The title of the study that I'll be presenting today is the comparison of strain and conventional echocardiography in predicting ICU needs in MIS-C. My name is Esther Kim. I'm a third year pediatric critical care fellow at Children's National Hospital in Washington, DC. My research interests focus mainly on point-of-care ultrasound and its uses in improving patient care. I completed my medical degree at the University of Virginia in Charlottesville, Virginia, followed by a pediatric residency at Children's National Hospital. I have no financial disclosures or conflicts of interest. Some learning objectives for my presentation today. First, we'll review the clinical definition and critical care needs of MIS-C patients. Then we'll describe conventional and strain echocardiography. And then finally, we'll compare these two modalities in predicting the needs for ICU-level therapies among this patient population. MIS-C, or multisystem inflammatory syndrome in children, is defined by the CDC as pediatric patients less than 21 years of age with fever for greater than 24 hours, lab evidence of inflammation, severe illness requiring hospitalization, multisystem involvement of two or more organs with no alternative diagnosis. And finally, these patients also have to have recent or current infection or exposure to COVID-19. This disease was first noted in April of 2020. And in the U.S. as of October 31st, 2022, there have been over 9,000 confirmed cases, 74 deaths. The median age of the patient population is about nine with 60% of the patients being male and 56% of the patients being Hispanic or African-American. In the context of pediatric critical care, many patients present with fluid refractory shock, often cardiogenic in nature. A large proportion of MIS-C patients require admission to the ICU, with some studies citing rates of up to 80%. About a third of MIS-C patients require non-invasive positive pressure ventilation, about 20% require mechanical ventilation, and up to half or more of patients require vasoactive support. We thought it would be important to distinguish which of these high-risk patients will and will not develop shock from cardiac dysfunction and require intensive care support. To try to detect this early, we looked at strain echo as a possible solution. Strain is defined as the percent deformation due to an applying force, delta L over L. In echocardiography, the naturally occurring speckle pattern of the myocardium is used to measure cardiac strain. So cardiac strain is defined as the local shortening, local thickening, and local lengthening of the myocardium as a measure of regional left ventricular function. It's a quantitative value. It gives us magnitude and direction of myocardial deformation and motion. This short clip shows the measurement of global longitudinal strain in the four-chamber view using speckle tracking echocardiography. When we think of conventional echocardiography, the typical ventricular function measurements that we think of are ejection fraction and shortening fraction. These values, while useful, are limited in that they're dependent on loading conditions and image quality. Because strain looks at regional movements, it is often more sensitive in detecting early or regional myocardial dysfunction. The aim of our study was to compare the relative strengths of conventional echocardiography and strain echocardiography in predicting the needs of ICU-level therapies in MIS-C patients. Our study design was a retrospective observational cohort study. The patient population included pediatric patients admitted to a large quaternary care pediatric hospital with a diagnosis of MIS-C between March 2020 and January 2022. The first echo during hospitalization was analyzed for both left ventricular ejection fraction and global longitudinal strain. Abnormal EF was defined as less than 55%, and abnormal strain was defined as greater than or equal to negative 17.2%. We started with 188 total patients admitted with a diagnosis of MIS-C during the study period. Of these, 128 patients had both conventional and strain echo available. We then further excluded patients whose echoes were done after the initiation of vasoactives, as vasoactive support has been shown to affect both ejection fraction and strain values. And then finally, our study included the remaining 90 patients. The median age of our patients was about 7.3 years. 66% of our patients were male, and then the race ethnicity breakdown was about 11% white, non-Hispanic, 50% black, 35% Hispanic, and 3% other. This chart depicts the clinical course of the patients. About half of the 90 patients were admitted to the ICU. 37% required ICU level care. 37% required ICU level care. 30% required positive pressure ventilation, and 28% required vasoactive support. The median total hospital length of stay among our patients was about nine days, and there were no mortalities. In terms of our results, we first looked at conventional and strain echo findings among all patients. Looking at the left circle, we found that 79% of patients had preserved ejection fraction, while on the right, we see that only 46% of patients had preserved strain. When we broke down the patients between these two groups, we found that the median strain in the normal ejection fraction group was also normal at 19.2%, negative 19.2%. And then among the patients with abnormal ejection fraction, the median strain in this group was also abnormal, negative 11.7. So this would be expected. Normal ejection fraction had a median strain that was normal, and the abnormal ejection fraction group had a median strain that was abnormal. Interestingly though, when we divided the patients into patients with normal strain and abnormal strain, the median ejection fraction, while significantly different, were normal in both groups. We then further divided the patients categorically by their clinical requirements. This left graph depicts the ejection fraction, and the right graph depicts the strain. Patients were divided by those who didn't require IC-level care and those who did, patients who didn't require positive pressure ventilation and those who did, and patients who didn't require vasoactive support and those who did. The dotted lines are the threshold for normal ejection fraction and normal strain. When we look at the left graph, what we found was that while the median ejection fraction was significantly different in each of the groups, they were again, both above the normal threshold, greater than or equal to 55%, regardless of whether they did or did not require IC-level care, positive pressure ventilation, or vasoactive support. Meanwhile, in this right graph, we see that the median strain was significantly different between the two groups, but more importantly, the median strain was abnormal among patients who required IC-level care, required positive pressure ventilation, and required vasoactive support. Sensitivity analysis of these groups showed that strain was superior to ejection fraction in identifying those patients who would need IC-level care, positive pressure ventilation, and vasoactives. Negative predictive values of strain was also superior to ejection fraction for all these outcomes. ROC curves were done for these groups as well, adjusted for age and gender. For IC-level care, strain had a higher AUC compared to ejection fraction. The case was the same for positive pressure ventilation and for vasoactive support. Our conclusions from our study are twofold. One, that strain echo is superior to conventional echo in detecting cardiac dysfunction in MIS-C patients. And two, compared to ejection fraction, the presence of abnormal strain better identifies patients who will need IC-level therapies. We did have several limitations to our study, mainly that it's a retrospective single-center study. There was a subset of patients who were excluded due to inadequate images as strain was collected retrospectively. Additionally, there were 38 patients excluded from the study as they were already on vasoactives at the time of the echo. In practice, these patients had already declared themselves as requiring escalation of care. But perhaps if echoes were conducted early on, they can help guide further care in this patient population as well. Finally, our study only looked at one single measure of strain, global longitudinal strain in the four-chamber view. Inclusion of other values of strain, particularly detecting diastolic strain, may add to the utility of strain echo in future studies. A prospective study that includes early strain echo to help identify patients at high risk for requiring IC-level therapies. We also fully acknowledge that we are seeing much less MIS-C as COVID cases are declining, but these principles are applicable for other populations that would benefit from early detection of myocardial dysfunction, particularly oncology patients receiving cardiotoxic treatments and acute rejection and heart cancer. These are several articles that are referenced in my talk today. Thank you very much.

strain echocardiography

conventional echocardiography

ICU needs

multisystem inflammatory syndrome in children

MIS-C