We examined data from a large prospectively collected dataset which followed up patients after Exeter total hip replacement (THR) - the Exeter Primary Outcome Study. We studied 78 patients who had total hip replacement for osteoarthritis on a morphologically normal hip, and in whom the other hip was also morphologically normal for comparison. All selected patients had complete patient outcome data at 1 and 5 years, and x-rays at 1 year were available.

We measured accuracy of reconstruction on AP Pelvis radiographs. The parameters measured were difference in height of lesser trochanter from horizontal pelvic line (LLD); length from ASIS to greater trochanter (indicating abductor length, AL); length from symphysis to centre of femoral head (SFH); length from centre of head to axis of femur (offset). We examined clinical outcomes including gain of Oxford Hip Score (DOHS) at 1 and 5 years, absolute OHS at 1 and 5 years, and absolute SF-36 Physical Functioning and Role Physical scores at 1 and 5 years.

Examining the ratios of AL, SFH, Offset, Offset to SFH ratio and LLD between side of THR and unaffected side all showed marked variation. Thus the ratio of SFH varied from 0.85 to 1.11. For further analysis we divided the patients into 3 groups on the basis of variability of the ratio from 1; we examined whether the clinical scores above varied between these 3 groups. 1 way ANOVA demonstrated no significant difference for any of the clinical outcomes for any of the 5 grouped x-ray variables.

Accuracy of reconstruction has been shown to affect the risk of dislocation after THR. Our results indicate that clinical outcomes of total hip replacement appear resilient to some degree of surgical inaccuracy. Future trials of navigation should be designed to demonstrate not only improved accuracy but also improved clinical outcome.