During the study period 555 consecutive patients with potentially eligible hip fracture were registered at the emergency department. Of 333 patients assessed for eligibility by the orthopedic surgeon on duty at the Emergency Department, 288 patients were included in the original randomized trial (14 did not meet the inclusion criteria, 9 refused to participate, 4 had a refracture and 18 were excluded for other reasons). Of the 222 patients not assessed for eligibility (mean age 83 years), 70% were women, and 51% had a cervical and 49% a pertrochanteric or a subtrochanteric fracture.
The number of patients excluded from our present analysis was 33; patients receiving warfarin (n = 24) and patients receiving high-dose acetylsalicylic acid, dipyridamol or clopidogrel (n = 9). Of the 255 trial participants eligible for this analysis, 118 (46.3%) used LdAA (50 mg in 1 patient, 75 mg in 68 patients, 150 mg in 2 patients, 160 mg in 40 patients, 250 mg in 2 patients and 320 mg in 5 patients). Patients using LdAA preoperatively were older and more likely to have hypertension, cardiovascular and/or cerebrovascular disease, renal dysfunction and ASA score of III or IV than patients not using LdAA (Table 1).
Bleeding and Transfusion
Patients using LdAA at the time of fracture had significantly higher APTT and INR values and were significantly more likely to receive postoperative blood transfusion (odds ratio 1.8), but did not differ significantly from patients not using LdAA regarding the proportion transfused or the amount of blood transfusions given before and during surgery (Table 2).
At 30 days after surgery 17 patients (6.7%) had died, 14 were preoperative LdAA users and 3 were non-users (11.9% versus 2.2%). The 90-day mortality was 10.6% and first-year mortality was 19.2%, both higher among preoperative LdAA users (17.8% vs. 4.4% and 29.7% vs. 10.2% respectively) (Figure 1). Among the preoperative LdAA users the type of surgery was hemiarthroplasty in 14 of the 35 1-year non-survivors (40%) and in 37 of the 83 survivors (45%). Among the preoperative LdAA users, 1-year mortality was 31.7% (26 of 82) in women and 25% (9 of 36) in men (P = 0.52) and among nonusers it was 9.3% (10 of 108) in women and 13.8% (4 of 29) in men (P = 0.50). Among the LdAA users treated with ≤75 mg, 1-year mortality was 30.4% (21 of 69 patients) and among those treated with ≥150 mg it was 28.6% (14 of 49 patients) (P = 1.0).
Kaplan-Meier survival estimates during first year after surgery for hip fracture according to preoperative use of low-dose acetylsalicylic acid (LdAA), analysis time is number of days since surgery.
In the Cox regression analysis use of LdAA preoperatively was associated with significantly higher first-year mortality, with an adjusted HR of 2.35 (Table 3). In the Cox regression model that included a term for the interaction between sex and preoperative LdAA use, the sexspecific HR was estimated to be 2.88 for women and 1.25 for men. However, this difference in effect of LdAA was shown not to be statistically significant when tested using a likelihood-ratio test (P = 0.25).
Early Postoperative Complications
The total number of patients who suffered postoperative complications up to three months after the fracture did not differ between LdAA users and non-users (Table 4).
But patients using LdAA preoperatively were significantly more likely to be afflicted with thromboembolic events, such as deep vein thrombosis and pulmonary embolism (5.7% vs. 0.7%, P = 0.03). Among LdAA users those who did not survive one year the number of patients with complications were 22 (63%) compared to the survivors 32 (39%).
BMC Musculoskelet Disord. 2011;12(254) © 2011 BioMed Central, Ltd.
© 1999-2006 BioMed Central Ltd