitted to the medical intensive care unit. Patients for whom there was a clinical decision being made about fluid management by the clinical team were enrolled. We asked attending physicians caring for the patients to predict whether the patients would be fluid responsive (i.e.: whether they would expect the patient to increase their cardiac output after a fluid bolus) based on their clinical judgment and how certain they were of this assessment. We obtained four different ultrasound measurements at the time of clinical assessment: inferior vena cava respiratory variation (IVC), LVOTVTI, carotid artery maximal velocity (carotid Vmax), and femoral artery maximal velocity (femoral Vmax). The four sonographers performing the scans were all ultrasound fellowship-trained emergency physicians and were blinded to the admitting diagnosis and the physician’s clinical assessment of FRes.

Results: Overall, there was poor agreement among the different FRes assessments evaluated. The only measures that showed fair agreement were LVOTVTI and carotid Vmax (κ = 0.34) as well as carotid Vmax and femoral Vmax (κ = 0.26). IVC respiratory variation showed poor agreement with all other measures studied (κ = 0 to -0.01). Clinician judgment showed poor agreement with all ultrasound measures studied (κ = -0.15 to 0.18). Clinicians rated their judgment as “highly certain” 45 percent of the time, “somewhat certain” 35 percent of the time, and “not at all certain” 20 percent of the time.

Projected Impact: This research will hopefully contribute to our understanding of how ultrasound may be used by emergency physicians to support clinical judgment when making complex decisions regarding fluid management in septic and critically ill patients. Even when performed by expert sonographers, the variety of ultrasound measurements currently in use to assess FRes show only poor to fair agreement with one another. Despite experienced clinicians expressing a high or moderate degree of certainty in their clinical assessment of FRes 80 percent of the time, there was poor agreement of these assessments with all ultrasound measures of FRes. Given the increasing recognition of the importance of accurate assessment of FRes in critically ill patients, future studies will need to clarify which measures should be used by emergency physicians when assessing FRes as these data suggest that the same patient could be deemed fluid responsive or nonresponsive depending on the method of assessment used.

Dr. Crager is a critical care fellow in the Department of Anesthesia at Stanford Hospital in Stanford, California.