Geert Kleinnibbelink

RV Strain in Pulmonary Hypertension 6 155 models 40 including TAPSE and/or RFVAC to increase the predictive value in PH patients. 3, 41 Several studies included in our meta-analysis revealed RVLS to has superior predictive value over RVFAC and TAPSE, indicating that RVLS may be a more sensitive predictor for RV dysfunction. 8, 10, 15 Implementing RVLS in these multi-parameter predictive models therefore may increase their predictive value for future events. In addition to predicting future events, a relative reduction in RVLS might be indicative for (adjustment of ) pharmacological therapy and/or surgery. Improvement of RVLS after pharmacological therapy and/or surgery has shown to be related to lower risks for mortality and PH- related events. 16, 24 These data further support the use of RVLS in clinical practice, as RVLS changes across time are associated to clinically relevant outcomes in PH patients. Future studies determining reference values and confirming clinically-relevant cut-off values are warranted to improve clinical decision-making and implementation of RVLS in practice. Limitations The studies within this meta-analysis were non-uniform in design and varied in the inclusion criteria, methods to measure RVLS (intervendor and technique variabilities), follow-up periods and endpoints. We corrected for these between-study variation using a random effects model in our meta-analysis. Additionally to minimize the impact of intervendor and technique variability we reported the relative reduction of % of RVLS rather than absolute values. We also included studies which used RVFWS (n=7) and RVGLS (n=1) or both (n=3) to determine the predictive value of RVLS in PHpatients. Unfortunately, the small amount of studies investigating RVGLS did not allow for a comparison between the predictive value of RVGLS and RVFWS. Similarly, we were not able to compare data obtained with 2D vs. 3D echocardiography and/or machines from different vendors. Due to differences in methodology and statistical approach, not all relevant studies could be included in our analysis. Studies using ROC-analysis 18, 20, 22, 25, 30 , Kaplan Meier survival curves 18, 19, 22, 25 , odds ratios 20 or predictive models 12, 13, 19 reported outcomes that align with the findings of the present meta-analysis.