The evaluation of LV function is an important assessment for the noncardiologist physician, with great clinical utility and can be performed subjectively without any specific measurements. In the present study, we observed increasingly high rates of concordance between the LV subjective evaluations performed by students and echocardiographers throughout the training session, with a high concordance rate (96.9%) even at the second stage of the evaluation. Similarly, Melamed et al. reported that intensivists were able to distinguish normal LV function from altered function in 86% of their adult patients . In addition, Vignon et al. demonstrated a correct assessment of the LV function in 92% of the patients , and Manasia et al. showed accurate determinations in 84% of the patients . Studying pediatric patients, Spurney et al. assessed the concordance of subjective LV function performed by pediatricians and cardiologists and demonstrated a strong correlation (96%) . Longjohn et al. also demonstrated a good interobserver agreement (κ = 0.87) between pediatricians and echocardiographers in subjectively differentiating normal LV function from reduced LV function .
The clinical utility of EF measurements for the management of critically ill patients has been established in previous pediatric studies [25, 26]. We obtained a strong correlation between the quantitative EF measurements performed by the students and the echocardiographers in all three phases of training, with a mean of the absolute difference that was consistently less than ten EF points. Similar results were also reported by Pershad et al. in a pediatric population, in which the mean difference in the shortening fraction of the LV measurements was 4.4% after a brief theoretical/practical training course .
Regarding the subjective analysis of RV function, we observed that the concordance rate between students and echocardiographers was high in the 3 stages of evaluation; unfortunately, in our sample, there were only 2 children with RV dysfunction. Among the students, only 14 (87.5%) felt they were capable of performing this assessment; therefore, our data do not allow us to make conclusions about the empowerment and learning curve of students on the subjective analysis of RV function.
The measurement of CI is widely accepted and used as a hemodynamic monitoring tool in ICUs . Minimally invasive devices, which are based primarily on pulse pressure analyses, are being developed for the measurement of CI in the adult patient population [29–31]. However, due to technical difficulties related to the unique physiological characteristics of children, these methods do not present an effective option in the arsenal of hemodynamic monitoring devices for the pediatric population. Although the calculation of a CI value via a transthoracic echocardiogram requires measurements of the LVOT diameter and the aortic VTI and is therefore considered to be technically challenging, we obtained a good correlation after the third evaluation with a mean difference of only 0.56 L/min/m2 between the measurements performed by the students and the echocardiographers. Evaluation of the CI by trained physicians in our study was performed in an innovative manner and demonstrated relevant results. This assessment might provide a new option in the arsenal for the hemodynamic monitoring of critically ill children, which emphasizes the need for extensive training that we observed in our study.
Respiratory changes in the diameter of the IVC in patients on mechanical ventilation are related to the individual’s volemic state; hypovolemic patients present greater effects of positive pressure ventilation in venous return and a greater variation in the diameter of the IVC [21, 32]. In our study, we observed a strong correlation at the second evaluation (16 training examinations), with a 90% concordance in the patient classification as volume-responsive (dIVC greater than 18%) or volume-unresponsive (dIVC less than 18%). In a pediatric study, Pershad et al. reported a strong correlation between the IVC diameter measurements performed by students in training compared with the measurements obtained by experienced echocardiographers; however, neither the dIVC nor another dynamic measurement of the IVC was evaluated during this study .
With respect to the evaluation of PE, we observed a decrease in concordance at the third evaluation. This reduced concordance rate could be explained by the small number of PEs that we had at this evaluation phase. Three cases were classified as mild by the echocardiographers but regarded by the students as absent. This misclassification could be attributed to the difficulties in distinguishing mild effusion from epicardial fat; most likely this type of PE would not have yielded any clinical implications for patient management. The literature reveals data that indicate that non-echocardiographer physicians can identify moderate and severe PE and cases of cardiac tamponade [22, 33]. We believe that PE diagnosis could be performed by the ICU physician, especially in cases of large volumes; however, the limited number of cases, together with the absence of severe PE and cardiac tamponade, does not allow us to confirm this statement.
The analysis of valve regurgitation does not appear to be of great clinical importance in the emergency evaluation of critically ill children because of the rare occurrence of serious regurgitation with hemodynamic repercussions in children who do not present congenital heart disease. However, we performed the evaluation of MR and TR with the primary objective of addressing the student’s ability to identify TR, for a further measurement of the pulmonary pressure through the TR. We observed a good concordance rate in identifying MR after 16 training examinations; however, we did not see major improvements in the identification of TR during the training. Another relevant observation is that only 75% of students felt capable of detecting and grading valve regurgitation after the training. Therefore, although our work is the first to report the identification and graduation of MR and TR by pediatricians, our data do not allow us to state that this training was accomplished in a successful manner, and further analysis in this direction is required.
Royse et al. trained 100 students using multimedia presentations and practical lessons in healthy volunteers, without training the students on actual patients . The students’ assessment was performed by analyzing videos and yielded concordance rates of 95% for the evaluation of volemia and 99% for the analysis of LV function. Recently, Tanzola et al. reported on the training of 10 anesthesiology residents via theoretical sessions and “hands-on” sessions using normal subjects. The training included subjective analyses of RV and LV function, volume assessment and pericardial disease, and the students’ evaluations were performed using 50 multiple-choice questions and videos. The results were presented as positive by improving post-test scores; however, there was no practical evaluation of the training . The examination of ICU or ER patients in loco presents technical difficulties related to the capture of echocardiographic images due to factors specifically related to critically ill patients, such as the use of mechanical ventilation, limited mobility and difficulty in positioning the patient for the examination; these issues were not considered in the above-mentioned studies.
The present study, compared with previous studies, is noteworthy because it was conducted in the ICU and the examinations were performed on critically ill patients and provided practical training to a greater number of professionals. Our study is the first to establish the learning curve throughout the training of the pediatricians. The present study demonstrates that pediatric intensivists and emergency physicians are capable of performing the focused bedside echocardiography approach in critically ill children and emphasizes that whenever doubt exists regarding the presence of anatomical or functional abnormalities, the case should always be discussed with and reassessed by a pediatric echocardiographer [13, 14].
The primary limitations of the present study are as follows: 1) the echocardiography examinations were performed on selected patients based on a predefined training schedule; thus, the study was conducted without including all of the echocardiographic possible abnormalities (e.g., there were no cases of cardiac tamponade or severe PE) during the training and evaluation stages, which limited our assessment of the learning curve regarding these echocardiographic alterations; 2) the design and sampling protocol of the study did not facilitate the evaluation of certain factors, such as accuracy and precision.