Digital templating in hip hemiarthroplasty: Is it possible to accurately predict femoral head size from magnification alone?

Templating is an integral part of pre-operative planning in elective hip arthroplasty to achieve favourable long-term outcomes, but its applications in trauma surgery remain limited. When templating from radiographs without a calibration marker, there is always an element of magnification which must be accounted for. Our aim was to establish our institute-specific magnification and to determine whether using this to predict femoral head size in hemiarthroplasty was more accurate than using set magnifications previously reported in the literature. Fifty consecutive patients who underwent hip hemiarthroplasty were retrospectively identified, their pre-operative radiographs reviewed and femoral head measured with templating software. Intra- and inter-observer reliability analyses were performed. Using this value, and two set values of 15% and 21% magnification, we attempted to predict femoral head sizes of our original cohort. The results were compared using paired t-test to ascertain if there was any significant difference in accuracy. We established our institute-specific magnification as 17%. Inter- and intra-observer reliability were excellent. However, using this magnification we were only able to correctly predict to within ± one femoral head size in 49% of patients. There was no significant difference in accuracy comparing our institute-specific magnification with other magnifications from the literature. We would not recommend using magnification factor in digital templating software as this could potentially lead to errors in predicting final femoral head size in hip hemiarthroplasty and adversely affect patient outcomes. © 2022.

Citation

Daniel Thurston, Saad El-Ashry, Sreenadh Gella, Kanthan Theivendran. Digital templating in hip hemiarthroplasty: Is it possible to accurately predict femoral head size from magnification alone? Journal of clinical orthopaedics and trauma. 2022 Sep;32:101952

PMID: 35942324

search → result