Factors associated with implant survival following total hip replacement surgery: A registry study of data from the National Joint Registry of England, Wales, Northern Ireland and the Isle of Man

The choice of the RD&E as the exposure group (rather than any of the other units with “better than expected” survival results) is due to a lack of selection bias, in that every patient in this unit receives the same femoral stem, regardless of age, sex, or indication. The study population was all THRs implanted in the NJR; subgroup analysis was performed for THRs using any type of cemented stem (hybrid or all-cemented constructs) as well as THRs using the Exeter V40 femoral stem (hybrid or all-cemented constructs), the stem used by the RD&E. Unadjusted survival estimates were calculated using the Kaplan–Meier (KM) method for all included THRs, stratified by the exposure of interest [14]. FPSA modeling offers an advantage over more traditional semiparametric techniques, such as Cox regression, because in addition to allowing effects to vary with time via cubic splines, they allow us to estimate a baseline hazard function. Socioeconomic status was assessed using deciles of the Index of Multiple Deprivation (IMD) organized by Lower Layer Super Output Areas (LSOAs), and BMI was treated as a categorical variable using World Health Organization categories (40 kg/m2).