Agreement of cardiac output measurements by esophageal Doppler and transesophageal echocardiography with intermittent pulmonary artery thermodilution during pharmacologic manipulation of hemodynamics in anesthetized dogs.

Abstract

To compare cardiac output (CO) measurements by transesophageal echocardiography (TEECO) and esophageal Doppler monitor (EDMCO) with pulmonary artery thermodilution (PATDCO) in anesthetized dogs subjected to pharmacological interventions. The effect of treatments on EDM-derived indexes was also investigated. 6 healthy male dogs (10.8 ± 0.7 kg). Dogs were anesthetized with propofol and isoflurane, mechanically ventilated, and monitored with invasive mean arterial pressure (MAP), end-tidal isoflurane concentration (ETISO), PATDCO, TEECO, EDMCO, and EDM-derived indexes. Four treatments were administered to all dogs by randomization. Baseline data were collected before each treatment: (1) dobutamine infusion; (2) esmolol infusion; (3) phenylephrine infusion; and (4) ETISO > 3%. Data were collected after 10-minute stabilization and after 30 minutes of washout between treatments. Statistical tests were pairwise t test, Bland-Altman analysis, Lin's concordance correlation (ρc), and polar plot analysis with P < .05 set as significance. The mean ± SD relative bias (limits of agreement) for TEECO was 0.35 ± 25.2% (-49.1% to 49.8%) and for EDMCO was -27.2 ± 22.5% (-71.4% to 17%) versus PATDCO. The percent error for TEECO and EDMCO was 27.6% and 44.1%, respectively. The ρc value was 0.82 for TEECO and 0.66 for EDMCO. TEECO and EDMCO showed good trending ability. EDM-derived indexes displayed significant changes specific to the drug administered (P < .001). For minimally invasive CO monitoring, TEE may provide more favorable performance than EDM in clinical settings; however, EDM-derived indexes yield valuable hemodynamic information that reliably follows trends in CO, thus supporting critical decision-making in canine patients.