Abstract
Objective
To assess the ability to visually estimate the fractional shortening in dogs and the impact of experience on those assessments.
Methods
Right parasternal short- and long-axis cine loops from 25 dogs with varying fractional shortening (6.9% to 61.2%) were distributed online to observers with different levels of training in anesthesiology or cardiology and different experience (advanced, intermediate, beginner, or none) interpreting echocardiography. Observers were asked to estimate the fractional shortening and assign 1 of 5 categories: hyperdynamic (> 45%), normal (25% to 45%), mildly decreased (20% to 24%), moderately decreased (15% to 19%), or severely decreased (< 15%). The answers for all cases combined were graded as correct/incorrect based on each of the 5 fractional shortening categories and compared between participants of each level experience. In addition, answers were reanalyzed using 2 categories for fractional shortening, ≥ 20% and < 20%, to test the identification of clinically important decreased function.
Results
Answers from 240 participants were received. The overall rate of correct answers averaged between 45% and 47% for all levels of experience. This improved from 79% to 83% across all levels of experience when a 2-category scale was used. Participants with advanced experience performed slightly better. The success rate for identifying fractional shortening < 20% when the dogs had moderate and severe decreases was 50% to 54% and 92% to 96%, respectively.
Conclusions
Visual estimations of the fractional shortening can be used reliably to identify cases of severe but not moderately decreased function.
Clinical Relevance
A basic echocardiographic exam may identify dogs with severely decreased fractional shortening. Measurements are required to distinguish those with moderate function.
Clinicians commonly assess left ventricular (LV) systolic function in dogs with transthoracic echocardiography (TTE) using images acquired with either M-mode or 2-D to measure fractional shortening, fractional area change, or ejection fraction.1–3 While ejection fraction is typically used in humans, fractional shortening is more commonly reported in veterinary medicine. As an approximation of systolic function, visual estimation of ejection fraction based on a basic (preanesthetic) echocardiogram has demonstrated value in humans.4,5 The importance of quantifying LV systolic function may extend beyond animals with confirmed or suspected cardiac disease,6,7 as systolic dysfunction has also been associated with sepsis, inflammatory diseases, cancer, radiation therapy, and toxicosis, among other causes.8–13 Similar assessment in dogs might reduce anesthesia-associated morbidity and mortality by allowing detection of depressed systolic function before induction.
Therefore, we investigated the ability of observers with varying degrees of cardiology or anesthesiology training and experience with echocardiography to estimate the fractional shortening from cine loops of dogs previously examined at Cornell University. Our goals were to evaluate the ability of observers to (1) visually estimate the fractional shortening within 1 of 5 categories, (2) identify clinically relevant decreases in fractional shortening, and (3) assess the impact of specialty training and experience on the fractional shortening assessments.
Methods
This retrospective study was undertaken with the approval of the Cornell University Veterinary Clinical Studies Committee (protocol 090424-14) and considered exempt from IACUC and Institutional Review Board review. A survey was created in a cloud-based platform (Qualtrics) and distributed to subscribers of the email list of the American College of Veterinary Anesthesia and Analgesia (1,289 subscribers) and the American College of Veterinary Internal Medicine/European College of Veterinary Internal Medicine (ACVIM/ECVIM) Cardiologists listserv, hosted by the Veterinary Information Network (886 subscribers). The anonymous Qualtrics survey included 30 slides and could be accessed only once per participant. The initial 4 slides contained an introduction and instructions and an example of a dog with normal fractional shortening (39.5%).
Participants first provided their level of training according to the following categories: diplomate or resident of the American College of Veterinary Anesthesia and Analgesia or European College of Veterinary Anesthesia and Analgesia, diplomate or resident of the ACVIM or ECVIM (Cardiology), doctor of veterinary medicine (DVM) practicing anesthesia, DVM practicing cardiology, or other. In addition, observers reported their experience in performing or interpreting TTE in dogs according to these categories: advanced (performs or interprets TTE weekly and feels confident in their findings), intermediate (performs or interprets TTE every month with some confidence), beginner (performs or interprets a handful of TTE per year but does not feel confident in their findings), or none (does not perform or interpret TTE). The survey was not open to students or veterinary technicians.
Participants then observed cine loops of echocardiograms obtained from 25 dogs examined by the cardiology service of the authors’ institution. Each case consisted of 2 TTE views: a right parasternal LV short axis at the level of the papillary muscles and a right parasternal long axis (4 chamber), obtained from the same dog at the same appointment. All images had been previously obtained (between August 2022 and July 2024) by or under direct supervision of a diplomate of the ACVIM (Cardiology) and stored in our institution's picture archiving and communication system. The measured fractional shortening for these cases ranged between 6.9% and 61.2%, distributed in 5 cases per category and randomly organized, using a random sequence generator (www.random.org). The fractional shortening was averaged from 3 consecutive beats. Several cardiac cycles were displayed in each clip, which played in a continuous loop. Upon observation of both clips, participants assessed the fractional shortening within 1 of 5 categories, designed from several sources,14,15 including personal consultations with veterinary cardiologists: hyperdynamic (fractional shortening > 45%), normal (fractional shortening 25% to 45%), mildly reduced (fractional shortening 20% to 24%), moderately reduced (fractional shortening 15% to 19%), and severely reduced (fractional shortening < 15%). Once a selection was made, the next case was presented, without the possibility of returning to previous cases or modifying answers. The survey was accessible for 11 days (November 6, 2024 to November 17, 2024). The complete list of cases, including the breed, age, weight, and measured fractional shortening, which was not available to the participants, is shown in Supplementary Table S1.
Analysis
Descriptive summaries are provided for the participants’ experience and credentials. The answers for each case (all 25 cases combined) were graded as correct/incorrect based on each of the 5 fractional shortening categories. The overall percentage of correct answers was compared between participants of each experience level using ANOVA and Tukey post hoc tests after assessing the parametric/nonparametric distribution of the data by inspection of the histograms. The significance level was set at P < .05.
Given that dogs with hyperdynamic, normal, or even mildly reduced fractional shortening are unlikely to suffer complications associated with decreased systolic function during anesthesia, we also imposed a simplified grading system whereby any answer indicating a fractional shortening ≥ 20% (hyperdynamic, normal, and mildly reduced fractional shortening) was considered correct as long as the measured fractional shortening was ≥ 20%, without distinction of sub-categories within that range. Similarly, the identification of a fractional shortening < 20% was considered correct if the measured fractional shortening was < 20%, without distinction between moderate and severe decreases. This simplified score was performed to test the participant's ability to detect cases with a substantial decrease in fractional shortening that might be considered clinically important during a perianesthetic evaluation. The percentage of correct answers per fractional shortening category (using 5 or 2 categories) was reported and stratified by level of experience. The χ2-test was used to compare the performance between levels of experience when using a 2-scale grading for fractional shortening.
Results
Two-hundred forty participants completed the survey, comprising 73 diplomates of the American College of Veterinary Anesthesia and Analgesia or European College of Veterinary Anesthesia and Analgesia, 21 anesthesiology residents, 86 diplomates of the ACVIM or ECVIM (Cardiology), 12 cardiology residents, 7 DVMs practicing anesthesia, 30 DVMs practicing cardiology, and 11 participants with other credentials (3 Asian College of Veterinary Internal Medicine, 1 large animal internal medicine, 2 small animal internal medicine, 1 emergency and critical care, 2 Royal College of Veterinary Surgeons diploma in cardiology, and 2 unspecified). Anesthesiologists and anesthesiology residents generally had low levels of experience, while cardiologists and cardiology residents reported advanced levels of experience in performing and/or interpreting echocardiograms (Figure 1).
Number of participants arranged by their credentials (diplomate or resident of the American College of Veterinary Anesthesia and Analgesia [ACVAA] or European College of Veterinary Anesthesia and Analgesia [ECVAA], diplomate or resident of the American College of Veterinary Internal Medicine [ACVIM] or European College of Veterinary Internal Medicine [ECVIM (Cardiology)], doctor of veterinary medicine [DVM] practicing anesthesia or cardiology, and “other,” which included diplomates of the Asian College of Veterinary Internal Medicine [Cardiology], small and large animal internal medicine, emergency and critical care, Royal College of Veterinary Surgeons diploma in cardiology, or unspecified) and level of experience (performs or interprets transthoracic echocardiography [TTE] weekly and feels confident in their findings [Advanced]; performs or interprets TTE every month with some confidence [Intermediate]; performs or interprets a handful of TTE per year but does not feel confident in their findings [Beginner]; and does not perform or interpret TTE [None]).
Citation: American Journal of Veterinary Research 86, 5; 10.2460/ajvr.24.12.0384
When considering all 25 cases combined, participants had a relatively poor ability to correctly classify the echocardiograms into the 5 categories of fractional shortening, with mean percentages of correctly classified images < 50% (Figure 2). Participants with the highest level of experience had a marginally higher probability of correct classification (P = .003) than the remaining 3 experience levels (beginner group scored higher than intermediate and no experience, both P < .025). When fractional shortening was divided into 2 categories (fractional shortening < 20% vs ≥ 20%), accuracy increased substantially, with mean percentages of approximately 80%. Again, participants with the highest levels of experience had marginally higher accuracy than participants with lesser levels of experience (P < .001).
Mean (line), 95% CI (colored area), and individual (circles) percentage of overall correct answers per level of experience when considering all 25 cases combined. Answers were graded into 5 (black) or 2 (red) categories of fractional shortening. The 5 categories for fractional shortening were hyperdynamic (> 45%), normal (25% to 45%), mildly reduced (20% to 24%), moderately reduced (15% to 19%), and severely reduced (< 15%). Answers were also graded using 2 categories, with the fractional shortening divided into ≥ 20% or < 20%. Groups were compared with ANOVA and Tukey post hoc tests for answers assigned within 5 (P = .022) and 2 (P < .001) categories. Different letters over or under the answers indicate statistical significance between groups of participants.
Citation: American Journal of Veterinary Research 86, 5; 10.2460/ajvr.24.12.0384
Correct answers per fractional shortening category and the identification of moderately/severely decreased fractional shortening stratified by level of experience are shown in Figures 3 and 4. When examined by each of the 5 fractional shortening categories, but stratified into 2 categories (fractional shortening < 20% and fractional shortening ≥ 20%), clinicians with the highest levels of experience identification more frequently identified a fractional shortening ≥ 20% than participants with other levels of experience (P ≤ .0006), making fewer errors in each of the 3 fractional shortening categories. However, we found no differences in the frequency of correct classification of fractional shortening < 20% by level of experience (P > .6): all participants, regardless of level of experience, were equally likely to correctly identify a moderate to severe decrease in fractional shortening.
Percentage of correct answers for each level of experience and for each category of fractional shortening, when graded using all 5 categories. The 5 categories for fractional shortening were hyperdynamic (> 45%, red), normal (25% to 45%, orange), mildly reduced (20% to 24%, yellow), moderately reduced (15% to 19%, light blue), and severely reduced (< 15%, dark blue).
Citation: American Journal of Veterinary Research 86, 5; 10.2460/ajvr.24.12.0384
Percentage of correct answers for each level of experience and for each category of fractional shortening. The 5 measured categories for fractional shortening were hyperdynamic (> 45%), normal (25% to 45%), mildly reduced (20% to 24%), moderately reduced (15% to 19%), and severely reduced (< 15%). Answers were graded using 2 categories, with the fractional shortening divided into ≥ 20% (orange) or < 20% (blue). There were statistical differences between levels of experience when the fractional shortening was ≥ 20% (hyperdynamic, normal, and moderately decreased, all P < .0006) but not when the measured fractional shortening was < 20% (moderately and severely decreased, both P > .6).
Citation: American Journal of Veterinary Research 86, 5; 10.2460/ajvr.24.12.0384
Discussion
Our study demonstrates that clinicians can reliably identify a severe decrease in fractional shortening, regardless of prior echocardiographic experience. However, clinicians with lower levels of echocardiographic experience tend to occasionally, and more frequently than those with the highest level of experience, incorrectly identify a moderate to severe decrease in fractional shortening when none exists. All clinicians, regardless of echocardiographic experience have similar levels of difficulty in correctly identifying fractional shortening of 15% to 19%, with approximately half misidentifying dogs that have a fractional shortening ≥ 20% as having a fractional shortening of 15% to 19%, while few have difficulty correctly identifying either hyperdynamic systolic function (fractional shortening > 45%) or severely decreased systolic function (fractional shortening > 15%).
A fast visual assessment of the fractional shortening might be useful during the perianesthetic exam in dogs with an increased risk of systolic dysfunction, providing that reliable information can be obtained. A rapid visual estimation of fractional shortening to identify any clinically relevant decrease in systolic function might suffice in a hectic, suboptimal imaging environment. Our data support this idea, given that participating clinicians with varying levels of echocardiographic experience could reliably identify severe decreases in fractional shortening. While clinicians had more difficulty correctly identifying the degree of fractional shortening when asked to select 1 of 5 levels (and with minimal differences between experience levels), they performed much better when data were categorized into what we considered 2 clinically relevant categories. In this analysis, all groups of observers correctly identified the category 79% to 83% of the time on average. We also stratified those answers for each of the 5 categories of the measured fractional shortening. Importantly, the success rate for identifying dogs with severely decreased fractional shortening as < 20% ranged between 92% and 96% across all levels of experience, suggesting that even novice observers can identify those levels of dysfunction. However, the identification of dogs with moderately decreased fractional shortening (15% to 19%) ranged between 50% and 54%, regardless of experience level. Our data suggest that identifying dogs with normal or even hyperdynamic function and those with severe decreases in fractional shortening can be done with reasonable reliability regardless of the level of experience. The identification of mild to moderate dysfunction, however, is less reliable and does not seem to improve with experience. A different study16 observed a “ceiling effect” in the reproducibility of estimations based on observation, which may be inherent to visual assessment. Taken together, our observations suggest that while normal and severely decreased fractional shortening can be subjectively assessed, the identification of moderate decreases requires that the fractional shortening be measured.
Systolic dysfunction can arise from several causes, both intracardiac, such as dilated cardiomyopathy,6,7 and extracardiac, including sepsis, cancer, or toxicosis, among others.8–10,12 The diagnosis of such dysfunction could prove helpful as part of the perianesthetic preparation. However, the idea that a decreased fractional shortening by itself indicates LV dysfunction is an oversimplification. Fractional shortening is a load-dependent parameter and is influenced by both preload and afterload. It is not a true representation of contractility. Fractional shortening will almost surely be decreased in most dogs medicated with α2-agonists, a very common preanesthetic sedative. By itself, this does not indicate that contractility is inherently decreased, but rather that it is affected by an increased afterload. The same is true for other drugs commonly used during anesthesia that increase afterload, such as phenylephrine. Fractional shortening is also affected by preload and mitral valve insufficiency. In short, fractional shortening does not represent contractility and should be interpreted within a wider context.
While we saw only minor differences between levels of experience in the ability to “eyeball” fractional shortening, all images had been obtained by trained cardiology diplomates or residents. Our results, suggesting that the interpretation of these images was similar between participants with different levels of experience, should not be interpreted any more widely than precisely that. Image acquisition requires skill and experience, and poorly acquired images can affect interpretation. Second, fractional shortening should be interpreted within the context of both intracardiac and extracardiac variables. Other limitations include the presence of cases in which the measured fractional shortening value just falls into one category. If a fractional shortening < 15% is considered severely reduced, by definition a fractional shortening of 16% is not. Clearly, this would pose a real difficulty in correctly identifying cases visually. More importantly, even when the fractional shortening is measured, interpretation is required when deciding if values close to the thresholds are considered acceptable or not. The interpretation of the fractional shortening and whether a response is triggered by an observation escape the narrower scope of this study, which was to assess the ability of participants to qualify the fractional shortening based solely on observation.
In summary, cases of severely decreased fractional shortening can be reliably identified from a subjective observation of right parasternal short- and long-axis cine loops, even by clinicians with little experience. A quick diagnosis from a point-of-care TTE examination could prove useful as part of the perianesthetic assessment of dogs. Moderate decreases in fractional shortening cannot be reliably detected subjectively, and clinicians should obtain standard measurements in such cases.
Supplementary Materials
Supplementary materials are posted online at the journal website: avmajournals.avma.org.
Acknowledgments
The authors thank the Cardiology service at the Cornell Hospital for Animals for help with retrieving cases.
Disclosures
Dr. Martin-Flores and Dr. Rishniw are members of the AJVR Scientific Review Board but were not involved in the editorial evaluation of or decision to accept this article for publication.
No AI-assisted technologies were used in the composition of this manuscript.
Funding
The authors have nothing to disclose.
ORCID
M. Martin-Flores https://orcid.org/0000-0003-2014-9040
M. Rishniw https://orcid.org/0000-0002-0477-1780
References
- 1.↑
Cornell CC, Kittleson MD, Della Torre P, et al. Allometric scaling of M-mode cardiac measurements in normal adult dogs. J Vet Intern Med. 2004;18(3):311–321.
- 2.
Crippa L, Ferro E, Melloni E, Brambilla P, Cavalletti E. Echocardiographic parameters and indices in the normal beagle dog. Lab Anim. 1992;26(3):190–195.
- 3.↑
Bonagura JD. M-mode echocardiography. Basic principles. Vet Clin North Am Small Anim Pract. 1983;13(2):299–319.
- 4.↑
Gudmundsson P, Rydberg E, Winter R, Willenheimer R. Visually estimated left ventricular ejection fraction by echocardiography is closely correlated with formal quantitative methods. Int J Cardiol. 2005;101(2):209–212.
- 5.↑
Sievers B, Kirchberg S, Franken U, Puthenveettil BJ, Bakan A, Trappe HJ. Visual estimation versus quantitative assessment of left ventricular ejection fraction: a comparison by cardiovascular magnetic resonance imaging. Am Heart J. 2005;150(4):737–742.
- 6.↑
Bonagura JD, Visser LC. Echocardiographic assessment of dilated cardiomyopathy in dogs. J Vet Cardiol. 2022;40:15–50.
- 7.↑
Rath P, Jena B, Behera SS, Sathapathy S, Rath PK, Sarangi S. Comparative study of echocardiographic parameters in healthy and dilated cardiomyopathy-affected dogs. Vet Res Forum. 2024;15(8):403–409.
- 8.↑
Nelson OL, Thompson PA. Cardiovascular dysfunction in dogs associated with critical illnesses. J Am Anim Hosp Assoc. 2006;42(5):344–349.
- 9.
Naseri A, Akyuz E, Turgut K, Guzelbektes H, Sen I. Sepsis-induced cardiomyopathy in animals: from experimental studies to echocardiography-based clinical research. Can Vet J. 2023;64(5):871–877.
- 10.↑
Van Vleet JF, Ferrans VJ, Weirich WE. Cardiac disease induced by chronic adriamycin administration in dogs and an evaluation of vitamin E and selenium as cardioprotectants. Am J Pathol. 1980;99(1):13–42.
- 11.
Hanai K, Takaba K, Manabe S, Nakano M, Kohda A, Matsuo M. Evaluation of cardiac function by echocardiography in dogs treated with doxorubicin. J Toxicol Sci. 1996;21(1):1–10.
- 12.↑
Tagawa M, Shimbo G, Uemura A, Matsumoto K. Cardiomyopathy in a dog with multicentric lymphoma following treatment with several anthracyclines. Open Vet J. 2021;11(1):6–10.
- 13.↑
Kruckman-Gatesy CR, Ames MK, Griffin LR, et al. A retrospective analysis of stereotactic body radiation therapy for canine heart base tumors: 26 cases. J Vet Cardiol. 2020;27:62–77.
- 14.↑
Lang RM, Bierig M, Devereux RB, et al. Recommendations for chamber quantification: a report from the American Society of Echocardiography's Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr. 2005;18(12):1440–1463.
- 15.↑
Tissot C, Singh Y, Sekarski N. Echocardiographic evaluation of ventricular function-for the neonatologist and pediatric intensivist. Front Pediatr. 2018;6:79. doi:10.3389/fped.2018.00079
- 16.↑
Cole GD, Dhutia NM, Shun-Shin MJ, et al. Defining the real-world reproducibility of visual grading of left ventricular function and visual estimation of left ventricular ejection fraction: impact of image quality, experience and accreditation. Int J Cardiovasc Imaging. 2015;31(7):1303–1314.
