PROSPECTIVE RANDOMISED STUDY OF DIRECT ANTERIOR APPROACH VERSUS POSTERIOR APPROACH FOR BIPOLAR HEMIARTHROPLASTY OF THE HIP

Purpose

According to Bipolar Hemiarthroplasty of the Hip (BHA), several reports indicated earlier acquisition of walking ability in direct anterior approach (DAA) than posterior approach (PA), but there is still few randomized studies comparing accuracy of stem insertion in DAA and PA. Therefore, we performed a prospective study, focusing positioning of implant.

Patient and Methods

We chose 29 patients for this study, diagnosed as femoral neck fracture (Garden grade III or IV) from April 2012 to April 2013, excluding obesity patients (body mass index upper than 30kg/m²), muscular patients, osteoporotic patients (Sigh grade I or II), and tumor bearing patients. Patients were enrolled in a prospective, non-blinded study and were randomly assigned by envelope method to receive either DAA or PA. Taperloc BHA system (Zimmer) was adapted for all cases. Patient condition, operation condition and radiographic findings were investigated. On CT imaging, Canal fill was calculated Stem fill was studied at lower point of lesser trochanter that was considered to be load center. Static analysis was carried out using SPSS ver21 (IBM^c). Continuous variables between the two groups were compared using a no matched bilateral Student's t-test. For nominal variables we used χ²test. P value <0.05 was considered to be significant. (Table1)

Result

There was no significant difference in the head size (DAA43.7/PA45.1mm, p=0.628), stem size (DAA10.6/PA10.8, p=0.739) and total blood loss (DAA198.3/PA146.7ml, p=0.41). Duration of operation was significantly longer in the DAA group at 85.61 min in comparison with 61.75 min in the PA group (p<0.001). One fracture of the greater trochanter during surgery was occurred in DAA group. This case needed cable fixation around greater trochanter. No symptomatic pulmonary embolism occurred, but deep vein thrombosis was detected in three patients in DAA group and in two patients in PA group. Canal fill calculated by CT imaging was no significant difference between both groups (DAA67%/PA71%, p=0.12). All patients were discharged to a rehabilitation facility. Modified Harris Hip score at the discharge assessed early clinical result. There was no significant difference between both groups (DAA45.7/PA49.1, p=0.713).

To evaluate the learning effect of the operative outcome, we divide both groups into halves. We defined the first half of subgroups as early phase and the last half of subgroups as late phase. In the DAA group, significant improvement was observed in canal fill and there was a tendency of total blood loss to be lesser (Table 2). In the PA group, on the other hand, no improvement was observed for any values (Table 3).

Discussion and conclusion

We performed a prospective randomized study. The strengths of this study include the standardized protocol about stem component and surgical team. There were significant difference with duration operation in this study. This facilitates accurate radiographic assessment so that we could analyze more reliable data. This study is a preliminary study and probably sample size may be inadequate. We can also ignore the effect of learning curve especially in the DAA group.