Creación de un instrumento de decisión objetivo para la indicación de tratamiento quirúrgico vs conservador en la fractura de cadera

Mortality after hip fracture is high and knowing the risk of mortality on admission could be useful to adjust the need for pre- and perioperative care, help decision-making and allow risk adjustment in audits. At Hospital La Ribera (Alzira) we have created a model that, after admission for hip fracture, measures the risk of mortality at 30 (Alzira Hip Fracture Score: AHFS-30) and 180 days (AHFS-180). Methods: 701 patients over 69 years of age, admitted consecutively for any type of hip fracture were included and variables associated with mortality in previous studies were collected. A multivariate logistic regression was performed. Discrimination and calibration were evaluated at both the 30- and 180-day periods using the area under curve ROC and the Hosmer-Lemeshow goodness-of-fit test. The variables that were significantly related to mortality at 30 days were: male sex (2.09 CI95% 1.16-3.76; p=0.014), previous or current respiratory disease at admission (3.05 CI95% % 1.69-5.48; p<0.001), alterations in the electrocardiogram (2.84 CI95% 1.56-5.16; p=0.001) and social problem at admission (2.30 CI95% 1 .24-4.28, p=0.009). The variables that showed significance at 180 days were: respiratory disease (2.56 CI95% 1.56-4.18; p<0.001), alteration in the ECG (2.95 CI95% 1.83-4.74 ; p<0.001), terminal illness (7.76 CI95% 3.07-19.65; p<0.001), social problem at admission (2.107 CI95% 1.29-3.44; p=0.003), Barthel< 30 points (2.96 CI95% 1.25-7.01; p=0.014), dementia with GDS>5 points (2.24 CI95% 1.09-4.61; p=0.029) and delirium at admission (2.09 CI95% 1.25-3.51; p=0.005). Both the AHFS-30 and the AHFS-180 showed good discrimination with an area under the curve ROC of 0.843 (CI: 0.794-0.891; p < 0.001) and 0.829 (CI: 0.787-0.872; p < 0.001), respectively. The Hosmer-Lemeshow test did not indicate a lack of fit in either of the two models, AHFS-30: χ2(8) = 9.77; p = 0.281; AHFS-180: χ2(7) = 8.02; p = 0.331. Conclusions: The AHFS-30 and AHFS-180 are based on variables at admission for hip fracture and can be used to predict the risk of mortality at 30 and 180 days individually for each patient.