Furthermore, our study did have some limitations. Firstly, this study was based on an administrative database and lacks patient-level clinical information. NIS data regarding bipolar disorder are limited to inpatient hospitalization only and do not include any data from outpatient settings. Re-hospitalizations, which add to the total inpatient burden, are not accounted for in our study, given the nature of the data. Selection bias is possible, given the retrospective nature of the study. The prevalence of comorbidities in the study participants may differ than that seen in the general population and other bipolar populations, as our participants were selected from the hospital admissions and the inpatient database. However, despite these limitations, NIS is still an excellent inpatient population-based representation of disease associations and comorbidities. Despite the retrospective nature of the study, the chances of recall bias are probably minimal, given that it is an administrative database with primary and secondary diagnosis codes and other clinical information collected at the time of inpatient management, as well as on discharge. The main strength of this study is the nationally representative sample provided
