Delayed Primary Hip Arthroplasty for Geriatric Low-Energy Femoral Neck Fracture Is Not Associated With Worse Outcomes: A Study of the Arthroplasty for Hip Fracture Consortium.

BACKGROUND

There is no consensus on the association between mortality and surgical delay in hemiarthroplasty (HA) or total hip arthroplasty (THA) for elderly, low-energy, displaced femoral neck fractures (FNFs). This study assessed whether delayed surgery is associated with worse outcomes.

METHODS

This retrospective study reviewed FNF in patients > 60 years among nine academic tertiary-care institutions from January 2010 through December 2019. A total of 1,420 FNF patients underwent HA (n = 939) or THA (n = 481) < 48 hours (86.5%; n = 1,229) or ≥ 48 hours (13.5%; n = 191) from admission. Complications and mortalities at 90 days and 1 year were compared between groups. Multivariable regression and receiver operator characteristic curves assessed outcomes.

RESULTS

The delayed group had higher anesthesia society 3 and 4 scores (88.5 versus 81.4; P = 0.017), intensive care unit stays (22.5 versus 10.7%; P < 0.001) and achieved shorter ambulation distance at discharge (39.5 versus 51.6 feet; P < 0.001). There were no differences in 90-day or 1-year periprosthetic infections, revisions, or dislocations (P < 0.05). Delay ≥ 48 hours was associated with 90-day mortalities when controlling for significant covariates between delayed/accelerated groups (odds ratio, 1.00; 95% confidence interval, 1.00 to 1.01; P = 0.034), but with an area under the curve of 0.58 and Youden's Index of 25.5 hours. Time to surgery ≥ 48 hours was not associated with 1-year mortalities (odds ratio 1.00; 95% confidence interval 1.00 to 1.01; P = 0.45). Most delays were for medical (87.0%) reasons, and among logistical (13.0%) reasons, transfer from outside facilities or weekend admissions were frequently reported.

CONCLUSIONS

Our study does not support that delaying THA or HA for FNF has negative clinical implications. The association, although statistically significant, has poor clinical predictive value, and delays were necessary for medical optimization and were nonmodifiable in nature.

LEVEL OF EVIDENCE: Level III, Retrospective Cohort.