Outcome in hip fracture patients related to anemia at admission and allogeneic blood transfusion: an analysis of 1262 surgically treated patients

Anemia is more often seen in older patients. As the mean age of hip fracture patients is rising, anemia is common in this population. Allogeneic blood transfusion (ABT) and anemia have been pointed out as possible risk factors for poorer outcome in hip fracture patients. In the timeframe 2005-2010, 1262 admissions for surgical treatment of a hip fracture in patients aged 65 years and older were recorded. Registration was prospective from 2008 on. Anemic and non-anemic patients (based on hemoglobin level at admission) were compared regarding clinical characteristics, mortality, delirium incidence, LOS, discharge to a nursing home and the 90-day readmission rate. Receiving an ABT, age, gender, ASA classification, type of fracture and anesthesia were used as possible confounders in multivariable regression analysis. The prevalence of anemia and the rate of ABT both were 42.5%. Anemic patients were more likely to be older and men and had more often a trochanteric fracture, a higher ASA score and received more often an ABT. In univariate analysis, the 3- and 12-month mortality rate, delirium incidence and discharge to a nursing home rate were significantly worse in preoperatively anemic patients. In multivariable regression analysis, anemia at admission was a significant risk factor for discharge to a nursing home and readmission < 90 days, but not for mortality. Indication for ABT, age and ASA classification were independent risk factors for mortality at all moments, only the mortality rate for the 3-12 month interval was not influenced by ABT. An indication for an ABT was the largest negative contributor to a longer LOS (OR 2.26, 95% CI 1.73-2.94) and the second largest for delirium (OR 1.67, 95% CI 1.28-2.20). This study has demonstrated that anemia at admission and postoperative anemia needing an ABT (PANT) were independent risk factors for worse outcome in hip fracture patients. In multivariable regression analysis, anemia as such had no effect on mortality, due to a rescue effect of PANT. In-hospital, 3- and 12-month mortality was negatively affected by PANT, with the main effect in the first 3 months postoperatively.