THE USE OF M-MODE ECHOCARDIOGRAPHY IN DETERMINING CARDIAC OUTPUT IN DOGS WITH NORMAL, LOW, AND HIGH OUTPUT STATES: COMPARISON TO THERMODILUTION METHOD     

Clarke E.  Atkins  DVM  Michael B.  Curtis  DVM  PhD  Sheila M.  McGurik  DVM  PhD  Mark D.  Kittleson  DVM  PhD  Takashi  Sato  DVM  MS  Patti S.  Snyder  DVM  MS 《Veterinary radiology & ultrasound》1992,33(5):297-304

The utility of M-mode echocardiographic determination of left ventricular volume and cardiac output was evaluated by comparing four commonly used formulae for ventricular volume, derived from end-systolic (ESD) and end-diastolic (EDD) ventricular dimensions. Cardiac index (CI) was calculated by the following formulae, where HR = heart rate and BSA = body surface area (M²): 1. (EDD³- ESD³) * HR/BSA; 2. (0.85 * EDD³ - 1.2 *ESD³) * HR/BSA; 3. (EDD^3.16 - ESD^3.16) * HR/BSA; and 4. {[7/(2.4 + EDD)] * [EDD³] - [7/(2.4 + ESD)] * [ESD³)}* HR/BSA. Cardiac index, obtained invasively by thermodilution, was compared to values determined echocardiographically in normal dogs (Group 1) and in three states of altered cardiac function: after positive inotropic stimulation with graded doses (0,5,10, and 15 μg/kg/min) of intravenous dobutamine (Group 2); with experimentally induced heartworm disease (Group 3); and with spontaneous dilated cardiomyopathy (Group 4). Cardiac index, determined echocardiographically, was plotted against CI, determined with thermodilution, and regression equations calculated. In Group 1 (n = 15), a significant relationship (P < 0.01) was observed with each of the four formulae (r²= 0.54-0.68), however, individual estimates of CI were often inaccurate. With inotropic stimulation (Group 2;3 normal dogs, n = 12 data points), the relationship was significant (P < 0.05), but the coefficients of determination were less than in the unstimulated normal dogs (r²= 0.41-0.47). No significant relationship was seen between CI, determined echocardiographically and CI, determined by thermodilution, when overt cardiac disease was present (Groups 3 and 4; n = 7 and 13, r²= 0.13-0.21, respectively). Specific formulae were evaluated for accuracy by comparing their coefficients of determination and their ability to predict mean and individual measured CI. Using the paired Student's t-test, CI determined noninvasively were shown to differ significantly (P < 0.05) between methods and between formulae in many instances. No one formula was found to be consistently more accurate than the others. We conclude that, in normal dogs, while echocardiographic estimation of CI is significantly related to CI measured with thermodilution, individual variation diminishes clinical utility. However, this noninvasive technique should prove useful in comparing populations with minimal or no cardiac dysfunction. Secondly, the relationship is diminished, but not lost, after inotropic stimulation. M-mode echocardiographic estimation of CI offers little promise for accurate assessment of cardiac performance in cardiac disease states. Lastly, although none of the four formulae for CI compared herein were found to be consistently superior to the other three, formulae 3 and 4 appear to be most useful in estimating CI in normal dogs.  相似文献