Pulmonary hypertension has been recognized as a clinical problem for many years in veterinary medicine. Routine accurate clinical diagnosis of pulmonary hypertension in dogs specifically has been markedly enhanced by the widespread use of echocardiography.1 However, systematic assessment of function of the right chambers of the heart is not uniformly carried out. This lack of consistency is due in part to the high amount of attention given to evaluation of the left chambers of the heart, a lack of familiarity with ultrasonographic techniques available for imaging the right chambers, and a paucity of ultrasonographic studies providing reference intervals for size and function of the right chambers.2 To date, reference intervals and repeatability of right-chamber heart function indices for echocardiographic evaluation, such as peak systolic tricuspid annulus velocity, tricuspid annulus plane systolic excursion, and systolic longitudinal right ventricular strain, have been reported.3–6 In addition, right-chamber heart function tests involving echocardiography have been used in clinical settings for dogs.3,6,7 These indices, however, provide an assessment of only the regional or overall systolic function of the right chambers.
The Tei index (also called the myocardial performance index) is an index of overall myocardial function, including systolic and diastolic performance.8 This measurement involves a simple technique, and it correlates well with both the systolic and diastolic function of the right ventricle.9 Therefore, it has been used to evaluate right ventricular function and provide information on severity and prognosis for dogs with cardiac disease.9,10
The Tei index has been derived from conventional pulsed-wave Doppler8 and tissue Doppler echocardiography.11 However, an important limitation exists in that values derived from conventional pulsed-wave Doppler cannot be calculated in a single cardiac cycle; therefore, the Tei index is influenced by heart rate fluctuations (ie, respiratory sinus arrhythmia). On the other hand, tissue Doppler echocardiography can be used to simultaneously record diastolic and systolic phases, but values measured by use of that method in dogs are reportedly different from those obtained via conventional pulsed-wave Doppler echocardiography.12
Dual pulsed-wave Doppler echocardiography allows Doppler signals at 2 points to be simultaneously measured, setting 2 separate sample volumes in 1 image. Therefore, measurement of the RTX during the same cardiac cycle is possible, which may overcome the limitations associated with a conventional pulsed-wave Doppler approach. In human medicine, the intra- and interobserver reliability of RTXDPD are high.13 In addition, RTXtd values are higher than RTXPD and RTXdpd values.13 However, to the authors' knowledge, no reports exist on RTXdpd measurement in dogs. Moreover, degrees of agreement among RTX values derived from the 3 echocardiographic methods have not been reported for dogs.
The purpose of the study reported here was to evaluate the repeatability and reproducibility of RTXDPD, RTXPD, and RTXtd measurements in healthy dogs. In addition, we also sought to investigate statistical relationships among RTXDPD, RTXTD, and RTXPD.
Materials and Methods
Dogs
Six laboratory Beagles (2 females and 4 males) between 1 and 3 years of age and weighing between 9.5 and 13.0 kg were used in this study. All dogs were determined to be healthy with unremarkable heart anatomy and myocardial function on the basis of complete physical, ECG, and standard echocardiographic examinations (including M-mode, pulsed-wave Doppler, and color flow Doppler imaging). All procedures were approved by the Laboratory Animal Experimentation Committee, Graduate School of Veterinary Medicine, Hokkaido University.
Echocardiographic measurements
Conventional echocardiographic examinations were performed by 1 echocardiographer (KN), who used an ultrasonographic machinea equipped with a sector probeb (3 to 7 MHz). Unsedated dogs were manually restrained for evaluation in left and right lateral recumbency. An ECG trace (lead II) was recorded simultaneously with echocardiographic imaging and automatically measured heart rate. All dogs were confirmed healthy by echocardiographic examination.
The RTX was derived from DPD, conventional pulsed-wave Doppler, and tissue Doppler findings by 2 echocardiographers (KN and TM). For this process, RTX was defined as the sum of the ICT and IRT, divided by ejection time (Figure 1). For each RTX, mean values of 3 separate cardiac cycles were used to assess repeatability. Each RTX value was calculated after image acquisition.
Echocardiographic images illustrating a technique used to measure RTX in dogs by means of DPD (A), pulsed-wave Doppler (B and C), and tissue Doppler (D) techniques. A—The upper waveform is tricuspid inflow, and the lower waveform is pulmonary artery flow. B—Tricuspid inflow is shown. C—Pulmonary artery flow is shown. D—Tricuspid valve annular velocity is shown. The simultaneously recorded ECG appears at the top. For all calculations, RTX = (a – b)/b. a = Interval from tricuspid valve closure to opening (pulsed-wave Doppler recordings) or the end of the A' wave to the beginning of the E' wave (tissue Doppler recordings). A = Tricuspid valve late diastolic flow. A' = Tricuspid valve late diastolic velocity. b = Ejection time or S' duration. E = Tricuspid valve early diastolic flow. E' = Tricuspid valve early diastolic annular velocity. S' = Tricuspid valve systolic annular velocity.
Citation: American Journal of Veterinary Research 77, 7; 10.2460/ajvr.77.7.715
To calculate RTXDPD, tricuspid inflow and pulmonary artery flow were measured simultaneously by means of DPD echocardiography with a left parasternal short-axis view, and ICT and IRT were derived by subtracting the ejection time from the amount of time that elapsed between cessation of the tricuspid valve A wave to the onset of the tricuspid valve E wave in 1 image (Figure 1).13 Ejection time was measured from the beginning of one to the beginning of the next pulmonary arterial spectrum.
To calculate RTXPD, tricuspid valve inflow was measured from a left parasternal short-axis view. Next, pulmonary arterial flow was measured from a left parasternal short-axis view. Finally, ICT and IRT were derived by subtracting ejection time from the amount of time that elapsed between cessation of the tricuspid valve A wave (late diastolic flow) to the onset of the tricuspid valve E wave (early diastolic flow) in 2 separate images (Figure 1).14 No attempt was made to match R-R intervals for the in- and outflow signals because, in the authors' experience, it is difficult to match R-R intervals in clinical settings.
To calculate RTXTD, the tricuspid valve peak systolic annular velocity, peak early diastolic velocity, and late diastolic velocity were determined by tissue Doppler echocardiography with an apical 4-chamber view. Then, ICT and IRT were derived by subtracting the duration of the late systolic annular velocity wave for the tricuspid valve from the time that elapsed between the end of the late diastolic velocity wave and onset of the early diastolic velocity wave for the same valve on the basis of tissue Doppler recordings (Figure 1).2
Statistical analysis
Power calculations for sample size determination were made on the basis of data from a previous study.15 Presuming a similar intraobserver within-day, intraobserver between-day, and interobserver ICC, it was estimated that a sample size of 6 dogs would be required to provide a power of 90% to detect an ICC of 0.75,15 with the null hypothesis that the ICC would be 0, an α value of 0.05, and 3 measurements/technique/dog.16
Statistical analysis programsc,d were used to develop a linear mixed model, with measurement time (1 to 9 times), method (DPD, tissue Doppler, and pulsed-wave Doppler techniques), and their interaction as categorical fixed effects and dog identity as a random effect. The F test was performed to assess the effect of measurement time and method on RTX. Multiple comparisons were made by obtaining the LS mean of 1 observer's (KN) measurements and applying the Tukey honest significant difference test to assess differences among methods. The all-pairs Tukey test allows significance tests of all combinations of pairs, and the resulting honest significant difference intervals are greater than those provided with the Student pairwise t test for least significant differences.
The following linear model was used for within- and between-day and interobserver variability analyses17:
where Yijkl was the first value measured for dog k on day j by observer i, μ was the general mean, observeri was the differential effect (considered as fixed) of observer i, dogk was the differential effect of dog k, (observer X dog)ik represented the interaction between the observer and dog, (day X dog) represented the interaction between day and dog, and ϵijkl was the model error. The SD of within-day variability was estimated as the residual SD of the model, SD of between-day variability as the SD of the differential effect of day, and SD of interobserver variability as the SD of the differential effect of observer. The corresponding CVs were determined by dividing each SD by the mean.
The intraobserver within-day ICC was determined from data generated by the same observer (KN); this echocardiographer evaluated 6 dogs 3 times during the same day. The intraobserver between-day ICC was determined from data generated by 1 blinded observer (KN); on each of 3 days, this echocardiographer made 3 evaluations of the 6 dogs. The interobserver ICC was determined from data generated by 2 blinded observers (KN and TM) on the same day; these echocardiographers evaluated 6 dogs 3 times during the same day. Agreement was considered high when the CV was < 20%17,18 and the ICC was > 0.75.15
Differences among measurements derived from the 3 methods were evaluated by use of Bland-Altman analysis, with modification for repeated measures as described elsewhere.19 Mean differences (bias) and 95% CIs were calculated. Differences among methods were considered significant when the 95% CI did not contain 0. Values of P < 0.05 were considered significant. Results are summarized as LS mean (95% CI).
Results
Least squares means for variables associated with RTX as measured by 1 observer using the 3 echocardiographic methods were summarized (Table 1). The RTXTD values were significantly higher than the RTXdpd and RTXPD values. In contrast, RTXDPD did not differ significantly from RTXPD. Intervals between tricuspid valve closure and opening and isovolumic time (sum of ICT and IRT) derived from the tissue Doppler method were longer than respective values derived from the DPD and pulsed-wave Doppler methods. No difference in LS mean heart rate was identified among the 3 methods.
Least squares mean (95% CI) RTX values as assigned to healthy adult Beagles (n = 6) by 1 observer using 3 echocardiographic methods and variables identified through mixed linear modeling as associated with those values.
| Variable | DPD | Tissue Doppler | Pulsed-wave Doppler |
|---|---|---|---|
| RTX | 0.27 (0.23–0.31)a | 0.50 (0.46–0.54)b | 0.25 (0.21–0.29)a |
| TCO (ms) | 252 (236–268)a | 286 (270–292)b | 245 (229–261)a |
| Ejection time (ms) | 199 (189–209)a | 191 (181–201)a | 196 (186–206)a |
| ICT + IRT (ms) | 53 (43–63)a | 95 (85–105)b | 49 (39–59)a |
| Heart rate (beats/min) | 93 (77–109)a | 91 (75–107)a | 94 (78–110)a |
TCO = Interval between tricuspid valve closure and opening.
Values in the same row with different superscript letters are significantly (P < 0.05; Tukey test) different.
Bland-Altman analysis revealed that RTXtd values were significantly higher than RTXDPD and RTXPD values (Figure 2). However, agreement was good between RTXdpd and RTXPD.
Bland-Altman plots of differences among 3 echocardiographic methods for RTX measurement in healthy adult Beagles (n = 6) by 1 observer. A—Bias (95% CI) between RTXtd and RTXdpd was 0.23 (0.18 to 0.28). B—Bias between RTXtd and RTXpd was 0.27 (0.19 to 0.35). C—Bias between RTXdpd and RTXpd was 0.04 (−0.03 to 0.10). Solid horizontal line represents the bias (mean difference), and dashed lines above and below the solid line represent the 95% CI.
Citation: American Journal of Veterinary Research 77, 7; 10.2460/ajvr.77.7.715
Intraobserver within- and between-day and interobserver CVs and ICCs of RTX derived from the 3 methods were summarized (Table 2). The RTXdpd had high (low CV and high ICC) within-day and interobserver repeatability, but between-day repeatability was not high. The RTXTD had high within-day repeatability, but between-day and interobserver repeatability were not high. The RTXpd lacked high within- and between-day repeatability and interobserver repeatability.
Within- and between-day (1 observer) and interobserver (2 observers) CVs and ICCs for 3 echocardiographic methods of RTX measurement in healthy adult Beagles (n = 6) performed 3 times/d for 3 days.
| Within-day | Between-day | Interobserver | ||||
|---|---|---|---|---|---|---|
| Method | CV (%) | ICC | CV (%) | ICC | CV (%) | ICC |
| DPD | 6.1 | 0.77 | 8.4 | 0.73 | 3.5 | 0.83 |
| Tissue Doppler | 6.0 | 0.80 | 7.7 | 0.63 | 24.6 | 0.62 |
| Pulsed-wave Doppler | 20.7 | 0.62 | 20.7 | 0.35 | 19.1 | 0.65 |
Agreement was considered high when the CV was < 20% and the ICC was > 0.75.
Discussion
In the present study, within-day and interobserver repeatability of RTXdpd measurements were high in a small number of healthy Beagles. Findings provided the first description of repeatability of RTXDPD, RTXPD, and RTXTD measurements in dogs and were consistent with values reported for humans.13 The DpD method allows simultaneous recording of Doppler signals at 2 points during the same cardiac cycle; therefore, RTXdpd measurement is not influenced by heart rate fluctuation.13 Because respiratory arrhythmia is common in dogs,20 measurements obtained with the DpD method versus other methods are suggested to be more accurate in that species.
Intraobserver within- and between-day and interobserver repeatability and reproducibility of RTXpd measurement were low in the present study. This is in disagreement with the results of a previous study,14 in which the between-day CV for RTXpd measurement in 55 healthy dogs was 15.3%. This difference may be related to the number of measurements, in that 3 cardiac cycles were used in the present study versus 20 cycles in the other study.
High intraobserver and low interobserver repeatability of RTXTD measurement were obtained in the present study. To date, repeatability and reproducibility of RTXTD measurement in dogs has lacked adequate evaluation. For humans, high21–23 and low24 repeatability and reproducibility of RTXTD measurement have been reported. The RTXTD can also be measured in a single cardiac cycle; therefore, it is not influenced by heart rate fluctuations. Low interobserver repeatability in humans and the dogs of the present study may be attributable in part to that fact that limits of different intervals for tissue Doppler echocardiography are often poorly defined and may be too sensitive to mild changes, such as hemodynamic shifts or slight differences in the obtained images.24
The RTXTD values were higher than the RTXdpd and RTXpd values of the dogs of the present study. This finding was consistent with findings of previous studies involving humans21,22 and dogs.12 The higher RTXTD was mainly attributable to the longer interval between tricuspid valve closure and opening and isovolumic time derived from the tissue Doppler method, compared with values obtained with DpD and pulsed-wave Doppler methods. The reason for differences between RTXTD and other RTX measurements may have been related to differences in methods used and measurement sites.13,22 The RTXTD is measured by use of intervals based on myocardial motion, whereas the RTXdpd and RTXpd are measured by use of intervals based on blood flow.22 Moreover, RTXTD is measured only at the right ventricular inlet portion, in contrast to RTXdpd and RTXPD, which are measured at both the right ventricular inlet and outlet portions. Therefore, RTXTD may be unrelated to the overall right ventricular function.13,22 It is important to consider that RTXTD had higher reference values than did RTXdpd and RTXPD; therefore, Tei indexes should not be used interchangeably.
In humans with right ventricular overload, RTXdpd is a better predictor of exercise capacity than is RTXTD and RTXPD.13 Therefore, in dogs with right ventricular overload, as occurs with pulmonary hypertension, RTXdpd may also be a better predictor of right heart dysfunction. Additional studies are needed to validate the clinical usefulness of RTXdpd measurement in dogs with right heart dysfunction.
The present study had several limitations. First, a small number of healthy laboratory Beagles was used; therefore, caution should be exercised when attempting to extrapolate the repeatability and reproducibility data to dogs with right heart dysfunction. Indeed, in humans, the degree of disagreement among RTX values in patients with heart disease is higher than that in healthy subjects.23 Second, no reference standard of the right ventricular function, such as cardiac catheterization, was evaluated in the present study. Therefore, we could not assess which of the 3 methods for RTX measurement was superior. Additional studies are needed to validate the correlation between RTX and right ventricular function obtained by cardiac catheterization and other noninvasive echocardiographic indices, such as tricuspid valve annular plane systolic excursion or fractional area change. Third, DPD echocardiography is a novel application of ultrasonography that is available on only few ultrasonographic systems, so the usefulness of RTXDPD measurement may be limited in clinical settings.
The study reported here revealed that RTXDPD measurement was a feasible and reliable method for evaluation of cardiac function in a small number of healthy dogs. The RTXDPD values were not significantly different from the RTXPD values; however, RTXtd values were significantly higher than RTXdpd and RTXPD values. Therefore, RTX values derived from different methods should be interpreted with caution and not used interchangeably because values differ with each method. Investigations involving dogs with heart disease are warranted to determine the clinical applicability of RTXDPD measurement.
Acknowledgments
Supported in part by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (No. 25850203).
The authors thank Dr. Yoichi Ito for assistance with the statistical analysis.
ABBREVIATIONS
| CI | Confidence interval |
| CV | Coefficient of variation |
| DPD | Dual pulsed-wave Doppler |
| ICC | Intraclass correlation coefficient |
| ICT | Isovolumic contraction time |
| IRT | Isovolumic relaxation time |
| LS | Least squares |
| RTX | Right ventricular Tei index |
| RTXDPD | Right ventricular Tei index derived from DPD echocardiography |
| RTXTD | Right ventricular Tei index derived from tissue Doppler echocardiography |
| RTXPD | Right ventricular Tei index derived from conventional pulsed-wave Doppler echocardiography |
Footnotes
HI VISION Preirus, Hitachi Medical Corp, Chiba, Japan.
EUP-S52, Hitachi Medical Corp, Chiba, Japan.
JMP, version 8.0, SAS Institute Inc, Cary, NC.
SPSS, version 21, SPSS Inc, Chicago, Ill.
References
1. Kellihan HB, Stepien RL. Pulmonary hypertension in dogs: diagnosis and therapy. Vet Clin North Am Small Anim Pract 2010; 40: 623–641.
2. Rudski LG, Lai WW, Afilalo J, et al. Guidelines for the echocardiographic assessment of the right heart in adults: a report from the American Society of Echocardiography endorsed by the European Association of Echocardiography, a registered branch of the European Society of Cardiology, and the Canadian Society of Echocardiography. J Am Soc Echocardiogr 2010; 23: 685–713.
3. Pariaut R, Saelinger C, Strickland KN, et al. Tricuspid annular plane systolic excursion (TAPSE) in dogs: reference values and impact of pulmonary hypertension. J Vet Intern Med 2012; 26: 1148–1154.
4. Visser LC, Scansen BA, Schober KE, et al. Echocardiographic assessment of right ventricular systolic function in conscious healthy dogs: repeatability and reference intervals. J Vet Cardiol 2015; 17: 83–96.
5. Sieslack AK, Dziallas P, Nolte I, et al. Quantification of right ventricular volume in dogs: a comparative study between three-dimensional echocardiography and computed tomography with the reference method magnetic resonance imaging. BMC Vet Res 2014; 10: 242–255.
6. Tai TC, Huang HP. Echocardiographic assessment of right heart indices in dogs with elevated pulmonary artery pressure associated with chronic respiratory disorders, heartworm disease, and chronic degenerative mitral valvular disease. Vet Med (Praha) 2013; 12: 613–620.
7. Serres F, Chetboul V, Gouni V, et al. Diagnostic value of echo-Doppler and tissue Doppler imaging in dogs with pulmonary arterial hypertension. J Vet Intern Med 2007; 21: 1280–1289.
8. Tei C, Ling LH, Hodge DO, et al. New index of combined systolic and diastolic myocardial performance: a simple and reproducible measure of cardiac function–a study in normals and dilated cardiomyopathy. J Cardiol 1995; 26: 357–366.
9. Teshima K, Asano K, Iwanaga K, et al. Evaluation of right ventricular Tei index (index of myocardial performance) in healthy dogs and dogs with tricuspid regurgitation. J Vet Med Sci 2006; 68: 1307–1313.
10. Paradies P, Spagnolo PP, Amato ME, et al. Doppler echocardiographic evidence of pulmonary hypertension in dogs: a retrospective clinical investigation. Vet Res Commun 2014; 38: 63–71.
11. Hori Y, Kano T, Hoshi F, et al. Relationship between tissue Doppler-derived RV systolic function and invasive hemodynamic measurements. Am J Physiol Heart Circ Physiol 2007;293:H120–H125.
12. Hori Y, Kunihiro S, Hoshi F, et al. Comparison of the myocardial performance index derived by use of pulsed Doppler echocardiography and tissue Doppler imaging in dogs with volume overload. Am J Vet Res 2007; 68: 1177–1182.
13. Choi JO, Choi JH, Lee HJ, et al. Dual pulsed-wave Doppler tracing of right ventricular inflow and outflow: single cardiac cycle right ventricular Tei index and evaluation of right ventricular function. Korean Circ J 2010; 40: 391–398.
14. Baumwart RD, Meurs KM, Bonagura JD, et al. Tei index of myocardial performance applied to the right ventricle in normal dogs. J Vet Intern Med 2005; 19: 828–832.
15. Rosner B. Multisample inference. In: Rosner B. Fundamentals of biostatistics. 7th ed. Boston: Brooks/Cole, 2010;568–571.
16. Walter SD, Eliasziw M, Donner A. Sample size and optimal designs for reliability studies. Stat Med 1998; 17: 101–110.
17. Chetboul V, Athanassiadis N, Concordet D. Observer-dependent variability of quantitative clinical endpoints: the example of canine echocardiography. J Vet Pharmacol Ther 2004; 27: 49–56.
18. Simpson KE, Devine BC, Gunn-Moore DA, et al. Assessment of the repeatability of feline echocardiography using conventional echocardiography and spectral pulse-wave Doppler tissue imaging techniques. Vet Radiol Ultrasound 2007; 48: 58–68.
19. Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res 1999; 8: 135–160.
20. Matsunaga T, Harada T, Mitsui T, et al. Spectral analysis of circadian rhythms in heart rate variability of dogs. Am J Vet Res 2001; 62: 37–42.
21. Acharya G, Pavlovic M, Ewing L, et al. Comparison between pulsed-wave Doppler- and tissue Doppler-derived Tei indices in fetuses with and without congenital heart disease. Ultrasound Obstet Gynecol 2008; 31: 406–411.
22. Duzenli MA, Ozdemir K, Aygul N, et al. Comparison of myocardial performance index obtained either by conventional echocardiography or tissue Doppler echocardiography in healthy subjects and patients with heart failure. Heart Vessels 2009; 24: 8–15.
23. Kakouros N, Kakouros S, Lekakis J, et al. Tissue Doppler imaging of the tricuspid annulus and myocardial performance index in the evaluation of right ventricular involvement in the acute and late phase of a first inferior myocardial infarction. Echocardiography 2011; 28: 311–319.
24. Rojo EC, Rodrigo JL, Pérez de Isla L, et al. Disagreement between tissue Doppler imaging and conventional pulsed wave Doppler in the measurement of myocardial performance index. Eur J Echocardiogr 2006; 7: 356–364.
