Comparisons of fatigue performance between datasets are relatively common practice, although there is not a standardized metric for conducting such analyses. In this work, two recently published nitinol fatigue datasets with relevance for cardiovascular medical devices are used to probe a range of comparative metrics. Results highlight the importance of selecting an appropriate metric as well as the importance of test planning as both can affect the final conclusion regarding the similarity of two datasets. Additionally, an unsuccessful attempt was made to predict high cycle fatigue strength based on low cycle (i.e., censored) fatigue data using a Coffin-Manson relation, which reinforces the importance of conducting fatigue tests to as many cycles as the intended device lifetime. Lessons learned from this set of analyses can inform future standards work as well as engineering practice.