Clinical applications of wall-stress analysis in the pediatric intensive care unit

Objective: To determine whether serial, noninvasive assessment of afterload, contractility, and Doppler-derived cardiac output reliably detects variations in cardiac function in unstable pediatric patients. Design: Prospective, blinded clinical trial. Setting: The pediatric intensive care unit at Massachusetts General Hospital. Patients: Fourteen critically ill pediatric patients. Interventions: Pediatric patients meeting criteria for hemodynamic instability underwent serial echocardiograms every 6 hrs until they met exit criteria, generating 75 studies. Measurements and Main Results: Shortening fraction, cardiac index (Cl), end-systolic wall stress (ESWS), and corrected velocity of circumferential shortening (Vcfc) were measured in each patient. Data points were plotted as a stress-velocity relationship, compared with published normal values, then correlated with changes in vital signs and pharmacologic interventions. Fourteen of 16 patients who were enrolled completed the study. A strong negative correlation between ESWS and Vcfc was confirmed (p < .001). As an internal measure of validity, Vcfc had a strong positive correlation with Cl measurements (p = .012). An increase in dopamine infusion was associated with a fall in ESWS (p = .02), an increase in Vcfc (p = .03), and an increase in the Cl as measured by Doppler (p = .035). The infusion of dopamine above renal perfusion levels moved patients from zones of normal or compensated contractility for afterload on a modified stress-velocity relationship to a zone of high contractility for afterload. Urine output was the only clinical index of cardiac function that had a significant correlation with the echocardiographic indices. Hemodynamically unstable patients followed similar patterns of deterioration and recovery on the modified stress-velocity graph. All surviving patients returned to a normal or compensated zone. Conclusions: Wall-stress analysis of cardiac function is easily and safely performed on mechanically ventilated pediatric patients with the production of consistently high-quality data. For internal validity, Vcfc and Cl measurements were correlated and were strongly positive. Wall-stress indices reliably detected patient deterioration, recovery, and response to changes in dopamine infusion. Patients who failed to return to areas of normal or compensated levels of contractility and afterload did poorly in this study. Noninvasive measures of afferload and contractility appear useful for monitoring cardiac function of critically ill children in an intensive care setting.