The investigators will use health administrative data to examine the variation of anesthesia type for hip fracture surgery, as well as sources of variation and predictors of variation.

To evaluate the sources of variance in utilization of neuraxial anesthesia the investigators will use multilevel multivariable modeling strategies. This will allow identification of sources of variation (hospital, vs physician, vs patient characteristics), as well as identification of independent predictors of neuraxial use at each of these three levels.

The investigators will also evaluate the independent association of hospital level variation in neuraxial use on our outcomes of interest using multilevel multivariable regression.

Inclusion Criteria:

• Had hip fracture surgery
• Admitted to hospital on a non-elective basis

Exclusion Criteria:

• Do not hold a valid provincial health insurance account

All Ontario residents aged 66 years or older on the day of their emergency hip fracture surgery.

30 Days

All-Cause Mortality: The occurrence of death due to any cause was reported. No other AEs were collected or monitored.

Health administrative data are not initially collected for research purposes. There are patient-level predictors, hospital-level variables, and anesthesiologist variables that we could not measure directly. Surgeon-specific variables such as preference for neuraxial versus general anesthesia could influence anesthesia decision making and should be considered. Patient-preference should contribute to warranted variation. The generalizability of our findings to other jurisdictions is uncertain.