Archives of Orthopaedic and Trauma Surgery

https://doi.org/10.1007/s00402-018-3080-0

HIP ARTHROPLASTY

The accuracy of digital templating in uncemented total hip

arthroplasty
Lukas A. Holzer1,2 · Georg Scholler1,3 · Stefan Wagner4 · Jörg Friesenbichler1 · Werner Maurer‑Ertl1 ·
Andreas Leithner1

Received: 30 June 2018

© The Author(s) 2018

Abstract
Introduction  Preoperative planning is an essential part of total hip arthroplasty (THA). It facilitates the surgical procedure,
helps to provide the correct implant size and aims at restoring biomechanical conditions. In recent times, surgeons rely more
and more on digital templating techniques. Although the conversion to picture archiving and communication system had
many positive effects, there are still problems that have to be taken into consideration.
Objectives  The core objective was to evaluate the impact of the planners’ experience on the accuracy of predicting com-
ponent size in digital preoperative templating of THA. In addition, the influence of overweight and obesity (according to
WHO-criteria), patient’s sex and component design on the accuracy of preoperative planning have been analysed.
Materials and methods  The retrospective study included 632 consecutive patients who had primary uncemented THA.
Digital templating was done using “syngo—EndoMap” software by Siemens Medical Solutions AG. Mann–Whitney U test
and Kruskal–Wallis test have been used for statistical analysis.
The accuracy of predicting component size has been evaluated by comparing preoperative planned sizes with implanted sizes
as documented by the surgeons. The planner’s experience was tested by comparing the reliability of preoperative planning
done by senior surgeons or residents. The influence of BMI on predicting component size has been tested by comparing the
accuracy of digital templating between different groups of BMI according to WHO-criteria. The same procedure has been
done for evaluating the impact of patient´s sex and component design.
Results  The implant size was predicted exactly in 42% for the femoral and in 37% for the acetabular component. 87% of the
femoral components and 78% of the acetabular cups were accurate within one size. Digital templating of femoral implant
size was significantly more reliable when done by a senior surgeon. No difference was found for the acetabular component
sizes. The BMI also had an impact on estimating the correct femoral implant size. In overweight patients, planning was
significantly more inaccurate than normal weight people. Differences were seen in obese patients. However, these were not
significant. Accuracy of acetabular components was not affected. The design of the prostheses and the patient’s sex had no
influence on predicting component size.
Conclusions  Inexperience and overweight are factors that correlate with inaccuracy of preoperative digital templating in
femoral components, whereas acetabular components seem to be independent of these factors.

Keywords  Total hip arthroplasty · Digital templating · X-rays · Experience · Obesity

* Lukas A. Holzer
lukas.holzer@medunigraz.at Introduction
1
Department of Orthopaedics and Traumatology, Medical Preoperative planning is an essential and integral part of
University of Graz, Auenbruggerplatz 5, 8036 Graz, Austria
total hip arthroplasty (THA). It facilitates determining the
2
AUVA Trauma Center Klagenfurt, correct implant size and helps restoring physiological bio-
Klagenfurt am Wörthersee, Austria
mechanical conditions such as leg length, centre of rotation
3
Department of Traumatology, Karl Landsteiner University and lateralization [1, 2]. It has been shown that component
of Health Sciences, Krems, Austria
loosening due to underestimation or periprosthetic frac-
4
Department of Sociology, Johannes Keppler University Linz, tures due to overestimation of implant size can be avoided
Linz, Austria

13
Vol.:(0123456789)
 Archives of Orthopaedic and Trauma Surgery

[3–5]. Moreover, preoperative surgical planning improves a tapered stem geometry and a grit-blasted surface which
postoperative range of motion and stability, shortens the provide proven initial and secondary fixation. It has a rec-
operative time and reduces wear caused by mal-positioning tangular cross-section that enables rotational stability. The
of the implant components [6]. Furthermore, it allows to Pinnacle cup is a spherical cup with a single radius. The
reduce costs by decreasing the necessity of large invento- Corail stem is designed to sit in the cancellous bone. It is
ries of implants [7]. In last decade, preoperative planning hydroxyapatite coated and has trapezoidal-like proximal
of THA changed as X-rays became digitalized. Accuracy cross-section to provide rotational stability. Due to its design
and reliability have been studied in both analogue as well the choice of size might get influenced by the quality of sur-
as digital techniques. Both techniques show similar results rounding bone stock.
of accuracy [7, 8].
As indicated above digital templating poses an essential Preoperative X‑ray technique
part of THA. Still, there is a variety of factors that might
negatively influence its accuracy. With the introduction of Standard preoperative digital radiographs of the hip were
picture archiving and communication system (PACS), tem- obtained in anterior–posterior view. The tube to film dis-
plates with a fixed magnification factor could not be used tance was 1.15 m. A metallic radio-opaque ball with a stand-
anymore. Image size of X-rays is no longer standardised and ardised diameter of 25 mm was used as a reference for deter-
can vary. Thus, digital images must be calibrated to scale the mining the magnification factor. The metallic radio-opaque
dimensions shown [9]. This led to difficulties in determining ball was placed next to the greater trochanter and had to be
the correct magnification factor, especially for obese patients projected in total.
[6, 10–12]. Highlighting surgical results, better results and
lower revision rates of orthopaedic consultants compared
with residents could be shown when performing THA [13, Preoperative digital templating
14]. But still, there are little data focusing the impact of the
planner’s experience affecting the accuracy of predicting Preoperative digital templating has been performed by either
component size using a digital preoperative planning soft- residents or consultant surgeons with the EndoMap soft-
ware [15]. Component design is expected to be important for ware system (Siemens Medical Solutions AG, Erlangen,
the surgical outcome, but there are little data concerning the Germany). All of the surgeries were performed by consult-
effect of implant design on planning accuracy [15]. ant surgeons. The magnification factor, leg length, femoral
Therefore, we aimed to analyse factors (planner’s expe- offset and femoral neck length were determined in the ante-
rience, body mass index (BMI), sex, implant design) that rior–posterior radiographs of the hip. Subsequently, digital
might influence the accuracy of preoperative digital templat- templates were used for estimating correct component size
ing in patients who underwent THA retrospectively. in the anterior–posterior radiographs of the hip. Leg length
discrepancies had to be avoided by choosing correct femoral
head size.
Materials and methods Accuracy of preoperative planning was determined
as described before by comparing the difference between
In this study, the data of 903 patients who underwent pri- planned and implanted component sizes as documented in
mary uncemented THA consecutively at our department the surgical report [8, 11, 15–18]. Perfect matches and a var-
between January 2012 and December 2015 were reviewed iance of +/− one size were considered to be adequate. Devi-
retrospectively. Female and male patients of 18–99 years of ations of more than one size were considered inaccurate.
age with primary osteoarthritis of the hip were included. Furthermore, planning accuracy has been related to the
Exclusion criteria were the following: prior surgical inter- planner’s experience defined by the status of consultant sur-
ventions in the hip, cemented THA, revision surgery, intra- geons or residents.
operative complications such as periprosthetic fractures,
malalignment of the femoral stem in postoperative a.p. Statistical analysis
X-rays (defined as 5° < varus or valgus). The exclusion cri-
teria were applied to all identified 903 patients. Statistical analyses have been calculated with SPSS version
The included implants were Allofit cup and Alloclassic 20 (IBM SPSS statistics, Chicago, IL, USA). The level of
femoral stem (both Zimmer Inc., Warsaw, IN, USA) and significance was p < 0.05. Descriptive statistics were applied
Pinnacle cup and Corail femoral stem (both DePuy Syn- for sex, age and BMI. The influences of the planner’s expe-
thes Inc., Warsaw, IN, USA). The Allofit cup is a spheri- rience the component manufacturer and the patient’s sex
cal cup that is flattened at the polar zone. The cup design on accuracy were analysed by Mann–Whitney U test. The
allows press fit implantation technique. The Alloclassic has impact of BMI according to WHO criteria on planning

13
Archives of Orthopaedic and Trauma Surgery

Table 1  Patients’ demographic data and residents in predicting stem size within +/− one size
Age (mean) 65.7 years ± 12.1 SD
was statistically significant (z = −2.111, p = 0.035).
Occurence Percentage
Regarding cup size, differences in planning accuracy were
not statistically significant (z = 0.642, p = 0.52).
Sex
 Male 282 45.0 BMI and planning accuracy
 Female 350 55.0
BMI (WHO-classification in kg/m²) Regarding body weight, patients were split into four groups
 Underweight 3 0.5 of BMI according to the WHO criteria: underweight, nor-
 Normal weight 181 28.6 mal weight, overweight and obese. Underweight persons
 Overweight 279 44.1 were ignored for calculation as this group consisted of
 Obese 169 26.7 three patients. The other groups have been analysed with
Kruskal–Wallis test and paired post hoc tests.
Accurate stem size was predicted in 91.2% (n = 165) for
accuracy was investigated by Kruskal–Wallis test and paired normal weight, in 73.5% (n = 233) for overweight and in
post hoc tests. 87% (n = 147) for obese people. According to this, a high
BMI leads to an inaccurate planning of femoral compo-
nent size [H(2) = 6.05, p = 0.049]. Nevertheless, paired post
Results hoc tests only documented a statistically significant differ-
ence of normal weight patients compared with overweight
With respect to exclusion and inclusion criteria, 632 out of (p = 0.043), but not with obese, as expected.
903 cases were included in this study. 55% (n = 350) were Concerning cup size, 81.2% (n = 147) of normal weight
female and 45% (n = 282) were male patients. The mean age patients, 77.4% (n = 216) of overweight and 76.9% (n = 130)
was 65.7 (± 12.1) years. Demographic data of patients can of obese have been planned adequately. Post hoc calculations
be seen in Table 1. showed no statistically significance. Therefore, the impact of
According to WHO criteria, 0.5% (n = 3) were under- BMI on planning accuracy of the cup could not be approved
weight, 29% (n = 181) were in normal range, 44% (n = 279) [H(2) = 0.20, p = 0.901].
were overweight and 27% (n = 169) were obese.
59% (n = 371) of preoperative planning have been per- Component design
formed by consultant surgeons, 41% (n = 261) by residents
in orthopaedic surgery. Stem components have been planned adequately in 85.3%
74% (n = 469) of the used component designs were DePuy (n = 402) in case of DePuy Corail, and in 89% (n = 145) in
Synthes Inc. (Warsaw, IN, USA), whereas 26% (n = 163) case of Zimmer Alloclassic. Cup size has been predicted
were from Zimmer Inc. (Warsaw, IN, USA). correctly for DePuy Pinnacle in 79.3% (n = 372) and for
Zimmer Allofit in 75.5% (n = 123). Mann–Whitney U test
General reliability attested no statistically significant difference in planning
accuracy regarding different implant designs (stem: z =
The exact stem size was predicted in 42% (n = 264). Further, − 1.13, p = 0.273; cup: z = − 0.234, p = 0.819).
45% (n = 283) were within a range of +/− one size. Thus,
87% (n = 547) of the stems have been measured accurately. Sex and planning accuracy
Data of stem planning accuracy are presented in Table 2.
37% (n = 231) of the cups were predicted correctly, Femoral components have been planned correctly in 89.1%
whereas further 42% (n = 263) were within +/− one size. (n = 312) for female patients and in 83.3% (n = 235) for male
Altogether, cups were estimated correctly in 78% (n = 494). patients. Cup planning has been adequate in 80.9% (n = 283)
Data of cup planning accuracy are presented in Table 2. for females and in 75.2% (n = 212) for males. Analysing gen-
der differences, Mann–Whitney U test showed no statistical
Planner’s experience

Consultants predicted femoral component size correctly

in 87.6% (n = 547) of the cases and residents in 85.1% 1
  U(261,371) = 44.060.
(n = 325) of the cases. The difference between consultants 2
  U(261, 370) = 49.628.
3
  U(469, 163) = 36.211.
4
  U(469, 162) = 37.562.

13
 Archives of Orthopaedic and Trauma Surgery

Table 2  Planning accuracy and deviation of implants in absolute values and percentage

Stem size Cup size
Occurence Percentage Occurence Percentage

Implant size
 Perfect match 264 42.0 231 37.0
 +/− 1 size 283 45.0 263 42.0
 +/− 2 size 67 10.6 99 15.7
 +/− 3 sizes and more 18 2.8 39 6.2
Adequate femoral planning Adequate acetabular planning

Planner’s experience
 Consultants 547 87.6 281 75.7
 Residents 325 85.1 214 82.0
z = − 2.11; p = 0.035 z = 0.64; p = 0.52
BMI
 Underweight Eliminated because of low occurence
 Normal weight 165 91.2 147 81.2
 Overweight 233 73.5 216 77.4
 Obese 147 87.0 130 76.9
z = 6.05; p = 0.049 z = 0.20; p = 0.901
Component design
 Zimmer 145 89.0 123 75.5
 DePuy 402 85.3 372 79.3
z = − 1.1; p = 0.273 z = − 0.23; p = 0.819
Sex
 Female 312 89.1 283 80.9
 Male 235 83.3 212 75.2
z = 1.52; p = 0.13 z = 1.87; p = 0.061

Deviation of 1 size is considered to be adequate

significant difference, too (stem: z = 1.525, p = 0.13; cup: 66% of their plannings were within +/− one size for unce-
z = 1.876, p = 0.061). mented stems and 52% for uncemented cups [8]. Bertz et al.
[16] predicted femoral component size in 95% (+/− one
size) and acetabular component size in 94% (+/− one size)
Discussion in 129 patients with either cemented or hybrid THA. But it
should be taken into consideration, that they also included
We retrospectively analysed the accuracy of 632 preopera- THA with cemented fixation. Cementation might bias results
tive digital THA templates. The femoral components could of preoperative planning accuracy. With respect to this data,
be predicted correctly with a range of +/− one size in 87% preoperative digital templating is a helpful tool in the preop-
and acetabular components in 78%. Similar results were erative management of THA. However, it has to be pointed
published by Davila et al. [11] who investigated planning out that our review of the literature focuses primarily on
accuracy in 36 patients who had undergone THA. They studies documenting stratified results on uncemented THA.
predicted stem size in 86% and cup size in 72% correctly With respect to the present data, higher levels of expe-
(+/− one size). Efe et al. [18] found that in 169 of their rience lead to a statistically significant higher percentage
included patients’ hips stems were planned in 82.3% and of adequate preoperative planning concerning femoral
cups in 77.5% correctly (+/− one size). The et al. studied components, but not for acetabular components. This
173 patients with THA who had preoperative plannings. corresponds to the results of Jung et al. [15]. They also
showed a positive effect of higher experience on predict-
ing stem size, but not on cup size. Carter et al. [19] docu-
5
  U(282, 350) = 52.528. mented the influence of experience on digital templating
6
  U(281,350) = 53.159. too, but they were only focusing on femoral components.

13
Archives of Orthopaedic and Trauma Surgery

They compared the results of three planners in different Conclusion

levels of experience and documented a positive impact of
experience on planning accuracy. High level of experience With 87% of adequately planned femoral components and
led to 95% adequate planning, moderate to 88% and low 78% of adequate acetabular components, preoperative digi-
experience to 82%. tal planning is reliable for predicting implant size in THA.
BMI affects planning accuracy of femoral components, Therefore, preoperative digital templating should be inte-
too. Nevertheless, statistically significance can only be grated as a routine part in the preoperative management
shown for the comparison of normal weight and over- of THA. Furthermore, higher levels of experience lead to
weight people. Due to mispositioning of radio-opaque significantly more precise predictions of stem size. BMI
reference objects when performing X-rays and conse- partly influences digital templating. Overweight is related
quently resulting magnification errors, a negative effect to inaccurate preoperative planning more often. Yet, this is
on planning accuracy could be expected [17]. However, not valid for predicting cup size. Therefore, special attention
this negative influence of obesity has not been confirmed should be paid in overweight and obese patients.
statistically. This matches with the results of Heep et al.
who could not show a correlation of body shape param- Acknowledgements  Open access funding provided by Medical Uni-
eters such as BMI on the magnification of a radio-opaque versity of Graz.
reference object. [20].
Furthermore, no gender-specific differences in planning Compliance with ethical standards 
accuracy could have been seen. Thus, preoperative digital
Conflict of interest  On behalf of all authors, the corresponding author
templating is reliable for both, women and men. Regard- states that there is no conflict of interest.
ing the missing deviations in calibration of X-rays between
men and women, a statistically significant difference was Ethical approval  The study has been approved by the institutional
not expected [20]. review board (EK 25–429 ex 12/13).
Component design of the two different implant systems
that were used in our study did not take influence on predict- Open Access  This article is distributed under the terms of the Crea-
ing implant size. In the present study straight stems were tive Commons Attribution 4.0 International License (http://creat​iveco​
mmons​.org/licen​ses/by/4.0/), which permits unrestricted use, distribu-
used. Jung et al. found superior planning accuracy in straight tion, and reproduction in any medium, provided you give appropriate
stems compared to short stems [15]. credit to the original author(s) and the source, provide a link to the
One reason for the higher percentage of accurate stem Creative Commons license, and indicate if changes were made.
plannings might be the surgical procedure. The correct posi-
tion and size of the cup can be estimated during the sur-
gery by a sufficient acetabular reaming, whereas stem size
is chosen by less obvious factors such as acoustics. There- References
fore, it is dependent on individual subjective perception and
experience. This is of special importance as no intraopera- 1. Eggli S, Pisan M, Müller ME (1998) The value of preopera-
tive planning for total hip arthroplasty. J Bone Joint Surg Br
tive X-rays of stem position have been made in the studied 80(3):382–390
patients. Furthermore, surgeons may tend to the planned 2. Hoikka V, Paavilainen T, Lindholm TS, Turula KB, Ylikoski
stem size with respect to the risk of fractures due to lack of M (1987) Measurement and restoration of equality in length
macroscopically control. of the lower limbs in total hip replacement. Skeletal Radiol
16(6):442–446
When using components of DePuy, a cup size of 52 and 3. de Thomasson E, Mazel C, Guingand O, Terracher R (2002)
bigger allows a femoral head size with a diameter of 36 mm. Etude critique des résultats de la planification préopératoire sur
This optimises stability and prevents dislocations [21–23]. l’anatomie de la hanche prothésée. Rev Chir Orthop Repar Appar
For this reason, surgeons may tend to use greater cup sizes Mot 88(3):229–235
4. Heinert G, Hendricks J, Loeffler MD. Digital templating in hip
than determined in preoperative planning. replacement with and without radiological markers. J Bone Joint
This study has limitations as data analysis was done ret- Surg Br 2009;91-B:459 – 62
rospectively. Furthermore, templating was done by either 5. Kim YH, Kim VE (1993) Uncemented porous-coated anatomic
consultants or residents. In case of a prospective study, digi- total hip replacement: results at 6 years in a consecutive series. J
Bone Joint Surg Br 75-B:6–13
tal planning should be done by both to analyse intraobserver 6. Della Valle AG, Padgett DE, Salvati EA (2005) Preoperative plan-
variability. Furthermore, consultants could be biased as they ning for primary total hip arthroplasty. J Am Acad Orthop Surg
did the planning and performed the THA. In a prospective 13(7):455–462
study and perfect setting templating should be performed 7. Kosashvili Y, Shasha N, Olschewski E, Safir O, White L, Gross
A et al (2009) Digital versus conventional templating techniques
blinded. Another point that limits our study is the lack of in preoperative planning for total hip arthroplasty. Can J Surg
a check on both intraobserver and interobserver reliability. 52(1):6–11

13
 Archives of Orthopaedic and Trauma Surgery

8. The B, Diercks RL, van Ooijen PMA, van Horn JR (2005) Com- 16. Bertz A, Indrekvam K, Ahmed M, Englund E, Sayed-Noor AS
parison of analog and digital preoperative planning in total hip (2012) Validity and reliability of preoperative templating in total
and knee arthroplasties. A prospective study of 173 hips and 65 hip arthroplasty using a digital templating system. Skeletal Radiol
total knees. Acta orthop 76(1):78–84 41(10):1245–1249
9. White SP, Shardlow DL (2005) Effect of introduction of digital 17. Della Valle AG, Comba F, Taveras N, Salvati EA (2008) The util-
radiographic techniques on pre-operative templating in orthopae- ity and precision of analogue and digital preoperative planning for
dic practice. Ann R Coll Surg Engl 87(1):53–54 total hip arthroplasty. Int Orthop 32(3):289–294
10. Franken M, Grimm B, Heyligers I (2010) A comparison of four 18. Efe T, El Zayat BF, Heyse TJ, Timmesfeld N, Fuchs-Winkelmann
systems for calibration when templating for total hip replacement S, Schmitt J (2011) Precision of preoperative digital templating in
with digital radiography. J Bone Joint Surg Br 92(1):136–141 total hip arthroplasty. Acta Orthop Belgica 77(5):616–621
11. Davila JA, Kransdorf MJ, Duffy GP (2006) Surgical plan- 19. Carter LW, Stovall DO, Young TR (1995) Determination of accu-
ning of total hip arthroplasty: accuracy of computer-assisted racy of preoperative templating of noncemented femoral prosthe-
EndoMap software in predicting component size. Skeletal Radiol ses. J Arthroplast 10(4):507–513
35(6):390–393 20. Heep H, Xu J, Löchteken C, Wedemeyer C (2012) A simple and
12. Sinclair VF, Wilson J, Jain NP, Knowles D (2014) Assessment of convenient method guide to determine the magnification of digital
accuracy of marker ball placement in pre-operative templating for X-rays for preoperative planning in total hip arthroplasty. Orthop
total hip arthroplasty. J Arthroplast 29(8):1658–1660 Rev 4(1):e12
13. Marston RA, Cobb AG, Bentley G (1996) Stanmore compared 21. Jameson SS, Lees D, James P, Serrano-Pedraza I, Partington PF,
with charnley total hip replacement. A prospective study of 413 Muller SD, Meek RMD, Reed MR (2011) Lower rates of dislo-
arthroplasties. J Bone Joint Surg 78(2):178–184 cation with increased femoral head size after primary total hip
14. Inglis T, Dalzell K, Hooper G, Rothwell A, Frampton C (2013) replacement. J Bone Joint Surg 93:876–80
Does orthopedic training compromise the outcome in total hip 22. Allen CL, Hooper GJ, Frampton CMA (2014) Do larger femoral
arthroplasty? J Surg Educ 70(1):76–80 heads improve the functional outcome in total hip arthroplasty? J
15. Jung S, Neuerburg C, Kappe T, Wernerus D, Reichel H, Bieger R Arthroplast 29(2):401–404
(2012) Validität der computergestützten Hüfttotalendoprothesen- 23. Singh SP, Bhalodiya HP (2013) Head size and dislocation rate in
planung: Einfluss von Schaftdesign und Untersuchererfahrung. primary total hip arthroplasty. Indian J Orthop 47(5):443–448
Zeitschrift für Orthopädie Unfallchirurgie 150(4):404–408

13