There are a number of limitations of this study.
This is a retrospective analysis with lack of randomization, potential
severity bias, and likely population mismatch. Analysis of administrative data depends on identification of relevant
records using ICD-9 codes rather than more accurate clinical data
Our study findings may be limited by an unknown degree
of completeness of coding. Information bias may also
exist because of over- or underdiagnosis of the underlying
conditions, lack of clinical details, and inability to control
for severity of illnesses