Statistical analysis

All statistical analyses were performed on Stata 14.2 (StataCorp, College Station, TX, USA) or StatsDirect Statistical Software (version 3.2.10, 03/05/2020). Descriptive statistics summarized demographic and clinical data. Frequencies and percentages were used to present categorical data. Mean ± SD values were presented for continuous variables that were normally distributed. Otherwise, continuous data were presented as median ± IQR.

Multivariate Cox regression analysis was used to evaluate the association between CKD and post-fracture mortality risk. Independent variables were selected a priori. Age (for each 1 year increase), sex, intracapsular hip fracture, CKD stage (with CKD G3b being the reference), CCI (for each unit of increase) and CFS score (with CFS 1–2 being the reference category). The assumption of proportional hazards was assessed by reviewing the significance of time-variable interactions.

An area under the curve (AUC) value was calculated through receiver operating characteristic (ROC) curve analyses to assess each assessment tool’s prognostic value for 30 day and 1 year mortality. A two-tailed p value < 0.05 was considered statistically significant. As a secondary analysis, there was no a priori sample size calculation.