We retrospectively reviewed the medical records of all 869 patients who underwent 1025 elective primary THAs from May 2003 to December 2009. Eight patients were treated with heparin and warfarin for thromboprophylaxis and were excluded from the study: two patients (two hips) with a history of a thromboembolic event and six patients (six hips) with cardiac disease before surgery. This left 861 patients (1017 hips) for the study. Simultaneous bilateral THAs were performed in 25 patients, who were considered to have had one procedure. Therefore, the total number of procedures was 992 in 861 patients. Of the 861 patients, 65 (74 hips) took aspirin-containing compounds or other antiplatelet agents before surgery. Because their medications were discontinued 5 to 7 days before surgery, these patients were not excluded [7
]. There were 414 women and 447 men with a mean age of 51.1 years (range, 18–83 years) at the time of the operation. Their mean BMI was 24.2 kg/m2
(range, 14.6–38.8 kg/m2
), and the most common diagnosis for THA was osteonecrosis of the femoral head (564 hips, 55.5 %) (Table ). In 857 patients who survived longer than 6 months postoperatively, the mean followup was 42.8 months (range, 6–91 months).
Cementless fixation was used for the acetabular cup and femoral stem in all patients. Regional anesthesia was used in 876 procedures, and general anesthesia in 116 procedures. All procedures were performed with the patients in a lateral position. The posterolateral approach was used in 723 procedures, and the anterolateral approach was used in 269 procedures. In 93 procedures, wider exposure was necessary and the approach was extended to a triradiate approach, combined anterior and posterior approach, or transtrochanteric approach during the operation. The median operating time was 125 minutes (range, 50–535 minutes).
No pharmacologic or mechanical prophylaxis was used postoperatively in any of the 861 patients. However, thigh-length antiembolic stockings were applied and the patients were encouraged to use an ankle pump while in bed during the hospitalization. On postoperative Days 1 to 3, closed suction drainage was removed and patients were mobilized to a wheelchair. On postoperative Days 3 to 10, patients walked with restricted weightbearing and use of assistive devices (wheelchair, walker, crutches, or cane). As the walking ability improved, the assistive devices were changed as determined appropriate by a physical therapist. The mean length of hospital stay was 15.0 days (range, 6–28 days).
After the operation, we routinely monitored patients for clinical signs of DVT including pain and tenderness in the calf or thigh, swelling or erythema of the surgically treated limb, and a positive Homans’ sign. We suspected DVT or PE in 32 patients and consulted the cardiovascular physicians. A diagnosis of DVT was confirmed by duplex ultrasonography or lower extremity CT angiography. A PE was confirmed by a ventilation/perfusion scan or pulmonary CT angiography. Patients who were diagnosed as having a DVT or PE were treated with warfarin. Patients were monitored for 1 to 3 weeks in the ward.
Most deaths attributable to PE related to surgery reportedly occur within 3 months and any death of unknown cause that occurs within 3 months of surgery is considered to be the result of PE [5
]. We confirmed the fatal PE, if present, from the death certificate.
After discharge, patients were followed routinely at 6 weeks, 3 months, and 6 months postoperatively with specific attention given to the development of DVT or PE, although no patients were recalled specifically for this study.
Four patients died of causes unrelated to the operation within 6 months after surgery. Eight hundred thirty-eight patients visited the outpatient clinic for followups once or more after 6 months postoperative. Nineteen patients, who were unable to return, were visited or contacted by telephone by two nurses and one private locator.
We determined the incidence of symptomatic DVT, PE, and fatal PE. To determine confounding factors, univariate comparisons between the VTE group and the non-VTE group were made based on the demographic data and operative parameters, including age, gender, BMI, administration of aspirin, type of anesthesia, operation time, approach, simultaneous bilateral THAs, and duration of postoperative immobilization in bed. We used Fisher’s exact test for categorized data and the Mann-Whitney U test for continuous data. For the variables with a p value less than 0.1 in the univariate analyses, multivariate logistic regression analyses using the enter method were performed. The independent variables tested for the multivariate logistic regression analyses included age, gender, anesthesia, and duration of immobilization, as confounding factors; the dependent variable was whether the DVT occurred postoperatively. From the multivariate regression analyses, it was assessed which variables were the risk factors for occurrence of DVT. Statistical analyses were conducted using SPSS for Windows statistical package (version 12.0; SPSS, Chicago, IL, USA).
The design and protocol of this study were approved by the institutional review board in our hospital, who waived informed consent.