• Users Online: 300
  • Print this page
  • Email this page


 
 
Table of Contents
ORIGINAL ARTICLE
Year : 2018  |  Volume : 7  |  Issue : 4  |  Page : 176-181

Sensitivity, specificity, and accuracy of left ventricular systolic function indices and structure in detecting early systolic dysfunction assessed by speckle-tracking two-dimensional strain: An echocardiographic cross-sectional study


1 Department of Internal Medicine, Faculty of Medicine and Biomedical Sciences, University of Yaounde 1, Paris, France
2 Department of Cardiovascular, Institute Mutualiste Montsouris, Paris, France
3 Department of Internal Medicine, Faculty of Medicine and Biomedical Sciences, University of Yaounde 1; Internal Medicine Service, Cardiology Unit, Yaounde Central Hospital, Paris, France
4 Department of Internal Medicine, Faculty of Medicine and Biomedical Sciences, University of Yaounde 1, Paris, France; Internal Medicine Service, Cardiology Unit, Yaounde General Hospital, Yaounde, Cameroon

Date of Web Publication31-Dec-2018

Correspondence Address:
Dr. Ahmadou Musa Jingi
Department of Internal Medicine, Faculty of Medicine and Biomedical Sciences, University of Yaounde 1
France
Login to access the Email id

Source of Support: None, Conflict of Interest: None


DOI: 10.4103/rcm.rcm_18_18

Get Permissions

  Abstract 


Background: Reduced global longitudinal strain (GLS) is an early marker of subclinical left ventricular (LV) dysfunction, permitting timely interventions to slow disease progression. This technique is not widely available in echocardiographs in routine use. Aim: We sought to know if LV systolic function indices and structural left heart changes could predict a reduced GLS. Methods: We carried out a cross-sectional analytic study in May 2017. We measured GLS (reference test), LV ejection fraction, LV midwall shortening (MWS), LV mass index, LV diastolic diameter, LV volumes, and left atrial volume (predictors). We calculated the sensitivity, specificity, accuracy, predictive values, and likelihood ratios of the predictor variables. We assessed the discriminatory power of the indices with the Youden Index and area under the receiver operator characteristic curve (AUC). Results: A total of 32 participants (14 males) were retained for this study. Their mean (standard deviation) age was 62 (15.3) years. Eccentric LV hypertrophy (LVH) was the most frequent LV geometric pattern – 14 (43.8%) participants. A reduced GLS was the most frequent LV functional anomaly – 20 (62.5%) participants. A low MWS <36.5% had a good predictive power of a reduced GLS – sensitivity: 80%, specificity: 83.2%, accuracy: 81.3%, and AUC: 0.817. The presence of LVH had a fair prediction power of reduced GLS – sensitivity: 70%, specificity: 81.8%, accuracy: 65.6%, and AUC: 0.741. The composite of MWS <36.5% and or LVH had a fair discriminatory power (AUC: 0.783, Youden Index: 0.567), with a good sensitivity – 90%. Conclusion: Low MWS of the LV and the presence of LVH were found to be good predictors of reduced GLS.

Keywords: Accuracy, echocardiography, sensitivity, specificity, two-dimensional speckle tracking


How to cite this article:
Jingi AM, Czitrom D, Hamadou B, Mankoubi L, Sondes S, Kingue S. Sensitivity, specificity, and accuracy of left ventricular systolic function indices and structure in detecting early systolic dysfunction assessed by speckle-tracking two-dimensional strain: An echocardiographic cross-sectional study. Res Cardiovasc Med 2018;7:176-81

How to cite this URL:
Jingi AM, Czitrom D, Hamadou B, Mankoubi L, Sondes S, Kingue S. Sensitivity, specificity, and accuracy of left ventricular systolic function indices and structure in detecting early systolic dysfunction assessed by speckle-tracking two-dimensional strain: An echocardiographic cross-sectional study. Res Cardiovasc Med [serial online] 2018 [cited 2019 Jul 23];7:176-81. Available from: http://www.rcvmonline.com/text.asp?2018/7/4/176/249045




  Introduction Top


Reduced strain and strain rate are early markers of subclinical left ventricular (LV) dysfunction in many disease conditions.[1],[2],[3],[4],[5],[6],[7],[8],[9] They have a higher prognostic value than the widely used LV ejection fraction (LVEF) and are powerful independent predictors of vascular events, atrial fibrillation, heart failure, and all-cause mortality.[10],[11],[12],[13],[14],[15],[16] They are used to evaluate for myocardial viability in ischemic heart disease, in postheart transplant patients for acute graft rejection, and in early detection of cardiotoxicity of anticancer medicines.[6],[17] Global longitudinal strain (GLS) is highly reproducible, with low intra- and interobserver variability.[18] This new echocardiography index of heart function is not widely available in echocardiographs used in most low-middle income settings (LMICs). It is not known which of the frequently used indices of LV systolic function best predicts a low GLS. Such data are needed to guide daily clinical practice, especially in LMICs. We carried out this cross-sectional analytic study to determine which LV systolic function index best predicts a low GLS. We also sought to know if structural changes of the left heart can predict a low GLS of the LV.


  Methods Top


Ethical statement

This work was approved by the Ethics Committee of Institut Mutualiste Montsouris (IMM) – Paris (France). This work was carried out in accordance with the Declarations of Helsinki.[19] We report this work according to the Standards for Reporting Diagnostic tests checklist.[20]

Study design and setting

In May 2017, we carried out a cross-sectional descriptive and analytic study in the echocardiography laboratory of IMM in Paris, France. This health institution has a high volume of cardiovascular activities, including noninvasive and invasive cardiology and open-heart surgeries. There are about 27 cardiologists in the cardiovascular unit.

Participants

These were adults aged ≥18 years, of both sexes, who gave oral consent to participate. Participants were patients who were referred to the echocardiography laboratory for cardiac ultrasound. We excluded from the study, the patients who had a history of aortic and valvular surgery (replacement and reparation), heart transplant, severe aortic stenosis, and obstructive hypertrophic cardiomyopathy. We also excluded cases of significant LV outlet obstructions. For each participant, we collected a significant history associated with the request of the ultrasound. We also obtained data on their height and the most recent weight through the interview. These were used to calculate the body mass index and body surface area.

Test methods

All the echocardiograms were prospectively performed by a single and experienced echocardiographist (DC), with the patient in the left lateral decubitus position using a GE Vivid E9 echocardiograph with an M5 ultrasound probe, coupled to the electrocardiography. We carried out the following measurements of LV systolic function indices: GLS, LVEF using the Simpson biplane method of discs, midwall shortening (MWS) – synonymous to fractional shortening (FS), and cardiac output. We also measured the size and dimensions of the chambers. The average of three measurements was used in the analysis.

Working definitions

These were based on the American Society of Echocardiography (ASE) recommendations.[21],[22] Low LVEF <55% (mild: 45%–54%, moderate: 30%–44%, and severe: <30%), low FS (or MWS) <25% (mild: 20%–24%, moderate: 15%–19%, and severe: ≤14%), low cardiac output <3.5 l/min, low GLS<–18%, LV hypertrophy (LVH) – LV mass index (LVMi) >115 g/m2 in men (mild: 116–131 g/m2, moderate: 132–148 g/m2, and severe: ≥149 g/m2) and LVMi >95 g/m2 in women (mild: 96–108 g/m2, moderate: 109–121 g/m2, and severe: ≥122 g/m2), LV dilation – LV diastolic diameter >31 mm/m2 in men and >32 mm/m2 in women, and left atrial (LA) dilation as – LA volume >29 ml/m2.

Sample size and power

The early detection of asymptomatic LV dysfunction is essential for timely interventions to slow the progression of disease. We needed a surrogate marker of reduced GLS with a sensitivity of >90% and a specificity close to 90%. We hypothesized that 60% of high-risk patients seen in the echocardiography laboratory had a reduced GLS. By accepting a margin of error of ≤10% around the 95% confidence interval for the sensitivity, we used the table provided by Buderer[23] to estimate a sample size of at least 58 participants.

Statistical analysis

Data were analyzed using the software SPSS Statistics for Windows, Version 16 (SPSS Inc, Chicago, Ill., USA) and an online calculator (MedCalc version 7). We considered GLS as the reference measurement (standard) to be predicted. We calculated the sensitivity, specificity, accuracy, predictive values, and likelihood ratios of having a low GLS <–18%, using the other LV systolic function indices as well as left heart structural changes (test variables). We compared the performance or discriminating power of the indices by computing the area under the receiver operator characteristic (ROC) curve (AUC) and the Youden Index (sensitivity + specificity – 1). An AUC of 0.9–1 was considered excellent, 0.8–0.9 was good, 0.7–0.8 was fair, and <0.7 was poor. The Youden Index gives the best trade-off of sensitivity and specificity pair for an optimal threshold. The Youden Index closest to 1 indicates a good test performance and 0 indicates a worst test performance for the threshold considered. We present discrete variables as frequencies and proportions and continuous variables as means ± standard deviation (SD). P < 0.05 was considered statistically significant for the observed associations.


  Results Top


Participants

The clinical and echocardiographic characteristics of the study population are summarized in [Table 1] (frequencies and percentages) and [Table 2] (means and standard deviations). A total of 32 participants (14 males) were retained for this study. Their mean (SD) age was 62 (15.3) years and ranged from 29 to 87 years. Eccentric LVH was the most frequent LV geometric pattern, and reduced GLS was the most frequent LV functional anomaly. The mean values of GLS according to different grades of LVEF, MWS, and LVMi are shown in [Figure 1]. There was a significant decrease in the GLS from normal to severely abnormal LVEF and MWS (P < 0.001 and P = 0.001 for trend, respectively). This trend was observed but not significant for LVMi (P = 0.190). The mean values of GLS according to LV geometry pattern are shown in [Figure 2]. There was a trend of GLS from normal to concentric LVH pattern. Those with eccentric LVH pattern had higher values of GLS than those with concentric LVH pattern.
Table 1: Clinical and echocardiographic presentation of the study population in frequencies (n) and percentages (%)

Click here to view
Table 2: Mean values (standard deviation) of the clinical and echocardiographic characteristics

Click here to view
Figure 1: Mean values of global longitudinal strain according to the severity of left ventricular ejection fraction, midwall shortening, and left ventricular mass index

Click here to view
Figure 2: Mean global longitudinal strain (%) according to the left ventricular geometry

Click here to view


Test results

The performance of the routinely used LV function indices and left heart changes in predicting reduced GLS is shown in [Table 3]. Structural left heart changes (LVH and LA dilation) had the highest sensitivities for reduced GLS but had low specificities. Low MWS and the presence of LVH had the highest AUC (P < 0.05, respectively) but had low sensitivity and specificity pair (Youden Index of 0.3 and 0.32, respectively). Cardiac output and LV volumes had the lowest AUC for reduced GLS – AUC: 0.577, P = 0.471 for LV end-diastolic volume and AUC: 0.679, P = 0.094 for LV end-systolic volume. The LV systolic volume had a better predictive power than LV diastolic volume – sensitivity: 35%, specificity: 100%, and accuracy: 59.4% versus sensitivity: 20%, specificity: 91.7%, and accuracy: 46.9% for LV systolic and diastolic volumes, respectively. We derived optimal thresholds for the optimal sensitivity and specificity pair for LVEF, MWS (FS), and LVMi using ROC curves. The derived threshold was similar for LVEF (threshold: <55.5%, sensitivity: 40%, specificity: 99.9%, and Youden Index: 0.399) and LVMi (threshold: >113.5 g/m2, sensitivity: 70%, specificity: 81.8%, and Youden Index: 0.518). The derived optimal cutoff of MWS was higher than the standard cutoff (<36.5% – new threshold vs. <25% – old threshold). The performance of MWS <36.5% in predicting a low GLS is shown in [Table 4]. This had a very good discriminatory power (AUC: 0.817, Youden Index: 0.633) with a sensitivity of 80%. The composite of MWS < 36.5% and or LVH had a fair discriminatory power (AUC: 0.783, Youden Index: 0.567), with a good sensitivity – 90%.
Table 3: Performance of left ventricular function indices and structural heart changes in predicting a reduced (global longitudinal strain <-18%)

Click here to view
Table 4: Test performance of midwall shortening or fractional shortening with or without left ventricular hypertrophy in predicting a reduced (global longitudinal strain <-18%) using optimal cutoff derived from receiver operator characteristic curve

Click here to view



  Discussion Top


Reduced GLS is an early marker of subclinical LV systolic dysfunction. Early detection with appropriate interventions can slow the progression of the disease to overt clinical heart failure. GLS is, however, not widely available in most echocardiographs in routine use. We sought to determine functional and structural heart changes that could predict a low GLS. Low MWS or FS of the LV and the presence of LVH were found to be good predictors of reduced GLS.

Our findings should be interpreted in light of some limitations. Our sample size was small (55% of estimate) due to the limited recruitment time (1 month – the Principal investigator (PI) had 1 month of training on advanced cardiac ultrasound) and the high selectivity of patients (excluding those with heart surgery and those with significant LV outflow tract obstruction). This could limit precisions of our calculations, and also reduce the ability of some of the indices to detect a reduced GLS. Most of our participants were Caucasians in a high-income setting, thus, our findings may not be valid for use in low-income settings, where these indirect markers of reduced GLS are intended for use. The indices we studied are surrogates of GLS. It does not give us an idea of the heart segment with motion anomaly thus not localizing as in the bull's eye image. We used two-dimensional (2D) speckle tracking, which is inferior to three-dimensional (3D) speckle tracking because of the multiple acquisitions of images at different times with 2D speckle. However, the 3D speckle tracking is still to be fully studied and validated. One observer performed all the ultrasounds; thus, we could not assess interobserver variability. However, studies designed to assess intra- and interobserver variability have shown high reproducibility of 2D speckle tracking.[18]

GLS is a new and indispensable index of myocardial function.[7] Many studies have shown the feasibility and robustness of strain and strain rate in detecting subclinical LV dysfunction early in different disease conditions. To the best of our knowledge, no study on the surrogates of low GLS has been carried out. We found LVH to be associated with a reduced GLS. LVH has been shown to be associated with a poor cardiovascular outcome.[24],[25] In this study, those with concentric LVH appeared to have lower GLS than those with eccentric LVH. Those with concentric LVH have been shown to have higher mortality risk than those with eccentric LVH.[26] This suggests that LV geometry pattern can predict GLS, and thus stress the need for detecting LVH and initiate appropriate treatment when GLS cannot be measured.[27] It is not known whether targeting the reduction of LVH with appropriate treatment will improve the GLS. From our experience, measuring LV mass using ultrasound is more challenging than the acquisition of GLS, especially in hypoechoic patients.

MWS or FS seems a simple index to measure in daily clinical practice. This had the best discriminatory power and best predicted a reduced GLS for a threshold of MWS <36.5% compared to LVEF. This finding has not been reported elsewhere to the best of our knowledge. In computing MWS, the difference between the end-diastolic and end-systolic diameters reflects the sum of the septal and posterior wall excursions – radial deformation or strain. Although MWS seems simple to measure, the challenge is cutting through the LV perpendicularly. This also has as a limitation, assessing just the midwall of the anteroseptal and inferolateral segments of the LV. We also found that the LV end-systolic volume index appeared to be a better predictor than the end-diastolic volume index, with a test performance comparable to LVEF. The small sample size does not permit us to conclude on this index.


  Conclusion Top


Low MWS or FS of the LV and the presence of LVH–concentric and eccentric were found to be good predictors of reduced GLS. Our findings need to be confirmed in a larger sample and also need to be validated in a younger population in a low-income setting.

Acknowledgments

The authors would like to thank the support staff of the echocardiography laboratory for assisting with the patients. The authors also thank the patients for consenting to participate in advancing knowledge in science.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.



 
  References Top

1.
Nesbitt GC, Mankad S. Strain and strain rate imaging in cardiomyopathy. Echocardiography 2009;26:337-44.  Back to cited text no. 1
    
2.
Tadic M, Pieske-Kraigher E, Cuspidi C, Genger M, Morris DA, Zhang K, et al. Left ventricular strain and twisting in heart failure with preserved ejection fraction: An updated review. Heart Fail Rev 2017;22:371-9.  Back to cited text no. 2
    
3.
Mada RO, Duchenne J, Voigt JU. Tissue Doppler, strain and strain rate in ischemic heart disease “how I do it”. Cardiovasc Ultrasound 2014;12:38.  Back to cited text no. 3
    
4.
Hoit BD. Strain and strain rate echocardiography and coronary artery disease. Circ Cardiovasc Imaging 2011;4:179-90.  Back to cited text no. 4
    
5.
Dandel M, Hetzer R. Echocardiographic strain and strain rate imaging – Clinical applications. Int J Cardiol 2009;132:11-24.  Back to cited text no. 5
    
6.
Dandel M, Lehmkuhl H, Knosalla C, Suramelashvili N, Hetzer R. Strain and strain rate imaging by echocardiography - basic concepts and clinical applicability. Curr Cardiol Rev 2009;5:133-48.  Back to cited text no. 6
    
7.
Mirea O, Duchenne J, Voigt JU. Recent advances in echocardiography: Strain and strain rate imaging. F1000Res 2016;5. pii: F1000 Faculty Rev-787.  Back to cited text no. 7
    
8.
Yip G, Abraham T, Belohlavek M, Khandheria BK. Clinical applications of strain rate imaging. J Am Soc Echocardiogr 2003;16:1334-42.  Back to cited text no. 8
    
9.
Wang Q, Gao Y, Tan K, Li P. Subclinical impairment of left ventricular function in diabetic patients with or without obesity: A study based on three-dimensional speckle tracking echocardiography. Herz 2015;40 Suppl 3:260-8.  Back to cited text no. 9
    
10.
Zhang KW, French B, May Khan A, Plappert T, Fang JC, Sweitzer NK, et al. Strain improves risk prediction beyond ejection fraction in chronic systolic heart failure. J Am Heart Assoc 2014;3:e000550.  Back to cited text no. 10
    
11.
Russo C, Jin Z, Elkind MS, Rundek T, Homma S, Sacco RL, et al. Prevalence and prognostic value of subclinical left ventricular systolic dysfunction by global longitudinal strain in a community-based cohort. Eur J Heart Fail 2014;16:1301-9.  Back to cited text no. 11
    
12.
Russo C, Jin Z, Sera F, Lee ES, Homma S, Rundek T, et al. Left ventricular systolic dysfunction by longitudinal strain is an independent predictor of incident atrial fibrillation: A Community-based cohort study. Circ Cardiovasc Imaging 2015;8:e003520.  Back to cited text no. 12
    
13.
Russo C, Jin Z, Sera F, Lee ES, Hommes, Rundek T, et al. global longitudinal strain is an independent predictor of incident atrial fibrillation in the elderly: A community-based cohort study. Circulation 2014;130 Suppl 2:A 13905.  Back to cited text no. 13
    
14.
Kalam K, Otahal P, Marwick TH. Prognostic implications of global LV dysfunction: A systematic review and meta-analysis of global longitudinal strain and ejection fraction. Heart 2014;100:1673-80.  Back to cited text no. 14
    
15.
Krishnasamy R, Isbel NM, Hawley CM, Pascoe EM, Burrage M, Leano R, et al. Left ventricular global longitudinal strain (GLS) is a superior predictor of all-cause and cardiovascular mortality when compared to ejection fraction in advanced chronic kidney disease. PLoS One 2015;10:e0127044.  Back to cited text no. 15
    
16.
Stanton T, Leano R, Marwick TH. Prediction of all-cause mortality from global longitudinal speckle strain: Comparison with ejection fraction and wall motion scoring. Circ Cardiovasc Imaging 2009;2:356-64.  Back to cited text no. 16
    
17.
Thavendiranathan P, Poulin F, Lim KD, Plana JC, Woo A, Marwick TH, et al. Use of myocardial strain imaging by echocardiography for the early detection of cardiotoxicity in patients during and after cancer chemotherapy: A systematic review. J Am Coll Cardiol 2014;63:2751-68.  Back to cited text no. 17
    
18.
Yamada A, Luis SA, Sathianathan D, Khandheria BK, Cafaro J, Hamilton-Craig CR, et al. Reproducibility of regional and global longitudinal strains derived from two-dimensional speckle-tracking and Doppler tissue imaging between expert and novice readers during quantitative dobutamine stress echocardiography. J Am Soc Echocardiogr 2014;27:880-7.  Back to cited text no. 18
    
19.
World Medical Association declaration of Helsinki. Recommendations guiding physicians in biomedical research involving human subjects. JAMA 1997;277:925-6.  Back to cited text no. 19
    
20.
Bossuyt PM, Reitsma JB, Bruns DE, Gatsonis CA, Glasziou PP, Irwig LM, et al. The STARD statement for reporting studies of diagnostic accuracy: Explanation and elaboration. Ann Intern Med 2003;138:W1-12.  Back to cited text no. 20
    
21.
Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: An update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. Eur Heart J Cardiovasc Imaging 2015;16:233-70.  Back to cited text no. 21
    
22.
Yingchoncharoen T, Agarwal S, Popović ZB, Marwick TH. Normal ranges of left ventricular strain: A meta-analysis. J Am Soc Echocardiogr 2013;26:185-91.  Back to cited text no. 22
    
23.
Buderer NM. Statistical methodology: I. Incorporating the prevalence of disease into the sample size calculation for sensitivity and specificity. Acad Emerg Med 1996;3:895-900.  Back to cited text no. 23
    
24.
Hoang K, Zhao Y, Gardin JM, Carnethon M, Mukamal K, Yanez D, et al. LV mass as a predictor of CVD events in Older adults with and without metabolic syndrome and diabetes. JACC Cardiovasc Imaging 2015;8:1007-15.  Back to cited text no. 24
    
25.
Milani RV, Lavie CJ, Mehra MR, Ventura HO, Kurtz JD, Messerli FH, et al. Left ventricular geometry and survival in patients with normal left ventricular ejection fraction. Am J Cardiol 2006;97:959-63.  Back to cited text no. 25
    
26.
Ghali JK, Liao Y, Cooper RS. Influence of left ventricular geometric patterns on prognosis in patients with or without coronary artery disease. J Am Coll Cardiol 1998;31:1635-40.  Back to cited text no. 26
    
27.
Cuspidi C, Meani S, Valerio C, Fusi V, Sala C, Maisaidi M, et al. Effects of angiotensin II receptor blockade-based therapy with losartan on left ventricular hypertrophy and geometry in previously treated hypertensive patients. Blood Press 2006;15:107-15.  Back to cited text no. 27
    


    Figures

  [Figure 1], [Figure 2]
 
 
    Tables

  [Table 1], [Table 2], [Table 3], [Table 4]



 

Top
 
  Search
 
    Similar in PUBMED
   Search Pubmed for
   Search in Google Scholar for
 Related articles
    Access Statistics
    Email Alert *
    Add to My List *
* Registration required (free)  

 
  In this article
Abstract
Introduction
Methods
Results
Discussion
Conclusion
References
Article Figures
Article Tables

 Article Access Statistics
    Viewed492    
    Printed99    
    Emailed0    
    PDF Downloaded85    
    Comments [Add]    

Recommend this journal