5. Limitations
Present data was obtained from a single centre and from a limited number of patients, so it should be regarded as a “hypothesis-generating” study. This limited sample size also drastically reduces the power of the study and therefore limits the reliability of the results. The retrospective design of the study prevents making detailed analyses on the causes of death for patients within the third PDI tertile to ascertain whether these cases had RVF immediately before their death. The sample size is too small to compare PDI with other established risk predictors and risk scores, so it is unclear whether PDI offers incremental usefulness over existing methods. While this was an “all-comers” study, patients who were deemed to have at high risk for RVF did not undergo LVAD implantation and that constitutes an unavoidable “selection bias”. As aforementioned, this bias may also explain why a low PDI failed to predict RVF, contrary to our initial expectations. Finally, as this analysis is retrospective, not all confounders might have been accounted for.