Diagnostic Adjunct Techniques in the Assessment of Hypovolemia: A Prospective Pilot Project

Abstract

IntroductionMeasuring the hypovolemic resuscitation end point remains a critical care challenge. Our project compared clinical hypovolemia (CH) with three diagnostic adjuncts: 1) noninvasive cardiac output monitoring (NICOM), 2) ultrasound (US) static IVC collapsibility (US-IVC), and 3) US dynamic carotid upstroke velocity (US-C). We hypothesized US measures would correlate more closely to CH than NICOM. MethodsAdult trauma/surgical intensive care unit patients were prospectively screened for suspected hypovolemia after acute resuscitation, excluding patients with burns, known heart failure, or severe liver/kidney disease. Adjunct measurements were assessed up to twice a day until clinical improvement. Hypovolemia was defined as: 1) NICOM: ≥10% stroke volume variation with passive leg raise, 2) US-IVC: <2.1 cm and >50% collapsibility (nonventilated) or >18% collapsibility (ventilated), 3) US-C: peak systolic velocity increase 15 cm/s with passive leg raise. Previously unknown cardiac dysfunction seen on US was noted. Observation-level data were analyzed with a Cohen's kappa (κ). Results44 patients (62% male, median age 60) yielded 65 measures. Positive agreement with CH was 47% for NICOM, 37% for US-IVC and 10% for US-C. None of the three adjuncts correlated with CH (κ −0.045 to 0.029). After adjusting for previously unknown cardiac dysfunction present in 10 patients, no adjuncts correlated with CH (κ −0.036 to 0.031). No technique correlated with any other (κ −0.118 to 0.083). ConclusionsNone of the adjunct measurements correlated with CH or each other, highlighting that fluid status assessment remains challenging in critical care. US should assess for right ventricular dysfunction prior to resuscitation.