Source: NLM
We used the 2000-2014 US National Inpatient Sample (NIS) combined with Census Bureau data to develop projections for primary THA and TKA from 2020 to 2040 using polynomial regression to account for the nonlinearity and interactions between the variables, assuming the underlying distribution of the number of THA/TKA to be Poisson distributed. We performed sensitivity analyses using a negative binomial regression to account for overdispersion.