The results of this paper show that good quality PW Doppler flow spectrum of each cardiac valve can be obtained on standing unsedated goats like in most other animal species. The effects of the physiological factors such as body size , breed , age , lactation  and pregnancy [18, 23], which are known to affect PW Doppler measurements were not investigated in the present study. A colour-flow Doppler echocardiography performed each day before each PW-Doppler echocardioghraphic protocol did not reveal any abnormality. Particularly, any suspicion of interatrial communication or patent foramen ovale such as reported in 9 healthy Saanen goats  could be excluded.
To assess a pulmonary spectrum flow in this study, the transducer had to be advanced far cranially under the forelimb, which required an assistant pulling the right forelimb forward and upward during the examination. This procedure was not well tolerated by all of the investigated animals, and implies than the investigated goat is fully accustomed to being handled, which is not always the case in caprine field practice. Moreover, at this level, the cardiac window often appeared to be narrow as previously reported in goats [16, 18]. But when a heart base view of sufficient quality was obtained, the recorded pulmonary outflow velocity spectrum was generally speaking of good quality, suggesting a good alignment of the transducer with the pulmonary blood flow, as previously reported in horses [10, 24]. On the contrary, the tilted 2D right parasternal long axis four chambers view was in most of the goats of very good quality, but the obtained flows at the level of the mitral and tricuspid valves were generally underwhelming, as described in horses . Moreover, a true left parasternal apical view as used in dogs to measure aortic, mitral and sometimes tricuspid flows [10, 25], was not possible to obtain in goats because of the presence of gas in the reticulo-rumen .
The repeatability of the PW Doppler measurements in our study was poor, especially for Acc and Dec slopes of all intra-cardiac blood flows and, on the contrary to what could be expected on the basis of the good quality of the pulmonary flow spectrum, for all pulmonary flow measurements. PW Doppler measurements have been already described as poorly repeatable in other species [10, 26, 27]. Moreover, slopes of intra-cardiac blood flows are generally the less repeatable Doppler parameters . Aortic and pulmonary flow measurements were previously often reported as repeatable measurements [25, 28, 29], except in one study on horses in which they were poorly repeatable . The lack of repeatability of the pulmonary flow in the studied goats could be explained by the difficulty to obtain a good quality 2D image of the right ventricular outflow tract, then to place the sample volume day to day exactly at the good location. Recordings of the pulmonary flow from a left parasternal cranial long axis right ventricular outflow view or from a left parasternal short axis view with aorta and pulmonary artery, as used in small domestic animals [10, 25, 29], has not been investigated in this study but appeared to be difficult to obtain in goats.
The variability of the PW Doppler measurements in our study was quite high, but was in accordance with previous studies in dogs, cats and horses [7, 25, 26, 29]. Independently of the valve where the blood flow was measured, the within-day and between-day CV ranged from 4.68% to 41.53%, but most of these were inferior to 20% (Tables 1, 2, 3, 4). The main source of variability could be the poor alignment of the transducer with the direction of the blood flow [10, 26, 27]. Measurements of the Acc or Dec slopes and times from the different blood flows were the most variable parameters. This is in agreement with previous studies in horses  and in dogs .
In the studied goats, the mean values of the mitral velocity spectrum obtained from a tilted left parasternal long axis four chamber view were significantly different from those obtained from a tilted right parasternal long axis four chamber view, except for all parameters of the A peak and for ET of the E peak. Moreover, Emax and Emean were lower when the measurements were performed from the right side than from the left side, which suggests that in goats, the mitral flow should be interrogated from the left rather than from the right hemi thorax. This result is in agreement with previous studies on other domestic animals since, to record the mitral flow, a tilted left parasternal long axis four chambers view is recommended in horses , and a left parasternal apical view is recommended in sheep  and in dogs .
The Emax/Amax ratio is a parameter often used to evaluate the left ventricular diastolic function in man [2, 30]. Independently of the side from which it is measured, the Emax/Amax ratio of the mitral flow was rather similar to the tricuspid flow Emax/Amax. On the contrary to what was reported in sheep , in most goats of this study, Emaxwas higher than Amax for both mitral and tricuspid flows, and only one goat had Emax/Amax < 1 for mitral flow obtained from the right side. The same was observed in 8 of 40 investigated healthy horses  and was explained as a more accurate alignment of the transducer with the A wave of atrial contraction than with the E wave of the early rapid ventricular filling. Measurements of E peak and A peak seemed also to depend on HR. In goats as in sheep, it has been reported than the A peak is closer to the E peak with increasing HR, and when HR was more than 120 beats/min, fusion of the two peak can occur .
Measurements of aortic velocity spectrum are very interesting because they allow assessing left ventricular SV and CO . In this study, except for PEP and PEP/ET, the aortic velocity spectrum measurements obtained from the tilted left parasternal long axis five chambers view were significantly higher than those obtained from a tilted right parasternal long axis five chambers view. This is in agreement with the results obtained in horses , and could be explained by a better alignment between the transducer and the blood flow from the left hemi thorax. This could also explain why previous measurements of aortic flow parameters reported in Swedish goats and obtained from the right side  were lower than those obtained in this present study.