This study is limited by its retrospective nature. In assessing
for pre-injury anticoagulation, we relied on the surrogate
marker of elevated INR, which may not be accurate in all cases.
Further, we were unable to identify the contribution of additional
medications the patients were prescribed or taking at the time
of trauma, which may have an additional impact on outcome.
We attempted to control for the baseline health status of the
patient using the previously validated CCI, however this relies
on only the comorbidities that are coded in the trauma database.
While regular audits of the validity of our trauma database are
preformed, and our coding is accurate, the data are limited to
what is recorded. Finally, we are unable to provide insight into
what happened to patients once they were discharged from
LHSC, and it is possible that some patients discharged to another
hospital eventually returned home independently.
We believe the results of this study have import