Accurate assessment of left ventricular ejection fraction (LVEF) forms a routine part of daily echocardiographic practice. Currently, there is sufficient evidence to support the notion that assessment of LVEF by real-time three-dimensional echocardiography (RT3DE) should be considered the method of choice in routine practice. Nevertheless, implementation of RT3DE is still limited. There may be issues with, for example, work flow and training of sonographers. At our center, we have used a stepwise approach, aiming at the implementation of routine application of RT3DE for assessment of LVEF.
First, space was generated in the daily work flow of the echocardiography laboratory. Specific echocardiographic protocols were developed on the basis of several documents developed by the American Society of Echocardiography and the European Association of Echocardiography, each one describing the required acquisitions for a specific indication. In the first 6 months after the introduction of these protocols, the mean number of images per echocardiographic study decreased significantly. The time needed for acquisition and analysis of a real-time three-dimensional echocardiographic data set for LVEF assessment takes (on average) <3 min. This time became available by decreasing the number of two-dimensional images.
Second, local validation of LVEF by RT3DE was accomplished. Two sonographers were trained for LVEF assessment by RT3DE by the vendor of the ultrasound systems used in our clinic (2-day course). Following this, during a period of 1 month, they each acquired three-dimensional volumetric data sets in 15 patients. After this initial training, both sonographers were required to acquire ≥10 full-volume RT3DE data sets in patients scheduled for nuclear LVEF assessment. Measurement variability of LVEF assessment in the 25 patients in whom RT3DE and nuclear studies were available was 6.4 ± 4.1% (mean difference for LVEF 3.9 ± 2.3%; mean LVEF 64 ± 9%), and the correlation between RT3DE and nuclear LVEF was highly significant (R = 0.90, P < .001).
Third, LVEF assessment by RT3DE was implemented in daily practice by training the other five sonographers at our center and using the two RT3DE-experienced sonographers as the “gold standard.” Each novice RT3DE sonographer acquired 10 data sets that were also analyzed by one of the experienced RT3DE sonographers. Mean measurement variability was 4.9 ± 2.4% (mean difference for LVEF 2.9 ± 3.3%; mean LVEF 61 ± 15%). Correlation analysis showed good agreement between the measurements performed by the more experienced and novice RT3DE sonographers (R = 0.94, P < .001).
We believe that our stepwise approach, starting by generating space in the daily work flow, followed by local validation and finally application in daily clinical practice, may be successfully used to implement the routine use of RT3DE for LVEF assessment.