Eur J Cardiothorac Surg 2006;29:978-982
© 2006 Elsevier Science NL
Assessment of left ventricular ejection fraction may invalidate the reliability of EuroSCORE
Carl-Johan Jakobsen
*
,
Peter Torp,
Erik Sloth
Department of Anaesthesia & Intensive Care, Aarhus University Hospital, Skejby Sygehus, DK-8200 Aarhus N, Denmark
Received 20 December 2005;
received in revised form 7 February 2006;
accepted 8 February 2006.
* Corresponding author. Tel.: +45 89498751; fax: +45 89498809. (Email: cjj{at}dadlnet.dk).
 |
Abstract
|
|---|
Objective: To evaluate the impact of different echocardiographic methods for assessment of left ventricular dysfunction on the predictive value of EuroSCORE and thus its reliability in daily practice and benchmarking. Methods: Twenty-five recordings obtained by transthoracic echocardiography, optimal for study using four different echocardiographic methods, were examined and placed into relevant EuroSCORE classes. The four methods were eyeballing, Wall Motion Index, Simpson 2D volumetric method and Motion mode. Results: Bland and Altman plots showed that the three specific methods were not fully interchangeable. Using eyeballing as the key reference, only 44% of patients were placed in the same EuroSCORE class with all four methods. The quantitative echocardiographic methods in general overestimate the ejection fraction compared to eyeballing. Conclusion: The left ventricular dysfunction factor in the EuroSCORE risk assessment system needs a more precise definition.
Key Words: Risk assessment • Risk factor • Cardiac surgery • Mortality
|
1. Introduction
|
|---|
As a result of the need for cost containment in health care, providers of care are increasingly expected to produce reliable information on the cost-effectiveness of their procedures.
The European System for Cardiac Operative Risk Evaluation (EuroSCORE) was developed between 1995 and 1999 to provide a simple, additive risk model in European adult cardiac surgery [1–3] and has gained wide acceptance in Europe and elsewhere. Although primarily developed for preoperative risk assessment, it is now widely used for benchmarking the operative results, both between and within institutions [4–6].
Left ventricular dysfunction (LVD) is the most significant, non-acutely occurring, cardiac-related risk factor in the EuroSCORE system [1–3], and also one of the presumed objective parameters. However, on closer examination, the EuroSCORE definitions of cardiac dysfunction give cause for concern, as the methods of measurement may be far less objective and accurate than hitherto accepted. A reduction in left ventricular ejection fraction (LVEF) from 30% to 50% to <30% gives an increase in the score attributed to cardiac dysfunction from 1 to 3. As shown in Fig. 1
, the assessment of cardiac dysfunction may greatly influence the predicted mortality. With a basis score of 0–12 before adding the score for cardiac dysfunction, it is clear that increasing the overall score by 2 may change the predicted mortality by more than 200%. Therefore, the assessment of LVEF is crucial for the predictive value of EuroSCORE.
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 1. The logistic mortality of EuroSCORE 0–12 (basis), added the score for moderate (1 point) or poor (3 points) ventricular function (according to EuroSCORE definitions). The impact on logistic mortality is up to 200% compared with the mortality of the basis score.
|
|
LVEF is not synonymous with myocardial contractility but is, however, in daily clinical practice, considered to be the most valid measure of LV systolic function. Four echocardiographic methods are of particular interest for assessment LVEF; m-mode, Simpson 2D-volumetric method (S2D), Wall Motion Index (WMI) and eyeballing (EYE).
Motion mode (MM) as the longest established method is a one-dimensional recording where LVEF is calculated from changes in chamber dimensions obtained from a parasternal long axis view (Fig. 1). From the systolic and diastolic dimensions, the fractional shortening (FS) is calculated as (LV diastolic diameter – LV systolic diameter)/(LV diastolic diameter) x 100. FS x 2 is a rough measure of EF. Since this is a one-dimensional analysis, m-mode is very sensitive to unambiguous chamber geometry such as LV aneurysm or chamber dilatation will cause a reduction in FS and consequently EF [7]. Most echo machines, including the one used in this study, enable estimates of EF from m-mode recordings by means of the Teicholtz formula.
Simpson 2D-volumetric method is based on endocardial tracing of an apical 2- and 4-chamber image. This method is not sensitive to chamber geometry and also used in invasive ventriculography [8].
Wall Motion Index divides the LV into either 9 or 16 segments, which are individually scored from –1 (dyskinesia) to 3 (hyperkinesia) according to both wall motion and thickening. The total score divided by the number of segments gives WMI [9]. WMI multiplied by 30 reveals EF [10]. An advantage of this method is that it provides information on both regional and global contractility. However, like the S2D method, the analysis is time consuming and therefore not routinely used.
Eyeballing from a combination of parasternal and apical images is the most commonly used echocardiographic method in daily clinical practice. If performed by an experienced echocardiographic technician, eyeballing may even be competitive to volumetric methods of assessment in degree of accuracy [11].
The technological development during the last decade has noticeably improved the quality of the ultrasound image, e.g. second harmonic imaging [12]. Likewise, dedicated software for off-line analysis of the echocardiographic raw data has introduced new features that are expected to enhance image interpretation, compared with analysis obtained from videotapes and using old transducer technology.
We therefore hypothesized that echocardiographic raw data obtained with modern technology followed by off-line analysis using dedicated software could be used to test the consistency of different methods for LVEF assessment. The aim of this study was to evaluate the impact of four different and commonly used echocardiographic methods for assessment of LVEF on the predictive value of EuroSCORE, and thus its reliability in daily practice and benchmarking.
|
2. Method
|
|---|
After institutional approval, 25 preoperative standard transthoracic echocardiographic examinations were randomly selected from the clinical and scientific database at Skejby Sygehus, Aarhus University Hospital. All the patients were adults, and with the exception of two subjects who did not undergo cardiac surgery, scheduled for coronary bypass grafting (14) or aortic valve replacement (9). The 25 recordings were optimal for the study of all four methods and with a range of degrees of cardiac function. The recordings were examined by an experienced echocardiographic technician, and were then categorized by eyeballing into normal, moderate, or poor cardiac function.
A Vivid 7 echo-machine and a 2.5 MHz matrix transducer with second harmonic imaging were used for data acquisition (GE Healthcare, Horten, Norway). Echocardiographic raw data of the apical 4, 2, and long axis view were stored in cineloop format triggered on the R-wave in the corresponding ECG, and analysis of three consecutive sinus heart beats was performed using dedicated software (EchoPac, GE Healthcare, Horten, Norway)
Comparison of the methods was based on the definitions of cardiac dysfunction in EuroSCORE, where the categories are poor (LVEF <30%), moderate (LVEF 30–50%), and normal (LVEF >50).
Statistical analysis was based on the methods described by Bland and Altman [13] for assessing agreement between two methods of clinical measurements. The bias or mean differences and 95% safety limits (1.96 standard deviations) were calculated. The relative error, defined as 100 x (Method 1 – Method 2)/(Method 1 + Method 2)/2) was computed together with standard deviations of the relative error to calculate 95% limits of absolute measurements. Parametric statistics were used to compare the absolute values and results (paired samples t-test or Fisher's exact test). All tests were carried out using MedCalc software.
|
3. Results
|
|---|
The average ejection fractions from each of the three quantitative/volumetric methods are shown in Table 1
. In general, Wall Motion Index-16 showed the lowest ejection fractions, although not significantly so in comparison with Simpson 2D volumetric (p
= 0.29). The m-mode was significantly different from both WMI (p
= 0.010) and S2D (p
= 0.0138).
The bias (mean differences) and 95% limits of agreement (1.96 standard deviations) are shown in Fig. 2
. The absolute bias ranged from 1.6% points (S2D vs WMI) to 7.7% points (MM vs WMI) and with 95% safety limits ranging from 13.5% points to 20.5% points.
View larger version (12K):
[in this window]
[in a new window]
|
Fig. 2. Bland and Altman plots of echocardiographic methods. Bias or mean differences and 95% safety limits (1.96 standard deviations) are MM and 2SD 6.1 ± 20.5, MM and WMI 7.7 ± 18.8 and S2D and WMI 1.6 ± 13.5. (MM = Motion mode, S2D = Simpson 2D-volumetric method, WMI = Wall Motion Index and SD = standard deviation).
|
|
Table 2
shows the comparison between eyeballing and the three echocardiographic quantitative methods. Eleven (44%) were grouped in the same EuroSCORE class by all the measurement methods. If considering eyeballing as the standard method, there was significant difference between eyeballing and Motion mode (p
< 0.001), eyeballing and Simpson 2D (p
= 0.002). There was no significant difference in EuroSCORE class between eyeballing and Wall Motion Index (p
= 0.48) and Wall Motion Index and Simpson 2D (p
= 0.073) (all Fisher's exact test).
View this table:
[in this window]
[in a new window]
|
Table 2. The results of four evaluation methods for assessment of ejection fractions, diagnosis and operative procedure
|
|
|
4. Discussion
|
|---|
Echocardiography presents at least four well-defined methods for measuring the left ventricular ejection fraction [6–9], although our data show that the methods are not totally comparable. Several studies have established eyeballing as equivalent or even superior to volumetric methods [14,15] and we therefore use this as the key method in our study.
If one considers the normal variation of 10–15% (95% safety limits) between the individual echocardiographic quantitative methods, the Bland and Altman plots show that the measurements methods are not in agreement, although the Simpson 2D and Wall Motion Index are within acceptable clinical limits. Both of these appear, therefore, to be feasible methods for evaluating LVEF. However, only Wall Motion Index shows results which overall are not statistically significant compared with eyeballing (Table 2).
The fact that there is agreement between all the methods in EuroSCORE classification in only 44% of the assessed recordings is worrying. Furthermore, in patients with poor cardiac function by eyeballing, only 20% of the recordings show agreement between the three echocardiographic quantitative methods with regard to EuroSCORE classification. In the classes moderate dysfunction and normal cardiac function, 54% and 57% of the recordings, respectively, show consistency between the three quantitative methods.
In general, the three quantitative methods, given a ‘baseline’ figure of ejection fraction obtained by eyeballing, overestimate the ejection fraction comparing to eyeballing. In only three cases (12%) was cardiac function underestimated while in 12 cases (44%) cardiac function was overestimated.
EF assessed by radionuclide technology is often considered the golden standard. However, many centers, including ours, no longer perform angiography routinely thus enhancing the demands on the different echocardiographic algorithms for assessment of EF.
All the echo-loops in this study were analyzed by the same skilled echocardiographer with more than 10 years experience in clinical and research echocardiography. In addition, all data were recorded on high-end echo machines, using professional software. That both data acquisition and analysis have taken place under optimal conditions further undermines the role of LVEF in EuroSCORE.
Ultimately, an inaccurate assessment LVEF obtained by an imprecise method may result in a EuroSCORE of limited value. The maximal impact on the EuroSCORE of just two points either way gives a total effect on the logistic mortality of 1.94 [3], while a single point difference changes the logistic mortality by a factor of 1.51.
|
5. Conclusion
|
|---|
The inclusion of the assessment of left ventricular ejection fraction, without more precise definition, may reduce the reliability of EuroSCORE. The impact on the individual patient when EuroSCORE is used as a preoperative risk assessment tool, as originally envisaged, might be acceptable. However, since EuroSCORE is also used as a benchmarking tool for making performance comparisons both between surgeons and between centers, the findings that more than 50% of patients may be placed in the wrong EuroSCORE class, dependent on method of measurement, might have an unacceptable influence on both the average and the total predicted mortality.
Our study shows that even when the assessment is performed by an experienced echocardiographic technician, using the best technology and applying four standard methods, the impact on the EuroSCORE is ambiguous. The development of a more suitable definition of cardiac dysfunction for use in the EuroSCORE system must, therefore, be regarded as a matter of importance.
|
References
|
|---|
- Roques F, Nashef SA, Michel P, Gauducheau E, de Vincentiis C, Baudet E, Cortina J, David M, Faichney A, Gabrielle F, Gams E, Harjula A, Jones MT, Pintor PP, Salamon R, Thulin L. Risk factors and outcome in European cardiac surgery: analysis of the Euro-SCORE multinational database of 19,030 patients. Eur J Cardiothorac Surg 1999;15:816-822.[Abstract/Free Full Text]
- Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European system for cardiac operative risk evaluation (Euro-SCORE). Eur J Cardiothorac Surg 1999;16:9-13.[Abstract/Free Full Text]
- Roques F, Michel P, Goldstone AR, Nashef SA. The logistic EuroSCORE. Eur Heart J 2003;24(9):882-883.[Free Full Text]
- Pitkänen O, Niskanena M, Rehnberg S, Hippeläinen M, Hynynena M. Intra-institutional prediction of outcome after cardiac surgery: comparison between a locally derived model and the EuroSCORE. Eur J Cardiothorac Surg 2000;18:703-710.[Abstract/Free Full Text]
- Nashef SAM, Roques F, Hammill BG, Peterson ED, Michel P, Grover FL, Wyse RKH, Ferguson TB. Validation of European System for Cardiac Operative Risk Evaluation (EuroSCORE) in North American cardiac surgery. Eur J Cardiothorac Surg 2002;22:101-105.[Abstract/Free Full Text]
- Roques F, Nashef SAM, Michel P, Pintor PP, David M, Baudet E, The EuroSCORE Study Group Does EuroSCORE work in individual European countries?. Eur J Cardiothorac Surg 2000;18:27-30.[Abstract/Free Full Text]
- Teichholz LE, Kreulen T, Herman MV, Gorlin R. Problems in echocardiographic volume determinations: echocardiographic–angiographic correlations in the presence or absence of asynergy. Am J Cardiol 1976;37:7-11.[CrossRef][Medline]
- Schiller NB. Two-dimensional echocardiographic determination of left ventricular volume, systolic function and mass. Summary and discussion of the 1989 recommendations of the American Society of Echocardiography. Circulation 1991;84(Suppl. I):I-280-I-287.
- Schiller NB, Shah PM, Crawford M, DeMaria A, Devereux R, Feigenbaum H, Gutgesell HJ. Recommendations for quantitation of the left ventricle by two-dimensional echocardiography. American Society of Echocardiography committee on standards, subcommittee on quantitation of two-dimensional echocardiograms. Am Soc Echocardiogr 1989;2:358-367.
- Berning J, Rokkedal J, Launbjerg J, Fogh J, Mickley H, Andersen PE. Rapid estimation of left ventricular ejection fraction in acute myocardial infarction by echocardiographic wall motion analysis. Cardiology 1992;80:257.[Medline]
- Rich S, Sheikh A, Gallastegui J, Kondos GT, Mason T, Lam W. Determination of left ventricular ejection fraction by visual estimation during real-time two-dimensional echocardiography. Am Heart J 1982;3:603-606.
- Caidahl K, Kazzam E, Lidberg J, Neumann-Andersen G, Nordanstig J, Dahlqvist SR, Waldenstrom A, Wikh R. New concept in echocardiography: harmonic imaging of tissue without use of contrast agent. Lancet 1998;352:1264-1270.[CrossRef][Medline]
- Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurements. Lancet 1986;1:307-310.[CrossRef][Medline]
- Jensen-Urstad K, Bouvier F, Höjer J, Ruiz H, Hulting J, Samad B, Thorstrand C, Jensen-Urstad M. Comparison of different echocardiographic methods with radionuclide imaging for measuring left ventricular ejection fraction during acute myocardial infarction treated by thrombolytic therapy. Am J Cardiol 1998;81:538-544.[CrossRef][Medline]
- Amico AF, Lichtenberg GS, Reisner SA, Stone CK, Schwartz RG, Meltzer RS. Superiority of visual versus computerized echocardiographic estimation of radionuclide left ventricular ejection fraction. Am Heart J 1989;118:1259-1265.[CrossRef][Medline]
Top
Abstract
1. Introduction
