An ancillary survival analysis indicated that patients without LVEF assessment had worse survival. As in any observational study, residual confounding may account for this association. However, this underscores that the use of clinically indicated tests to measure the prevalence of LV dysfunction after MI and/or to ascertain outcomes is confounded by indication and will likely provide biased estimates. Indeed, estimates of the true prevalence of LV systolic dysfunction vary depending on the value of LV function in the group in whom it was not measured. To this end, we reported that the proportion of patients with preserved LVEF after MI among subjects with post-MI heart failure could vary notably, depending on assumptions for LVEF among patients without LV function measured.
These results provide important insights into the patterns of practice with regards to measurements of EF after MI. Potential limitations should be kept in mind, however, while interpreting the data. The indications for measuring LVEF by noninvasive and invasive methods are different. The purpose of this article was to report on the use of all tests in order to ascertain the frequency with which the information is available for decision making regarding the use of evidence-based therapies in MI. Thus, the respective indications for the use of each method are beyond the scope of this report and were not addressed.