Abstract
Background
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.
Materials & methods
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.
Results
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.
Conclusion
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.
Keywords: Hip fracture, Trauma, Hemiarthroplasty, Templating
1. Background
A crucial part of elective total hip arthroplasty (THA) is meticulous pre-operative planning to ensure that length, offset and centre of rotation are adequately restored. This was historically achieved by use of acetate templates superimposed over traditional radiographic films, but with the advent of Picture Archiving and Communication Systems (PACS), digital templating has increasingly come into vogue and has been shown to be at least as accurate, if not more so, than traditional methods.1, 2, 3 Obtaining pre-operative radiographs for templating often involves the placement of an external calibration marker (ECM) of a known diameter at a specific point within the field; the digital films can then be merged with a templating software package and calibrated in order to accurately predict implant sizes.
Despite its widespread use in elective orthopaedic surgery, digital templating has yet to be commonly applied in the trauma setting. This may be due to compromised radiograph quality, as patients are often agitated and in pain4,5 and therefore unable to hold a required position long enough to obtain ideal radiographs. In addition, the urgent nature of obtaining radiographs in the trauma setting often means that an ECM is rarely used. This, therefore, creates an element of magnification that must be accounted for if attempting to template. King et al. concluded that magnification factor in trauma extremity radiographs was greatest proximally4 (i.e.- radiographs of the femoral head were associated with a greater magnification factor than the distal tibia, for example).
The use of an ECM does not always guarantee accuracy however; incorrect positioning of the marker within the X-ray field leads to errors that can significantly affect predicted implant size.6,7 Numerous groups have reported that accounting for a set magnification factor, rather than using an ECM, produced more reliable results when templating for THA.6,8, 9, 10, 11 There is, however, great variability in the amount of magnification that must be accounted for, with estimates reaching as high as 27%,9 and consensus often being that magnification must be calculated for each individual institution to ensure accuracy.8,11
Neck of femur fracture remains one of the commonest orthopaedic injuries around the world, with the National Hip Fracture Database recording over 63,000 occurring in England, Wales & Northern Ireland in 2021 alone, with hemiarthroplasty the most commonly performed operation for intracapsular fractures.12
The aim of our study was to establish our institute-specific magnification factor when measuring femoral head size from standard radiographs without an ECM, and to compare this with other set magnifications described in the literature. We hypothesised that using an institute-specific magnification should provide more accurate results in predicting femoral head size.
2. Materials & Methods
We retrospectively identified 50 consecutive patients from December 2018 to June 2019 who underwent hip hemiarthroplasty for traumatic neck of femur fracture, from our institution's in-house neck of femur fracture database.
Pre-operative pelvic radiographs were loaded into Merge OrthoCase™ (IBM Watson Health, Cambridge, MA), a templating software package used in elective arthroplasty planning. One of the authors (DT) measured the widest point of the contralateral (intact) femoral head, from which the templating software was able to generate a diameter for the femoral head (see Fig. 1). Another author (KT) independently performed the same measurements for a subset of 16 radiographs, and both clinicians performed repeat measurements at separate times to generate data to assess inter- and intra-observer reliability. These data were calculated using intraclass correlation coefficients (ICC).
Fig. 1.
Measuring femoral head size using templating software.
Actual implanted femoral head size was obtained from operation notes. The magnification factor was calculated using the following formula:
| Templated head size / Actual head size x 100 |
Based on the mean magnification factor generated from this, we calculated a predicted head size for each patient using the following formula:
| Templated head size / Magnification ×100 |
The same was done using set magnifications of 15% and 21% which have previously been reported in the literature.
These predicted head sizes were then compared with the actual head sizes implanted and paired t-tests were used to compare the accuracy of templating at the different magnifications.
3. Results
Nine patients were excluded; one had no pre-operative radiographs (diagnosis was made on trauma CT scan), one had no documentation of the head size implanted in the operation note, and seven patients had a contralateral total hip replacement or hemiarthroplasty, thus precluding templating.
Average patient age was 85, male:female ratio was 1:2.2. The majority of patients (51%) received a cemented Exeter Trauma Stem. Other implants used included cemented Exeter stem with Unitrax head (27%), cemented Thompson (7%) and uncemented Thompson (15%). The range of head sizes used was 40–57 mm (see Table 1).
Table 1.
Demographic & Implant information.
| n (%) | |
|---|---|
| Demographics | |
| Female | 28 (68%) |
| Male | 13 (32%) |
| Average age (Range) | 85 (60–102) |
| Prostheses | |
| Exeter Trauma Stem (Cemented) | 21 (51%) |
| Exeter Stem with Unitrax head (Cemented) | 11 (27%) |
| Thompson (Cemented) | 3 (7%) |
| Thompson (Uncemented) | 6 (15%) |
| Average head size (Range) | 47 mm (40–57) |
The average magnification factor was 17% (range = 107.5–126.8). The intraclass correlation coefficient (ICC) for intra-observer reliability was 0.99 (confidence interval = 0.98–1). The ICC for inter-observer reliability was 0.99 (CI = 0.96–0.99).
We then calculated the predicted head size based on initial measurements, accounting for a magnification of 17%. Head size was correctly predicted in 22% of cases, to within ± 1 mm in 49% and within ± 2 mm in 68%. The mean variability was 0 (range = −5 – +4).
Using a set magnification of 15%, head size was correctly predicted in 17% of radiographs, to within ± 1 mm in 46% and within ± 2 mm in 71%. Mean variability was +1 (range = −4 – +5).
With a set magnification of 21%, head size was correctly predicted in only 10% of cases, to within ± 1 mm in 39% and within ± 2 mm in 63%. Mean variability was −2 (range = −6 – +2) (see Table 2).
Table 2.
Prediction of femoral head size by magnification factor.
| Femoral head size correctly predicted |
|||
|---|---|---|---|
| Tolerance (mm) | 17% magnification | 15% magnification | 21% magnification |
| 0 | 22% | 17% | 10% |
| ±1 | 49% | 46% | 39% |
| ±2 | 68% | 71% | 63% |
Overall, we found no statistically significant difference between the actual implanted femoral head size and predicted head size when accounting for 17% magnification (p = 0.82). There was also no significant difference in accuracy at predicting femoral head size when comparing our institute-specific magnification factor with other magnifications of 15% (p = 0.35) or 21% (p = 0.12).
4. Discussion
Templating in elective arthroplasty has been proven to reliably provide favourable outcomes.13,14 There is currently a paucity of literature, however, reviewing whether templating can confer similar benefits in the trauma setting.
Currently, standard practice in hip hemiarthroplasty is intra-operative measurement of the femoral head, using a calliper or template. Whilst this remains the most accurate way of assessing head size, if implant stocks are low or the required prosthesis is significantly smaller or larger than usual, this can generate lengthy intra-operative delays whilst the correct implant is located. The ability, therefore, to accurately template pre-operatively could be of great benefit.
Our results confirm that magnification factor is site-specific, due to differences between institutions in radiographic equipment used and technique employed. But, whilst other groups have previously suggested that the magnification factor must be calculated for each individual institution to ensure accuracy,8,11,15 our findings suggest that there is no significant difference in accuracy at predicting head size when using our own institute-specific magnification factor or an arbitrary magnification factor of greater or lesser value.
Derias et al. used a set magnification of 20% to template 50 patients for hip hemiarthroplasty and reported 90% accuracy at predicting femoral head to with ± two sizes.11 Another group have reported as high as 90% accuracy to within ± one head size using a set magnification of 21%.16 Even using our institute-specific magnification of 17%, our results do not support these findings, as we were not able to achieve anything close to 90% accuracy at predicting femoral head size, even within a tolerance of ± 2 mm, and it is widely acknowledged that there is high variability in magnification between institutions, making accurate predictions difficult.8,17,18
Crosswell et al. recently trialled several different models for assessing accuracy at predicting femoral head size, including two assumed values of 10% and 15% magnification, a regression-through-the-origin model and a model including an intercept parameter.19 They concluded that, whilst the latter was significantly more accurate at predicting head size, due to the necessary formula required to calculate head size in this model, they recommend the use of the regression-through-the-origin model, the magnification value of which was 17.8%.
Whilst we did not demonstrate any statistically significant difference between predicted head size assuming a magnification of 17% and the actual head size implanted, we were also unable to demonstrate that calculating our own institute-specific magnification was any more accurate that using other arbitrary values.
In practice we believe there is little role for templating from magnification in trauma; with the calculated range of −5 to +4 mm, it would be necessary to have 10 continuous sizes of implant available for every templated patient, except for in approximately seven out of every ten cases, where a range of five sizes would be adequate. Clearly this is not practical.
Many commercial digital templating software vendors allow the user to set a magnification factor if an ECM is not used prior to templating. Our study suggests that this should not be recommended. Furthermore, due to the potential large discrepancies, we would suggest that this function should be removed from any of these software or at least highlight that the final template should not be used for surgery planning.
We appreciate that the key limitation to our study is our small cohort of patients, which could potentially bias results. Using a larger sample size after calculating our institution's unique magnification factor may have yielded more accurate results. Although the OrthoCase™ software is CE marked for medical use there may be digital inaccuracies in the coding and software itself leading to discrepancies in the magnification and therefore its templating function within the platform. This would apply to any templating software vendor where different tolerances and inaccuracies may exist. Despite the described limitations, ours is one of few studies to look at templating in hip hemiarthroplasty from magnification factor. We also performed inter- and intra-observer reliability using the templating software which shows observer validity in our measurements, and this correlates with other groups who have similarly investigated inter- and intra-observer reliability using templating software to determine magnification.11
5. Conclusion
Whilst templating with an ECM has been demonstrated to aid in achieving favourable long-term outcomes in THA, attempting to template from magnification does not seem to reproduce these positive results. Utilising an ECM when obtaining radiographs for patients with suspected neck of femur fracture may prove to be more reliable at predicting femoral head size and we would suggest that future work in this area should focus on prospective cohort studies comparing ECM with magnification for templating.
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.
Author contributions
D Thurston: Collected and analysed data, wrote manuscript, S El-ashry: Designed study, wrote manuscript, S Gella: Wrote manuscript, K Theivendran: Designed study, collected data, wrote manuscript.
Declaration of competing interest
None.
Footnotes
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Contributor Information
Daniel Thurston, Email: d.thurston@doctors.org.uk.
Saad El-ashry, Email: saad.elashry@nhs.net.
Sreenadh Gella, Email: sreenadh.gella@nhs.net.
Kanthan Theivendran, Email: k.theivendran@nhs.net.
References
- 1.Iorio R., Siegel J., Specht L.M., Tilzey J.F., Hartman A., Healy W.L. A comparison of acetate vs digital templating for preoperative planning of total hip arthroplasty: is digital templating accurate and safe? J Arthroplasty. 2009;24(2):175–179. doi: 10.1016/j.arth.2007.11.019. Feb. [DOI] [PubMed] [Google Scholar]
- 2.Gamble P., de Beer J., Petruccelli D., Winemaker M. The accuracy of digital templating in uncemented total hip arthroplasty. J Arthroplasty. 2010;25(4):529–532. doi: 10.1016/j.arth.2009.04.011. Jun. [DOI] [PubMed] [Google Scholar]
- 3.Whiddon D.R., Bono J.V., Lang J.E., Smith E.L., Salyapongse A.K. Accuracy of digital templating in total hip arthroplasty. Am J Orthop (Belle Mead NJ) 2011;40(8):395–398. Aug. [PubMed] [Google Scholar]
- 4.King R.J., Craig P.R., Boreham B.G., Majeed M.A., Moran C.G. The magnification of digital radiographs in the trauma patient: implications for templating. Injury. 2009;40(2):173–176. doi: 10.1016/j.injury.2008.06.027. Feb. [DOI] [PubMed] [Google Scholar]
- 5.Kwok I.H., Pallett S.J., Massa E., Cundall-Curry D., Loeffler M.D. Pre-operative digital templating in cemented hip hemiarthroplasty for neck of femur fractures. Injury. 2016;47(3):733–736. doi: 10.1016/j.injury.2015.12.027. Mar. [DOI] [PubMed] [Google Scholar]
- 6.Heinert G., Hendricks J., Loeffler M.D. Digital templating in hip replacement with and without radiological markers. J Bone Joint Surg Br. 2009;91(4):459–462. doi: 10.1302/0301-620X.91B4.21730. Apr. [DOI] [PubMed] [Google Scholar]
- 7.Boese C.K., Wilhelm S., Haneder S., Lechler P., Eysel P., Bredow J. Influence of calibration on digital templating of hip arthroplasty. Int Orthop. 2019;43(8):1799–1805. doi: 10.1007/s00264-018-4120-7. Aug. [DOI] [PubMed] [Google Scholar]
- 8.Franken M., Grimm B., Heyligers I. A comparison of four systems for calibration when templating for total hip replacement with digital radiography. J Bone Joint Surg Br. 2010;92(1):136–141. doi: 10.1302/0301-620X.92B1.22257. Jan. [DOI] [PubMed] [Google Scholar]
- 9.Riddick A., Smith A., Thomas D.P. Accuracy of preoperative templating in total hip arthroplasty. J Orthop Surg. 2014;22(2):173–176. doi: 10.1177/230949901402200211. Aug. [DOI] [PubMed] [Google Scholar]
- 10.Archibeck M.J., Cummins T., Tripuraneni K.R., et al. Inaccuracies in the use of magnification markers in digital hip radiographs. Clin Orthop Relat Res. 2016;474(8):1812–1817. doi: 10.1007/s11999-016-4704-8. Aug. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Derias M., Khan M.A., Buchanan J. Digital templating without a calibration marker is accurate at predicting implant size for hip hemiarthroplasty. Eur J Trauma Emerg Surg. 2018 doi: 10.1007/s00068-018-0994-8. Sep. 12. [DOI] [PubMed] [Google Scholar]
- 12.National Hip Fracture Database Annual Report 2021. Available at: https://www.nhfd.co.uk/20/hipfractureR.nsf/report?open&NHFD+2021+Report. Accessed January 8th, 2022.
- 13.Della Valle A.G., Padgett D.E., Salvati E.A. Preoperative planning for primary total hip arthroplasty. J Am Acad Orthop Surg. 2005;13(7):455–462. doi: 10.5435/00124635-200511000-00005. Nov. [DOI] [PubMed] [Google Scholar]
- 14.González Della Valle A., Slullitel G., Piccaluga F., Salvati E.A. The precision and usefulness of preoperative planning for cemented and hybrid primary total hip arthroplasty. J Arthroplasty. 2005;20(1):51–58. doi: 10.1016/j.arth.2004.04.016. Jan. [DOI] [PubMed] [Google Scholar]
- 15.Hornová J., Růžička P., Hrubina M., et al. Magnification of digital hip radiographs differs between clinical workplaces. PLoS One. 2017;1 2(11) doi: 10.1371/journal.pone.0188743. Nov 30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Archibeck M.J., Tripuraneni K.R., Carothers J.T., Junick D.W., Munson N.R., Murray-Krezan C.M. Prospective, randomized, surgeon-blinded comparison of standard magnification assumption vs magnification marker usage for preoperative templating in total hip arthroplasty. J Arthroplasty. 2017;32(10):3061–3064. doi: 10.1016/j.arth.2017.05.016. Oct. [DOI] [PubMed] [Google Scholar]
- 17.Pickard R.J., Higgs D., Ward N. The accuracy of the PACS for pre-operative templating. J Bone Joint Surg Br. 2006;88-B(Suppl 2):264. [Google Scholar]
- 18.Varghese B., Muthukumar N., Balasubramaniam M., Scally A. Reliability of measurements with digital radiographs--a myth. Acta Orthop Belg. 2011;77(5):622–625. Oct. [PubMed] [Google Scholar]
- 19.Crosswell S., Akehurst H., Ramiah R., Navadgi B. Preoperative sizing of hip hemiarthroplasties to accurately estimate head size from non- standardised pelvic radiographs: can it be done? Injury. 2019;50(11):2030–2033. doi: 10.1016/j.injury.2019.08.025. Nov. [DOI] [PubMed] [Google Scholar]