Feasibility of a point‐of‐care ultrasound protocol for cardiorespiratory evaluation of horses in different clinical settings

Abstract Background A point‐of‐care ultrasound (POCUS) protocol for evaluation of the cardiac and respiratory systems in horses does not exist. Objectives (a) Describe the windows of a POCUS protocol for cardiorespiratory assessment of horses (CRASH); (b) Estimate the number of acoustic windows that can be acquired by a sonographer‐in‐training; (c) Estimate the time required to complete the protocol for specific groups of horses; (d) Describe the sonographic abnormalities detected in horses presented with cardiovascular, respiratory, or systemic disease. Animals Twenty‐seven healthy horses, 14 horses competing in athletic events, and 120 horses with clinical disease. Method A pocket‐sized ultrasound device was used to acquire 7 sonographic cardiorespiratory windows in various clinical scenarios. The duration of the examination was timed, and images were evaluated for diagnostic quality. Abnormalities in horses with clinical disease were determined by an expert sonographer. Results The CRASH protocol could be performed in healthy and diseased horses in hospital, barn, and competition settings between 5.5 ± 0.9 (athletic horses) and 6.9 ± 1.9 min (horses with clinical disease). Thoracic windows were obtained most consistently, followed by right parasternal long‐axis echocardiographic windows. Frequently detected abnormalities were pleural fluid, lung consolidation, B‐lines, and moderate‐to‐severe left‐sided heart disease. Conclusions The CRASH protocol was feasible using a pocket‐sized ultrasound device in various groups of horses, could be completed rapidly in a variety of settings, and frequently identified sonographic abnormalities when evaluated by an expert sonographer. The diagnostic accuracy, observer agreement, and utility of the CRASH protocol merit further evaluation.


| INTRODUCTION
Ultrasonography allows rapid, stall-side, real-time assessment of multiple body systems in horses. It is used frequently in specialty and general practice in hospital and field settings. Comprehensive ultrasonography and examination of complex disease often entails detailed evaluations that might require advanced training and can be time-consuming.
Point-of-care ultrasound (POCUS) or clinician-performed ultrasound (CPUS) 1 protocols are increasingly used in veterinary and human medicine because of technological and training advances. Focused ultrasonographic protocols are time-sensitive, goal-oriented assessments that answer focused clinical questions based on a patient's clinical signs.
Point-of-care ultrasound protocols for cardiac and lung evaluation first were introduced in human medicine in the 1990s. Since then, different standardized protocols that offer valuable information in specific clinical scenarios have emerged. 2 Some of these protocols have become particularly relevant during the current coronavirus 2019 (COVID-19) disease pandemic. 3,4 Multiple studies have evaluated the utility and accuracy of focused cardiac and lung ultrasound examinations as well as the training programs that have been designed to teach focused cardiac ultrasound examination in humans and small animals. [5][6][7][8][9][10][11][12][13] Focused ultrasound examinations that combine evaluation of both the respiratory and cardiac systems exist for humans but none currently is described for small animals or horses. 14 Common scenarios in which focused cardiac and respiratory ultrasound examination have been implemented for human and small animal medicine are emergency and critical care, resource-limited settings, and primary care practice. In these instances, comprehensive ultrasound examination or other forms of advanced imaging may not be readily available or needed, and a focused examination can increase diagnostic accuracy of common cardiac and respiratory diseases. 14 The American Society of Echocardiography and other medical organizations have published multiple consensus statements to establish definitions, protocols, recommendations, and limitations, and to review information regarding diagnostic accuracy of focused echocardiography and lung ultrasound examination. These guidelines often emphasize that focused ultrasound examinations are not intended to replace comprehensive echocardiograms, comprehensive sonograms, or the work of specialists but to augment diagnostic procedures and improve medical care. [15][16][17] Point-of-care ultrasound protocols for the evaluation of adult horses with colic (FLASH) and high-risk pregnancies have been evaluated. 18,19 Preliminary concepts of POCUS protocols for horses designed to define standards of training, indications and diagnostic accuracy of focused echocardiography, and POCUS for cardiorespiratory assessment recently were introduced. 20 The goal of these protocols is to make medical care accessible while maintaining or improving standards of care in general and specialty practice.
Our goal was to describe a protocol for focused cardiorespiratory ultrasound (CardioRespiratory Assessment with Sonography in the Horse [CRASH]) targeted to detect cardiac and respiratory diseases in the horse. This protocol combines assessment of both cardiac and respiratory systems because diseases of either system can sometimes be difficult to differentiate clinically, especially in emergent or time-sensitive scenarios. Our specific objectives study were: (a) to select and describe the windows of CRASH, a POCUS protocol for cardiore-

| Protocol
Seven windows were selected based on described windows that were expected to identify cardiovascular and respiratory abnormalities in the horse. Windows also were selected based on perceived ease of application and execution by non-experts on horses in various clinical scenarios. Abnormalities sought were pleural effusion, pericardial effusion, moderate or severe lung pathology, pneumothorax, moderate or severe pulmonary hypertension, moderate or severe left heart disease, valvular or myocardial disease, and moderate or severe hypertrophy or pseudohypertrophy of the left ventricle (LV), according to study definitions (Table 1). Standard echocardiographic windows for horses were chosen as recently reviewed, 21 and lung ultrasound windows were chosen based on literature describing lung sonographic abnormalities in the horse. [22][23][24] The following are the windows of the protocol and the possible abnormalities to be detected: • Right parasternal long axis 4-chamber view (R4C) -The transducer is positioned in the right fourth intercostal space at a level slightly above the olecranon, angled caudally, and rotated clockwise to the 1 o'clock position. The ventricles, atrioventricular valves and atria are imaged. Imaging the LV apex is prioritized and the dorsal aspect of the atrium may not be imaged in its entirety in all horses. This view is used to assess the structures and dimensions of the LV and left atrium (LA). This view also can be used to assess the presence of pericardial effusion, as well as abnormalities of the myocardium and mitral (MV) and tricuspid valves (TV).
• Right parasternal long axis view of the left ventricular outflow tract (LVOT) -Starting from a R4C view, the transducer is angled cranially, and rotated to the 2 o'clock position. It is used to assess the structure and dimensions of the aorta (Ao) and pulmonary artery (PA). Specific attention is paid to the relative size of the PA and the Ao as a marker for pulmonary hypertension. This view also can be used to assess the presence of pericardial effusion, as well as abnormalities of the myocardium and aortic valve (AoV) and TV.
• Right parasternal short axis view at the level of the chordal attachments (SAch). This view is obtained by rotating the transducer clockwise to the 3 to 4 o'clock position from a 4-chamber view. It is used to assess ventricular size. This view also can be used to assess the presence of pericardial effusion and abnormalities of the myocardium.  This view traditionally has been used for assessment of LA dimensions. This view also can be used to assess the presence of pericardial effusion, as well as abnormalities of the myocardium and MV.
The proposed CRASH protocol does not define normal ranges for measurement, and sizes are assessed subjectively and relative to internal references. Isopropyl alcohol on unclipped hair was used as a coupling agent. In all images, dorsal or cranial was at the right side of the screen. A pocket-sized ultrasound unit (Butterfly iQ Vet, Butterfly Network, Guilford, CT) was used with a compatible smart phone device (iPhone 12 Pro, Apple, Cupertino, California). This pocket-sized device has been used in 2 previous studies in horses that obtained common sonographic windows and detected multiple sonographic abnormalities. 25,26 Images for this study were acquired in B-mode with depth set to 30 cm on the "Abdomen Deep" setting, which offers a convex field of view. Images were saved in an online Digital T A B L E 1 Eight potential abnormalities to be recognized with examination and the criteria for determining the presence or absence.

Is there increased pleural fluid?
Increased pleural fluid is defined as fluid in the pleural space that is more than in the lateral aspect of the most ventral lung tip and more than approximately 1 cm in depth.

Is there increased pericardial fluid?
Increased pericardial fluid is defined as fluid visible in the pericardial space that is more than a few millimeters in depth 3. Is there moderate or severe lung pathology such as, consolidation, masses, abscesses and severe or coalescing B-lines or comet tails.
Consolidation is defined as hypoechoic area of lung with present bronchial or vascular markings. Mass and abscesses are defined as well circumscribed areas in the lung or pleural space that do not contain normal bronchial or vascular markings. Moderate or severe B-lines or comet tails are defined as interruptions of the smooth and regular hyperechoic echo of lung that merge or create patches.

Is there a pneumothorax?
Pneumothorax is the presence of air in the pleural space. This is visible as a hyperechoic echo of gas free in the pleural space (outside of the lung) and therefore the hyperechoic echo of lung is not seen sliding with respiratory motions 5. Is there evidence of moderate or severe pulmonary hypertension?
These are a pulmonary artery that is larger than the aorta's sinotubular junction or shape of the interventricular septum becoming concave right to left 6. Is there evidence of moderate to severe left side heart disease?
These are a left ventricular apex that is rounded and not cone shaped, the left ventricle is severely enlarged and compressing the right ventricle, or the left atrium is moderately or severely enlarged losing its rectangular shape and being disproportionate to the size of the right atrium. Imaging and Communications in Medicine (DICOM) server. The Butterfly iQ device has only 1 transducer type, and settings are determined by the preset and depth of examination using proprietary algorithms.
Imaging windows are described above and in previously publications 23 and reference images are shown in Figure S1.  Table 1) and would not have enabled answering the clinical questions described in Table 1. An image was considered acceptable if anatomic structures were visible and the image would enable a sonographer to answer the questions described in Table 1.

| Training phase
Sonographers 1 and 2 reviewed images obtained using the pocket-sized device and standard ultrasound machine to gain insights about variations associated with individual horse factors, device capability, or operator error, and to subsequently improve the image quality. given to assessment of the presenting complaint by the primary clinician, and therefore patient location was dependent on orders from the clinician on the case and not the primary investigator. The duration of the sonogram was the time between the first and last obtained images.

| Feasibility phase
For this phase, the review process involved identifying normal anatomical structures that typically are observed within each window (Figure 1).
Sonographer 2 then evaluated clips obtained by sonographer 1 and determined them to be of either acceptable or unacceptable quality. An image was considered unacceptable if anatomic structures described in Figure 1 were not visible and would not have enabled answering the clinical questions described in

| Training phase
For all horses of a homogenous group, sonographers 1 and 2 were able to obtain images in all windows in all 10 horses, and all images obtained in each window were considered of acceptable quality. The operator using a POCUS device was able to obtain acceptable images consistently.

| Feasibility study
Twenty-seven healthy horses were enrolled in the healthy horse group. Various breeds were represented (20 Quarter Horses, 6 Thoroughbreds, 1 Warmblood) with a mean age of 5 years (SD, 5 years) and body condition scores (BCS) ranged from 4 to 7 out of 9 with a median of 5 ( Figure S2). 35

| Clinical phase
Sonographic abnormalities identified in horses (n = 120) presented with cardiac, respiratory, or systemic disease were tabulated (Table 3).
T A B L E 2 Percentage of acceptable quality images based on review by sonographer 2 for each of the windows obtained by sonographer 1; duration of time required for examination by sonographer 1.  window, all detected abnormalities and windows were described (Table 3; Figure S3). For the 70 additional horses examined by sonographer 2, the duration of the sonogram was a mean of 5.9 min (SD ± 5.8 min) and the duration of the sonogram for horses imaged by sonographer 1 in the feasibility study was a mean of 6.9 min (SD ± 2.0 min). In 3 of the 70 horses (4%), no cardiac images of diagnostic quality could be obtained.

| DISCUSSION
Our results indicated that performing the proposed CRASH protocol was feasible using a pocket-sized ultrasound device in healthy, hospitalized, and athletic horses. The time required to complete the examination for all groups was suitable for most time-sensitive scenarios.
Thoracic windows were most consistently acquired and of clinically acceptable quality. Horses that had poor quality thoracic windows were those in which none of the windows of CRASH were of good quality. Increased body condition, long hair coat, or both often were recorded, but the association of these patient findings with the quality of images was not statistically analyzed (because it was not an a priori hypothesis). Our perception is that these patient factors con- improve until approximately 300 examinations are completed at which time skills start to plateau. 36 At the time of the study, sonographer 1 had not achieved this number and it is reasonable to believe that skills were developing but above average for the general population of equine practitioners who might use CRASH. In a recent study, participants who were non-specialist practitioners learned to perform and interpret a focused echocardiographic protocol for horses independent of previous experience in a 1-day course using Peyton's 4-step approach. 20 The described protocol was feasible in horses participating in equestrian events. Similarly, the use of focused cardiac ultrasound by sports medicine physicians has proved successful for acquiring images that recognize common causes of sudden athletic death such as hypertrophic cardiomyopathy and aortic root dilatation in human athletes. 37,38 The incidence of sudden athletic death in horses that participate in high intensity exercise disciplines is estimated to be over 200 times higher than that of humans. 39,40 Echocardiography is considered a second-tier diagnostic modality in pre-participation screening programs for human athletes, where history, physical examination, and 12-lead ECG are the most common components of the programs. 41 The conditions that cause sudden athletic death in horses and humans are different 42 and therefore it is plausible that prevention likely requires a different approach. The ability of a protocol such as that described here to detect preventable causes of sudden athletic death requires further study. as well as the higher incidence of respiratory disease as compared to cardiac disease in horses. Abnormalities are not always detected in the acoustic windows described for CRASH and it is therefore possible that certain abnormalities might go unrecognized using the described protocol. Clinicians performing CRASH should be aware of these limitations and not consider the examination a replacement for more comprehensive diagnostic testing or specialist evaluation. Also, the abnormalities detected with CRASH were not formally compared to findings of more comprehensive tests. A different study comparing results of the CRASH protocol to gold standards will be needed to accomplish this necessary step.
A limitation of the study was that a sonographer who was consid-

| CONCLUSIONS
We described a focused cardiorespiratory ultrasound protocol (CRASH) that could be performed in healthy and diseased horses in a hospital and barn setting in 5.5 to 6.9 min on average. A sonographer in training could acquire good quality images most consistently through thoracic windows followed by right parasternal long axis echocardiographic windows. Abnormalities that were commonly detected with the CRASH protocol by an expert sonographer included pleural fluid, lung consolidation, coalescing B-lines or comet tail artifacts, and moderate or severe left heart disease. Sonographic evidence of moderate or severe pulmonary hypertension or left ventricular hypertrophy or pseudohypertrophy also were common findings in our study population, whereas pericardial fluid, pneumothorax or moderate or severe myocardial or valvular echogenicity changes were detected rarely. The diagnostic accuracy and potential clinical utility of the CRASH protocol to improve veterinary care of horses deserves further evaluation.

Support provided by the American College of Veterinary Internal
Medicine Clinical Training Fellowship, the Texas A&M University Department of Large Animal Clinical Sciences, and the Texas A&M