Hocus POCUS! Parental Quantification of Left-Ventricular Ejection Fraction Using Point of Care Ultrasound: Fiction or Reality?

Abstract

Point of care ultrasound has become increasingly utilized in pediatric settings. The assessment of cardiac function is one such implementation of this. This study aimed to determine the feasibility of parents in acquiring images to assess function using a handheld ultrasound probe and the correlation of fractional shortening measurements by handheld ultrasound with hospital acquired echocardiography. This was a single-center prospective study of parents of pediatric patients admitted to the hospital. Parents underwent a 25-min education session on how to use the handheld ultrasound probe and then were asked to acquire a parasternal short-axis and apical four-chamber image on their own. Acquired images were reviewed by two physicians to determine adequacy of images to assess systolic cardiac function subjectively and objectively. Fractional shortening was measured using parent-acquired images and then compared to recent hospital acquired fractional shortening. A total of 25 parents of 21 patients enrolled and completed the study. Of the enrolled parents, 96% of both parasternal short-axis and apical four-chamber images acquired were deemed appropriate for subjective assessment of systolic function. Inter-reader variability of fractional shortening was moderate between two readers. Correlation of fractional shortening measured from parent-acquired images versus hospital acquired images was moderate. Parents were able to successfully obtain a parasternal short-axis and apical four-chamber image adequate to assess function and quantify fractional shortening after a 25-min education session. This pilot data demonstrate that further exploration of parent-performed point of care cardiac assessment may be warranted.

Introduction

The development and refinement of handheld ultrasound devices have led to an increasing use of point of care ultrasound by care providers throughout the medical system [1, 2]. This includes use by attending physicians, fellow physicians, resident physicians, medical students, nurses, advanced practice nurses, and paramedics across a variety of subspecialties. Physician use of point of care ultrasound in the emergency department and intensive care unit has most been described, although use has been described in other subspecialties such as cardiology, pulmonology, gastroenterology, and nephrology [3–14]. The current experience has demonstrated that not only it is point of care ultrasound feasible but that it may improve care delivery. Use of point of care ultrasound has demonstrated that point of care ultrasound can lead to decreased length of emergency department stay as well as decreased length of hospital stay, particularly in the setting of pediatric appendicitis [5–7, 12, 14].

An application which has seen increased use of point of care ultrasound is in the assessment of cardiac function. Several studies have now demonstrated that assessment of cardiac function using point of care ultrasound is feasible in children, with good agreement between emergency department physicians and cardiologists [3, 4, 8, 10, 13].

A majority of point of care ultrasound studies have focused on the use of point of care ultrasound by medical personnel. Limited data exist in the use of point of care ultrasound in pediatric patients by non-medical personnel. A study by Chen et al. demonstrated that parents were able to obtain measurements of the aortic root in Marfan patients with good correlation to hospital acquired measurements while Voleti et al. demonstrated that school personnel were able to successfully use point of care ultrasound to screen for rheumatic heart disease at the time of routine school health screening in the Republic of Palau [15, 16].

As the population of children with cardiac dysfunction grows and, in a time, when multiple factors have led to the increased implementation of telehealth, the feasibility of parent-performed cardiac function assessment has become more intriguing.

Methods

Study Design

This was a single-center, prospective study that was approved by the Advocate Aurora Healthcare institutional review board. Consent was obtained from the parents of all patients included in this study and consent was obtained from children seven years of age or older.

The study’s first and primary aim was to determine if parents could obtain a parasternal short-axis view of the left ventricle adequate to measure fractional shortening. The second aim was to determine if parents could obtain an apical four-chamber view adequate to measure ejection fraction. The third aim was to quantify inter-reader variability of fractional shortening estimates. The fourth aim was to quantify correlation of fractional shortening from point of care ultrasound to hospital acquired echocardiography. The fifth aim was to assess parent sentiment regarding point of care ultrasound.

Patient Selection

All pediatric (under 18 years of age) inpatients cared for primarily, or consulted on, by the pediatric cardiology service at Advocate Children’s Hospital were eligible for inclusion. Those with single-ventricle congenital heart disease were excluded as were those with left-handed ventricular topology. This was done to ensure that only patients with normal ventricular topology were included in the current study.

Patients deemed appropriate for inclusion were identified and their parents were approached in-person. The study aims and protocol were described in detail and any questions answered. Parents then were given time to decide whether they wished to participate in the study.

Data Collection

Data regarding the patient were collected from the electronic medical record. This included patient age, patient weight, and presence of congenital heart disease. Additionally, if an echocardiogram was obtained in the hospital within 48 h of the study point of care ultrasound, the left-ventricular ejection fraction was recorded.

Parents were given a questionnaire to fill out. Information about the parent including age, highest level of education, whether employed in healthcare were collected. Additionally, three survey items were included to assess parent sentiment regarding the utility of parental point of care ultrasound: (1) Was obtaining the echocardiographic images with minimal training; (2) Do you think you could perform this at home; (3) Do you think this could be helpful for parents and their children?

Point of Care Ultrasound Probe

A Philips Lumify S4-1 probe was utilized for the study. This probe utilizes a transducer able to scan between 4 and 1 megahertz with a 90° field of view. This probe has a 20.2 mm footprint with a scanning depth of 24 cm. The probe was plugged into an Android tablet on which it was paired with a Lumify application. This application allowed for image viewing, enhancement, acquisition, measurement, and post-processing. Pediatric imaging presets were not available. The same probe, tablet, and software were utilized for the entirety of the study. The Philips Lumify probe was provided on a complimentary loan by Philips (Amsterdam, Netherlands). The company was not involved with study design, data collection, data analysis, or manuscript preparation.

Parent Training and Image Acquisition

The same training was provided to each parent who participated in this study. A standardized script was used for the introduction to the probe and acquisition of images to try to standardize these portions as much as possible. This initial verbal education was designed to take 5 min. This was followed by up to five minutes for the parents to explore the probe and the tablet application. Next, the parents were given up to 15 min to obtain images on their children using the probe with assistance from the study team. During this 15-min period, provided assistance was primarily regarding where to find the pertinent imaging windows and how to position the probe to improve the image. After this there was a brief 5-min break and then the parents were given 10 min to acquire the images on their own without assistance. Each parent was asked to independently acquire a short-axis view of the left ventricle from the parasternal window and an apical four-chamber image.

Thus, the entire study encounter after consent was obtained took approximately 45 min. All training sessions were done by a single author (AJ) to try to standardize the sessions as much as possible.

Image Review

All images acquired by parents were reviewed by the two authors, independently. The short-axis image was reviewed subjectively with regard to whether it could be used to measure left-ventricular fractional shortening. The same image then had a fractional shortening calculated by measuring end diastolic and end systolic diameter. These measurements were done prior to obtaining the hospital acquired fractional shortening so that the authors remained blinded to this data.

The apical four-chamber image was then reviewed subjectively with regard to whether it could be used to measure left-ventricular ejection fraction. The results for these assessments were compared with discrepancies bring reviewed by both authors together and consensus achieved. Discrepancies of greater than 5% between fractional shortening were reviewed. Ultimately the average of the fractional shortening obtained by both readers was used for statistical analyses. If multiple images were obtained by the parent for the same view, then the authors discussed and selected what was considered the highest quality image to measure fractional shortening. Measurements were always done utilizing the same image by both authors. For images where the discrepancy in fractional shortening led to review of images, the original measurement was replaced with the measurement obtained on the review.

Statistical Analyses

Continuous variables are described as median and range while descriptive variables are described as absolute count and frequency.

Inter-reader variability between fractional shortening measurements was assessed using a Spearman correlation. A Bland–Altman analysis was also conducted.

Fractional shortening obtained from point of care ultrasound was compared to that obtained from a hospital echocardiogram within 48 h using a Spearman correlation. A Bland–Altman analysis was also conducted.

All statistical analyses were done using SPSS version 23.0. A p value of less than 0.05 was considered statistically significant.

Results

Cohort Information

A total of 25 parents of 21 patients participated in this study. Median parent age was 36 years and ranged from 26 to 49 years. Of these, 17 (68%) reported an educational level of beyond high school, 8 (32%) reported having a medical background, 20 (80%) had prior exposure to echocardiography, and 0 reported having performed an echocardiogram before (Table 1).

Table 1.

Parent and child demographic information

Parent age (years)	36 (26 to 49)
Child age (months)	55 (1 to 204)
Parent level of education
Beyond high school	17 (68%)
High school or less	8 (32%)
Medical background of parent	8 (32%)
Prior exposure to echocardiography	20 (80%)
Performed an echocardiogram before	0 (0%)

The children of these parents had a median age of 9 months and ranged from 1 to 204 months. Of these patients, 13 (62%) were admitted for a cardiac reason with 7 (33% of total children) of these being admitted for cardiac catheterization or surgery. Three (14%) children were admitted for a respiratory reason and the remaining 5 (24%) children were admitted for either gastrointestinal or orthopedic reasons (Table 2).

Table 2.

Reason for admission

Cardiac reason for admission	16 (65%)
Admitted for cardiac catheterization or surgery	9 (36%)
Respiratory reason for admission	4 (15%)
Non-cardiac non-respiratory reason for admission	5 (20%)

Children who were admitted for cardiac catheterization or surgery had the study echocardiogram done after the procedure. No child had a parental echocardiography done during sedation or anesthesia.

Acquisition of Images

A parasternal short-axis image was obtained by 25 (100%) of parents. Of these, 24 (96%) were adequate to assess cardiac function. An apical four-chamber image was obtained by 24 (96%) of parents. Of these, 24 (96%) were adequate to assess cardiac function.

Fractional Shortening

Fractional shortening, as measured by the two readers, had moderate, significant correlation (r = 0.55, p = 0.02) between the two readers. The mean difference of fractional shortening measurements between the two readers was 6.1 (95% confidence interval of − 1 to 12).

The average of the fractional shortening measured by the two readers had moderate, significant correlation with fractional shortening from a formal, inpatient echocardiogram performed within 48 h of the parent-performed echocardiogram (r = 0.62, p < 0.01) (Fig. 1).

Fig. 1.

Scatter plot demonstrating correlation between hospital acquired and parent acquired fractional shortening

Parent Questionnaire

All 25 (100%) parents responded that acquiring the images was easy with minimal training. All 25 (100%) parents responded that felt they could perform such echocardiography at home. All 25 (100%) parents responded that they felt this tool could be helpful for parents.

Discussion

This pilot study demonstrates that parents with no previous ultrasound experience can be taught how to obtain parasternal short-axis and apical four-chamber images of the left ventricle using point of care ultrasound. Furthermore, this can be done in children with a history of congenital malformations of the heart and with only brief training. This is only the second such study demonstrating the use of point of care ultrasound by parents.

Not only was there a high rate of acquired images with adequate quality to quantify left-ventricular fractional shortening but fractional shortening could be measured with low inter-reader variability and with moderate correlation to hospital acquired fractional shortening.

These data are particularly of note when considering the relatively short training period and use of a probe with off-the-shelf presets not necessarily chosen with pediatric patients in mind.

While much data have now been published regarding point of care ultrasound in the pediatric emergency medicine and pediatric critical care, these data are largely regarding point of care ultrasound utilized by clinicians [3–14]. Very few previous studies, to the authors’ knowledge, have been published regarding point of care ultrasound utilized by parents. Chen et al. demonstrated the ability of parents to obtain aortic root measurements in children with Marfan syndrome as part of a telehealth initiative. Aortic root measurements differed by a median of 3.4% and were not significantly different from measurements obtained by hospital acquired echocardiography. Voleti et al. demonstrated that those without previous echocardiography experience could be trained to screen for rheumatic heart disease in children during pre-existing school health screenings in the Republic of Palau [15, 16].

But what are the practical implications of such findings? While outside of the specific feasibility question of this study, it is important to discuss implications. Some families may have to drive over 84 miles to receive subspecialty cardiac care. For those with lower socioeconomic status, such a distance can be quite substantial in the potential absence of having access to a car, gasoline prices, and lack of public transportation. Effects of distance on access to pediatric cardiac care have been demonstrated by several studies [17]. Those with public insurance and higher deprivation index have been documented to be more likely to be able to complete pediatric cardiology visits, with up to 40% of new referrals not even being completed (Warren). Those with public insurance have also been demonstrated to be more likely to have to travel longer distances due to restrictions on centers at which they can receive care, specifically with respect to state borders [18, 19]. More recently, global events such as the coronavirus pandemic and the war in Ukraine have highlighted events of greater magnitude that may limit access to care for those across all socioeconomic strata.

As such, solutions have included scaling up virtual telehealth visits. Such visits can be limited by the ability to do a physical examination or diagnostic testing [2]. Devices that facilitate remote auscultation and vital sign measurement are already available. Indeed, some single-ventricle interstage programs are already utilizing such devices, routinely.

Point of care ultrasound can be an additive tool in the telehealth armamentarium. While not necessary for all patients, those with concerning systolic left-ventricular dysfunction in whom medications are being actively titrated and in whom there are ongoing issues with clinical symptoms, such a device could help efficiently titrate medications effectively in the outpatient setting while limiting clinic visits and, potentially, hospitalizations. This may also be a helpful tool in the event that symptoms change or progress, allowing for parents to assess left-ventricular function. Critics will cite cost as a barrier but costs have been cited for nearly every new medical device. As point of care ultrasound develops and matures as a technology, prices will likely drop. Additionally, if future analyses demonstrate that such deployment of point of care ultrasound helps decrease outpatient visit and inpatient hospitalizations then related cost savings may outweigh the cost of the device itself.

Potential clinical applications, related benefits, and related risks are not explored by the current study but are worth discussion as they establish direction for future investigation.

While findings of the current study are novel and additive, they are not without their limitations. First, there was a short period of time between the coaching and independent image acquisition. This was purposely done, however, to prove feasibility in the simplest of settings. Introducing a larger delay in future studies will certainly be a focus. Second, there were not many patients in whom ventricular function was depressed, a setting in which inter-reader variability of left-ventricular fractional shortening could be greater and correlation with hospital acquired fractional shortening could be lower. Future studies with a cohort of those with greater left-ventricular dysfunction will help further clarify this. Third, there may be selection bias in that parents who consented to participate in this study may have been intrinsically more interested and motivated in learning how to utilize point of care ultrasound and may represent a subset that outperforms the average. Fourth, some degree of difference in fractional shortening will be present whenever different devices and software are used. The degree of difference due to the probe and software itself is not quantifiable here but even despite some inherent differences the correlation with hospital acquired fractional shortening was good. Fifth, we allowed for hospital acquired fractional shortening to be within 48 h of the point of care ultrasound meaning that some difference in function may have just been a function of time, particularly in patients acutely admitted, thus impacting underestimating the correlation of point of care ultrasound to hospital acquired echocardiography. Sixth, image quality based on ultrasound probe presets may have been slightly impacted, leading to an underestimation of the true correlation. Future studies using custom presets would be of use. Seventh, use of an averaged value to quantify the correlation between the parent acquired and hospital acquired fractional shortening inherently leads to a reduction in the correlation. Thus, there are several mechanisms by which the correlation may be underestimated. Eighth, while the correlation between the two readers and between the parent acquired and hospital acquired images was statistically significant, the magnitude of correlation was still limited. Other limitations may have limited this correlation.

Despite the limitations of this study, these pilot data stress the importance for future investigation of parental echocardiography.

Conclusion

Parents were able to successfully obtain a parasternal short-axis and apical four-chamber image adequate to assess function and quantify fractional shortening after a 25-min education session. These pilot data demonstrate that further exploration of parent-performed point of care cardiac assessment may be warranted.

Author Contributions

AJ—study design, patient consent, data collection, manuscript preparation RL—study design, data analyses, manuscript preparation

Declarations

Conflict of interest

The authors declare no competing interests.

Ethical Approval

The study was conducted using a handheld ultrasound probe that was provided on complimentary loan from Philips (Amsterdam, Netherlands). The company was not involved with study design, data collection, data analysis, or manuscript preparation.