This study demonstrates that PSV can be measured in most patients with a low interobserver variability and may even be a better predictor of outcome than both traditional echocardiographic measurements of left ventricular function, such as EF, WMS and E/e'-ratio and more recent deformation parameters, such as myocardial strain in ACS patients.
The aim of this study was to test PSV in a real world setting. We therefore included consecutive unselected ACS patients and all images were acquired as in the clinical routine. In the univariable analyses, the median PSV value was lower in those with than in those without a subsequent cardiac event. When categorizing patients according PSV values, a low-, intermediate- and high-risk group could be identified with a 5-year cumulative risk of death, MI or readmission because of heart failure of 17%, 32% and 64%, respectively. Even after adjusting for differences in well-known predictors of outcome (including natriuretic peptides and estimates of kidney function), there was a strong association between PSV and outcome, with a 1.4 fold increased risk of the combined endpoint for every unit decrease in PSV.
When the prognostic value of PSV was compared with that of other echocardiographic parameters in a ROC analysis, PSV had the largest AUC although the differences were not statistical significant. When the echocardiographic parameters were tested one by one in a regression model including well-known predictor of outcome, only PSV and E/e'-ratio remained associated with outcome, whereas EF, WMS and strain failed to independently predict outcome. Finally, when PSV was included in the model, none of the other echocardiographic parameters carried any additional prognostic information.
The predictive value of different echocardiographic parameters may vary in different patient populations. Wang et al.  demonstrated a better prognostic value for both é, PSV and E/e' than for EF and WMS. In that study e' was the strongest predictor of cardiac death although PSV and E/e' also were strong predictors of cardiac death. However, the study by Wang and co-workers included patients with a broad spectrum of heart disease and only 16% had ischemic heart disease. The fact that the diastolic parameter e' was demonstrated to have a higher prognostic value than PSV might be explained by a relatively small proportion of patients having ischemic heart disease and a larger proportion of patients having heart failure. In such a population the proportion of patients with high filling pressure is expected to be higher than in our study population, which included ACS patients with predominantly normal or only mildly depressed EF.
Our study does not only confirm but also extends previous findings showing that E/e' has an incremental prognostic value to that of natriuretic peptides [8, 9]. Somewhat unexpectedly 2D-strain was not better to predict outcome than traditional measurements of systolic LV function such as EF and WMS. This is in contrast to earlier studies in the field. In a cohort of 649 ACS patients Antoni and co-workers showed that both strain and strain rate were superior to the traditional parameters, EF and WMS, to predict 1-year outcome  and similar finding were made by Bertoni and co-workers but in a population with chronic ischemic heart disease . In our study, we could not confirm a superiority of deformation parameters in comparison to EF and WMS. Clearly, the operator’s skills and experience, and their compliance to strict protocols are important determinants of the quality of echocardiographic examinations. It is important to note that the images in our study was collected from routine clinical echocardiography not always of highest quality, but still sufficient for analyzing PSV in a very large proportion of the patients, whereas in the study of Antoni et al. and Bertoni et al., all exams were performed within a dedicated laboratory with great experience of conducting studies and using deformation analysis. Still, even if a technique is good in the hands of experts in specialized centers, it can be problematic in the more generalized everyday clinical situation if the demand for technical skill is too high .
Therefore, the simplicity of PSV must be regarded as a great advantage, which can be an important determinant to why PSV in our study is superior to the other parameters. Another advantage is the fact that the method seems robust and insensitive to poor image quality with a low interobserver variability and low number of missing values. Thus the results of this study indicates that PSV could be superior to other parameters as a clinical routine method outside a core laboratory to predict outcome. Furthermore the strong correlation (0.91) between global PSV from all 6 basal segments and PSV from only septal and lateral wall indicates that the use of only one projection and two measurement might be as predictive as the global PSV which would further simplifiy the method.
The present study has some limitations. The sample size was rather small. Thus, lack of significant differences may still be caused by lack of power to detect such differences and our findings need to be confirmed in larger studies. Although interobserver variability was assessed for PSV in a subgroup of patients and all analyses were performed according to a protocol, only one person performed the echocardiographic analyses. Due to the clinical setting and image quality, strain rate was not included in this study. Strain rate is the derivate of strain and therefore it is reasonable to believe that the variability and number of missing values would be even higher compared to strain.
This study used a prespecified combined endpoint. Although PSV was significantly associated with all individual endpoints, the underlying mechanisms for these associations may differ. Systolic LV-dysfunction is known to cause both heart failure and death, whereas the association between systolic LV-dysfunction and subsequent risk of new MI may be explained by the association between systolic LV-dysfunction and severity of coronary artery disease.
In conclusion, PSV seems to be a robust and easily obtained echocardiographic measurement that may be very useful for risk stratifying ACS patients.