Limitations
Our study should be reviewed with certain limitations to our analysis.
Since we used admission-associated data rather than patient-level data
we depended on administrative diagnosis codes. Therefore, the outcomes
among the patients who were discharged cannot be defined due to limited
information of follow-up. We were also unable to account for unmeasured
confounders that could be associated with increased risk of TTS in AIS
patients and the complications occurring during hospital stay. Data on
medication records and follow-up were not available. Despite these
limitations, the benefit of using a large, multi-center, representative
sample affords a truly national perspective and preserves the validity
of associations reported.