Abstract
Objective: As the global population ages, hip fracture importance will increase. The high postoperative mortality and morbidity necessitate tools for accurate risk assessment to aid surgical decisions and inform patients and families. This study aimed to compare and validate ACS NSQIP, Nottingham hip fracture risk calculators, and C reactive protein/albumin ratio in predicting complications and 30 day mortality. Methods: 583 patients over the 65 years old who sustained hip fracture from simple fall and underwent surgical treatment for hip fracture were included in prospective study. Each patient was evaluated by ACS NSQIP and Nottingham hip fracture risk calculators and C reactive protein/albumin ratio was calculated from preoperative values. Patients were followed up for 30 days and all the complications were recorded. Results: ACS NSQIP and Nottingham hip fracture score showed AUC of .724 and .731 respectively. C reactive protein/albumin ratio performed less and showed AUC of .623 that is defined as poor predictor for 30 day mortality. Furthermore, in terms of predicting any complication, ACS NSQIP showed AUC of .645, Nottingham hip fracture score of .611 and C reactive protein/albumin ratio of .594. Nottingham hip fracture score than compared to ACS NSQIP yielded lower average of mortality rates (5.4% compared to 7.9%) in contrast of study findings of 8.1%. None of the ACS NSQIP scale predictive complication showed acceptable performance. When adjusted for fracture type, Nottingham hip fracture score showed .858 AUC in predicting 30 day mortality in femoral neck fractures. Conclusions: We recommend Nottingham fracture risk calculator use for 30 day mortality risk assessment in femoral neck fractures. In hip fractures combined-none of models showed strong discrimination. In our cohort C reactive protein/albumin ratio showed poor prognostic values in terms of mortality and complications.