Editorial Commentary: Big Databases Are Not All Created Equal: Interpret Their Studies With Caution.

"Big Data" studies are extremely important in orthopedic research, as randomized controlled trials with extremely large sample sizes are nearly impossible to perform. But, as always, the devil is in the details. Databases are only as good as the information that is put into them. And when evaluating these studies, let's not forget how to interpret them. Many factors-patient insurance status, age, socioeconomic status, ethnicity, and comorbidities; regional variations; hospital status (inpatient/outpatient); clerical errors, recording biases, and omission of relevant orthopedic outcome measures; temporal changes in billing codes; payer mix; population demographics and catchment area; and data collection methodology-mean that the results of a specific big data study may or may not be generalizable to other populations.