The Journal of Arthroplasty xxx (2014) xxx–xxx
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
Original Article
Assessment of Accuracy of Marker Ball Placement in Pre-operative Templating for Total Hip Arthroplasty Victoria F. Sinclair, MRCS (Eng) a, James Wilson, MRCS (Ed) a, Neil P.M. Jain, FRCS (Tr & Orth) b, David Knowles, Bsc MB BChir, FRCS, FRCS (Tr & Orth) a a b
Royal Lancaster Infirmary, Lancaster, United Kingdom University Hospital South Manchester, Manchester, United Kingdom
a r t i c l e
i n f o
Article history: Received 2 April 2013 Accepted 14 March 2014 Available online xxxx Keywords: templating marker ball position total hip arthroplasty
a b s t r a c t We report the accuracy of positioning of the calibration ball in the process of pre-operative templating for total hip arthroplasty (THA). The ball should be placed in the coronal plane of the hip to provide suitable accuracy. We reviewed 112 post-operative THA radiographs where a calibration ball had been placed. We templated the femoral head size of the implant after calibrating the templating system from the ball. The calibrated femoral head diameter was compared to the known prosthetic head size. A percentage error was calculated. Overall, incorrect placement of the calibration ball resulted in a mean percentage error in templating of 6.8% (median 5.7%; range 0–26%). Such error carries implications with the templating process and may result in incorrect component sizes, leg lengths and offset. © 2014 Elsevier Inc. All rights reserved.
Pre-operative templating in total hip arthroplasty has become a valuable part of pre-operative planning for the surgeon. It aids the surgeon to provide a hip arthroplasty that will reproduce the center of rotation of the hip joint. In doing so it corrects offset and leg length, guides component size and position and enables adequate soft tissue tensioning to create a stable hip arthroplasty [1–3]. In order to calibrate the scale of a digital radiograph of the hip, a calibration ball of known size must be placed in the same coronal (or antero-posterior) plane as the hip so as to be at a similar distance from the X-Ray plate. Magnification errors will occur if the ball is significantly closer to or further away from the X-Ray plate than the femoral head to be measured. By placing the ball to the side of the hip, either between the patient’s legs or lateral to the thigh, a small but unavoidable magnification error is introduced, hence the importance of not placing the ball in front of or behind the coronal plane of the hip as that would further increase this error. With the use of digital radiograph images now becoming commonplace, the traditional acetate templates historically used on an analogue radiograph have been replaced by the use of Computer Aided Design (CAD) software techniques. In order for these techniques to accurately template, a calibration device is required, usually in the form of a ball [4]. As mentioned it is traditionally placed either medially between the patient’s legs or laterally, both in an
The Conflict of Interest statement associated with this article can be found at http:// dx.doi.org/10.1016/j.arth.2014.03.013. Reprint requests: Neil P.M. Jain, FRCS (Tr & Orth), Specialist Trainee Registrar in Trauma and Orthopaedics, University Hospital South Manchester, Southmoor Road, Manchester M23 9LT, United Kingdom.
attempt to place the ball in the plane of the hip joint to minimize any subsequent errors in sizing. However difficulties have been described with these techniques, with the medial placement of the ball raising complaints about violation of a patient’s dignity or placement not in the plane of the hip in an obese patient (i.e. the ball is pushed too anteriorly by the excess fat on the anterior aspect of the patient’s thighs). In the obese patient there is also the possibility that lateral placement of the ball will also prove ineffective if the body mass of the patient results in such pelvic width that it fills the digital image and the ball is pushed so far laterally that either only a portion of it appears on the image or the ball does not appear at all [4]. Retrospectively, utilizing the fact that the implanted femoral head size is known, and the fact that it is by definition known to lie in the plane of the hip, it is possible to calculate the accuracy of the calibration ball placement. Calibrating the post-operative digital radiograph with the calibration ball, measuring the arthroplasty femoral head size on the calibrated radiograph and then comparing this measurement of femoral head size to the known head size implanted can achieve this. This provides a percentage error of the calibration ball templating process and infers the percentage error of the calibration ball placement. Therefore the aim of this study was to assess the accuracy of calibration ball placement in order to determine whether the operating surgeon could find the digital templating results to be reliable. Our purpose was to report on the errors observed with placement of the calibration ball in digital radiographs. Such errors would have an impact upon the reliability of templating and determine how accurate the templating process, which provides suggested sizes of arthroplasty components, is within the pre-operative planning process.
http://dx.doi.org/10.1016/j.arth.2014.03.013 0883-5403/© 2014 Elsevier Inc. All rights reserved.
Please cite this article as: Sinclair VF, et al, Assessment of Accuracy of Marker Ball Placement in Pre-operative Templating for Total Hip Arthroplasty, J Arthroplasty (2014), http://dx.doi.org/10.1016/j.arth.2014.03.013
2
V.F. Sinclair et al. / The Journal of Arthroplasty xxx (2014) xxx–xxx
Materials and Methods We performed a retrospective single blinded study that reviewed 112 total hip arthroplasties that had undergone pre-operative templating with the placement of a calibration ball. The post-operative digital radiographs were reviewed on the Picture Archiving and Communication Systems (PACS) service provided by GE Centricity. Each radiograph demonstrated at least one total hip arthroplasty with a femoral head component. A calibration ball had also been placed on the radiograph. Using the TraumaCad software hip templating program (Orthocrat Ltd, Petach-Tikva, Israel) the digital radiographs were calibrated according to the known size of the standard calibration ball, 2.54 cm (or 1 inch) in diameter placed by a radiographer in the coronal plane of the hip [5]. The radiographers supervising these radiographs had received appropriate departmental training in the placement of the ball to attempt to place it as best as possible in the coronal plane of the hip. This training involved the radiographers identifying the greater trochanter as an anatomical landmark on which to orientate them with the hip before the radiograph was performed with the patient supine with the patient’s legs in slight abduction and internal rotation. The beam source is 100 cm from the X-Ray plate and at 90° to the table [6]. Ideally the distance between the coronal planes of the hip would have been recorded and then compared to the accuracy. However, due to the variations in body mass and the difficulty in clinically determining where the plane of the hip actually is, this was not possible. The software of TraumaCad has been shown to reliably calibrate the radiograph according to calibrating ball size [6]. This is facilitated by the software, which essentially calculates the true distances along the radiograph digital image after referring to a known object on the screen of a known size, i.e. the calibration ball [5]. This allows the use of a digital ruler on the screen within the software program so that the operator is able to make linear length measurements on the radiograph that carry some accuracy. As a result a measurement of diameter of the arthroplasty femoral head size was achieved. This distance was recorded as ‘templated head size’ (THS). This process was carried out independently by two senior orthopedic trainees (VS and JW) each on two separate occasions. Their observations were compared to assess for their accuracy in templating. This provided a guide to any inter-observer and intra-observer error that the process may have incurred. The operation note of the patient was then reviewed to determine the ‘actual head size’ (AHS) used at surgery (28 mm in most cases). The THS was then compared to the AHS and any difference was recorded as a percentage of the AHS and termed the ‘percentage error’ (PE). The inter-observer error was recorded as a percentage difference between the THS for each trainee.
Results The 112 total hip arthroplasties were reviewed in 92 patients, 60 female and 32 male. The mean age of the patients was 72.5 years (range 52–98 years). The range of femoral head component sizes used is described in Table 1. The mean PE between THS and AHS was 6.8% (range 0–26%) with a median value of 5.7%.
Table 1 Distribution of Femoral Head Sizes Used with Their Percentage Errors (PE). Actual Prosthetic Head Size (mm) 22.24 28 32 36
Number of Patients
Mean PE (%)
Median PE (%)
Range of PE (%)
25 80 4 3
8.2 6.4 10.2 5.0
7.2 6.3 9.8 4.3
0.9–21.6 0–17.1 9.4–11.3 3.3–7.1
The variation between observers was shown to be a median interobserver error of 0.92% and a mean of 0.88% (range 0–3.57%). The variation seen for intra-observer error was a median of 0.36% and a mean of 0.51% (range 0–2.86%). The trend was to under-estimate the AHS with the THS, with an undersized THS being suggested in 63.4% of cases. No trend was observed relating to PE and AHS used. Discussion Pre-operative templating for total hip arthroplasty has gained popularity as part of the pre-operative planning process for the surgeon [1–3]. Most of the literature concerning this has thus far investigated the accuracy of templating in terms of the size of the components introduced at the time of surgery. The accuracy of templating is often described as being accurate to within one component size [5,7]. With our study, we have demonstrated a different methodology to that used by most templating studies [5,7,8], although still having been previously described [4]. We have demonstrated that the positioning of the calibration ball is responsible for a mean 6.8% error in accuracy in terms of size of the femoral head. Such inaccuracy may also suggest inaccuracy with the acetabulum size and femoral component size. Using the example of a 50 mm acetabular uncemented component a 6.8% inaccuracy would result in a potential variation in templated size between 47 mm (minus 6.8% error) and 53 mm (plus 6.8% error). This represents a 6 mm size difference in the potential acetabular component required compared with the potential size templated. Assuming that the error is an under sizing of the component then the error becomes 3 mm. However, this still represents 3 potential component size differences, which is greater than the usual one size described in the literature. Such differences, if heavily relied upon by the surgeon may have disastrous consequences. These include equipment and operative technique consequences. With streamlining and cost-cutting becoming more prevalent in United Kingdom hospitals, templating may become more relied upon in order to determine the appropriate component sizes required for a procedure. Inadequate accuracy of the templating technique requires the full inventory of component sizes to be available to confidently provide the suitable size for the procedure. Failure to have the full selection available may result in a patient safety issue. The templating findings from this study highlight the importance of having the full range of component sizes available for every individual operation and demonstrate the importance of an accurate inventory of stock. Regarding operative technique, if the acetabular component is oversized then intra-operative fracture may occur. If it is undersized then component loosening may be seen. The same principle of fracture in oversized and loosening in undersized components concerns a cementless femoral stem size. Such errors in component size also carry implications to the positioning of the prosthesis and therefore such factors as leg length and off-set thereby also influencing the stability of the joint [9]. In reality, templating forms merely part of the pre-operative plan and as a result is used merely as a guide rather than an actuality. One would hope that common sense would prevail and that the surgeon would not risk fracture in pursuit of achieving the templated size, rather the surgeon would use tactile feedback to appreciate when to limit their reaming process. Nevertheless, in the stressful environment of theater an inexperienced surgeon may not have such an appreciation for any imminent fractures and thereby relies upon their templating to a greater degree, highlighting the importance of our findings. Ultimately our findings suggest that while pre-operative templating may well be useful as part of formal pre-operative planning for total hip arthroplasty surgery, the inconsistent calibration ball placement contributes to the prediction of prosthesis size being inaccurate to a point where it cannot be relied upon in the surgery. The incorrect positioning of the calibration ball is highlighted by the finding that most
Please cite this article as: Sinclair VF, et al, Assessment of Accuracy of Marker Ball Placement in Pre-operative Templating for Total Hip Arthroplasty, J Arthroplasty (2014), http://dx.doi.org/10.1016/j.arth.2014.03.013
V.F. Sinclair et al. / The Journal of Arthroplasty xxx (2014) xxx–xxx
of the THS were undersized compared to the AHS which suggests that the calibration ball is measured relatively too big to the femoral head size. This occurs because the calibration ball is effectively anterior to the femoral head and therefore closer to the beam, ultimately not being in the same plane as the hip joint which remains the prerequisite to providing accurate templating. The process of templating is reproducible and has a low inter-observer and intra-observer percentage error from our observations, suggesting that the errors in THS and AHS are as a result of the calibration ball position rather than the templating process itself. Therefore a technique that provided a consistent form of calibration may provide a consistently accurate form of templating for component size [10]. This may be achieved by regulatory calibration ball placement. However, this is not as straightforward a prospect as one may initially believe. The body mass of every patient is different and as a result, surface landmarks of the hip such as the greater trochanter will be more easily palpable in some individuals than others. Similarly, femoral neck anteversion is variable between patients and this would have an impact upon where to place any calibration ball in order for it to be in the plane of the hip joint. By placing the calibration ball to the side of the hip, either between the patient’s legs or lateral to the thigh, a small but unavoidable magnification error is introduced, hence the importance of not placing the ball in front of or behind the coronal plane of the hip further increasing this error. In our study the beam source of the radiograph was 100 cm from the X-Ray plate and at 90° to the table. As mentioned, the distance between the X-Ray plate and the determined coronal plane of the hip would have been recorded and then compared to the accuracy to determine the effect of changes in this distance. This was not deemed possible due to the variations in body mass and the difficulty in clinically determining where the plane of the hip truly is, rather than the assumption made by the positioning of the legs and referencing the greater trochanter as performed in this study. A trigonometry calculation (Tan (angle) = opposite/adjacent) can offer some guidance to the effect of varied positioning of the calibration ball. Assuming the source beam is 100 cm from the X-Ray plate and that the true femoral head center lies 10 cm in front of the plate, and 90 cm from the source beam, and that each hip is 10 cm lateral of the midline, moving the calibration ball between the legs (i.e. 10 cm medially) while remaining in the coronal plane of the hip would result in a magnification error of − 0.61%. Moving the ball 10 cm laterally in the coronal plane of the hip would generate a 1.81% magnification error. However, moving the ball just 1 cm towards the beam (now 89 cm from the beam) would result in a magnification error of 1.12%, suggesting that greater error is caused by anteroposterior malposition and thereby emphasizing the importance of obtaining a calibration ball in the coronal plane of the hip if that is to be used for pre-operative templating purposes in digital radiographs. Our findings suggest that a mean error of 6.8% likely represents error induced from a combination of placement of the calibration ball laterally and incorrectly in the antero-posterior. Such error is likely representative of the difficulty in estimating the coronal plane of the hip for the radiographer when placing the calibration ball. This is not aided in patients that have a large amount of subcutaneous fat, which not only will make it difficult to palpate the greater trochanter but also by definition will force the position of the ball to even more lateral
3
than in a patient of less fat and be of difficulty in placement in the coronal plane of the hip. This fact highlights the necessity for vigilance by the radiographer performing the X-Ray. Ultimately, it may be that radiographs are not a suitable method for templating due to such difficulties to place any calibration ball in the plane of the hip. It is possible that Computer Tomography (CT) scanning would provide greater accuracy, however this would involve greater cost and greater radiation exposure [10]. Such a contrast in the benefits and risks highlights the issues surrounding the process of accurate pre-operative templating in hip arthroplasty and the difficulty in placing the calibration ball in the plane of the hip in radiographs. The strengths or our study are that we have highlighted the importance of the correct placement of the calibration ball for templating and in particular have confirmed that templating should not be used as an absolute guide to a surgeon as part of their preoperative planning and rather be used as a relative guide. We also highlight the suggestion that 3-dimensional imaging (such as with a CT Scan) may have a greater accuracy with regard to preoperative templating rather than the traditional 2-dimensional imaging (plain radiographs) [10]. Our weaknesses are the limits in the patient numbers, the limited number of observers and the fact that we could not compare the radiograph templating to any CT scan templating to confirm an increase in accuracy. Future studies could focus on a comparison of these templating modalities or modify the radiographic templating technique to have specific points for the X-Ray source, plate and calibration device, although, as has already been mentioned, this would be difficult to achieve practically due to the variations in human morphology. Ultimately, until the point where a more accurate method of templating is fully established and demonstrated to be reproducible, current radiographic templating using a calibration ball should only be used as a guide and not an absolute when the surgeon is determining correct implant size for the patient. References 1. Jassim SS, Ingham C, Keeling M, et al. Digital templating facilitates accurate leg length correction in total hip arthroplasty. Acta Orthop Belg 2012;78(3):344. 2. Efe T, El Zayat BF, Heyse TJ, et al. Precision of preoperative digital templating in total hip arthroplasty. Acta Orthop Belg 2011;77(5):616. 3. Levine B, Fabi D, Deirmengian C. Digital templating in primary total hip and knee arthroplasty. Orthopedics 2010;33(11):797. 4. 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. 5. Steinberg EL, Shasha N, Menahem A, et al. Preoperative planning of total hip replacement using the TraumaCad™ system. Arch Orthop Trauma Surg 2010;130 (12):1429. 6. Westacott DJ, McArthur J, King RJ, et al. Assessment of cup orientation in hip resurfacing: comparison of TraumaCad and computed tomography. J Orthop Surg Res 2013;11(8):8. 7. Gamble P, de Beer J, Petruccelli D, et al. The accuracy of digital templating in uncemented total hip arthroplasty. J Arthroplasty 2010;25(4):529. 8. Unnanuntana A, Wagner D, Goodman SB. The accuracy of preoperative templating in cementless total hip arthroplasty. J Arthroplasty 2009;24(2):180. 9. Padgett DE, Warashina H. The unstable total hip replacement. Clin Orthop Relat Res 2004;420:72. 10. Sariali E, Mauprivez R, Khiami F, et al. Accuracy of the preoperative planning for cementless total hip arthroplasty. A randomised comparison between threedimensional computerised planning and conventional templating. Orthop Traumatol Surg Res 2012;98(2):151.
Please cite this article as: Sinclair VF, et al, Assessment of Accuracy of Marker Ball Placement in Pre-operative Templating for Total Hip Arthroplasty, J Arthroplasty (2014), http://dx.doi.org/10.1016/j.arth.2014.03.013
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
Original Article
Assessment of Accuracy of Marker Ball Placement in Pre-operative Templating for Total Hip Arthroplasty Victoria F. Sinclair, MRCS (Eng) a, James Wilson, MRCS (Ed) a, Neil P.M. Jain, FRCS (Tr & Orth) b, David Knowles, Bsc MB BChir, FRCS, FRCS (Tr & Orth) a a b
Royal Lancaster Infirmary, Lancaster, United Kingdom University Hospital South Manchester, Manchester, United Kingdom
a r t i c l e
i n f o
Article history: Received 2 April 2013 Accepted 14 March 2014 Available online xxxx Keywords: templating marker ball position total hip arthroplasty
a b s t r a c t We report the accuracy of positioning of the calibration ball in the process of pre-operative templating for total hip arthroplasty (THA). The ball should be placed in the coronal plane of the hip to provide suitable accuracy. We reviewed 112 post-operative THA radiographs where a calibration ball had been placed. We templated the femoral head size of the implant after calibrating the templating system from the ball. The calibrated femoral head diameter was compared to the known prosthetic head size. A percentage error was calculated. Overall, incorrect placement of the calibration ball resulted in a mean percentage error in templating of 6.8% (median 5.7%; range 0–26%). Such error carries implications with the templating process and may result in incorrect component sizes, leg lengths and offset. © 2014 Elsevier Inc. All rights reserved.
Pre-operative templating in total hip arthroplasty has become a valuable part of pre-operative planning for the surgeon. It aids the surgeon to provide a hip arthroplasty that will reproduce the center of rotation of the hip joint. In doing so it corrects offset and leg length, guides component size and position and enables adequate soft tissue tensioning to create a stable hip arthroplasty [1–3]. In order to calibrate the scale of a digital radiograph of the hip, a calibration ball of known size must be placed in the same coronal (or antero-posterior) plane as the hip so as to be at a similar distance from the X-Ray plate. Magnification errors will occur if the ball is significantly closer to or further away from the X-Ray plate than the femoral head to be measured. By placing the ball to the side of the hip, either between the patient’s legs or lateral to the thigh, a small but unavoidable magnification error is introduced, hence the importance of not placing the ball in front of or behind the coronal plane of the hip as that would further increase this error. With the use of digital radiograph images now becoming commonplace, the traditional acetate templates historically used on an analogue radiograph have been replaced by the use of Computer Aided Design (CAD) software techniques. In order for these techniques to accurately template, a calibration device is required, usually in the form of a ball [4]. As mentioned it is traditionally placed either medially between the patient’s legs or laterally, both in an
The Conflict of Interest statement associated with this article can be found at http:// dx.doi.org/10.1016/j.arth.2014.03.013. Reprint requests: Neil P.M. Jain, FRCS (Tr & Orth), Specialist Trainee Registrar in Trauma and Orthopaedics, University Hospital South Manchester, Southmoor Road, Manchester M23 9LT, United Kingdom.
attempt to place the ball in the plane of the hip joint to minimize any subsequent errors in sizing. However difficulties have been described with these techniques, with the medial placement of the ball raising complaints about violation of a patient’s dignity or placement not in the plane of the hip in an obese patient (i.e. the ball is pushed too anteriorly by the excess fat on the anterior aspect of the patient’s thighs). In the obese patient there is also the possibility that lateral placement of the ball will also prove ineffective if the body mass of the patient results in such pelvic width that it fills the digital image and the ball is pushed so far laterally that either only a portion of it appears on the image or the ball does not appear at all [4]. Retrospectively, utilizing the fact that the implanted femoral head size is known, and the fact that it is by definition known to lie in the plane of the hip, it is possible to calculate the accuracy of the calibration ball placement. Calibrating the post-operative digital radiograph with the calibration ball, measuring the arthroplasty femoral head size on the calibrated radiograph and then comparing this measurement of femoral head size to the known head size implanted can achieve this. This provides a percentage error of the calibration ball templating process and infers the percentage error of the calibration ball placement. Therefore the aim of this study was to assess the accuracy of calibration ball placement in order to determine whether the operating surgeon could find the digital templating results to be reliable. Our purpose was to report on the errors observed with placement of the calibration ball in digital radiographs. Such errors would have an impact upon the reliability of templating and determine how accurate the templating process, which provides suggested sizes of arthroplasty components, is within the pre-operative planning process.
http://dx.doi.org/10.1016/j.arth.2014.03.013 0883-5403/© 2014 Elsevier Inc. All rights reserved.
Please cite this article as: Sinclair VF, et al, Assessment of Accuracy of Marker Ball Placement in Pre-operative Templating for Total Hip Arthroplasty, J Arthroplasty (2014), http://dx.doi.org/10.1016/j.arth.2014.03.013
2
V.F. Sinclair et al. / The Journal of Arthroplasty xxx (2014) xxx–xxx
Materials and Methods We performed a retrospective single blinded study that reviewed 112 total hip arthroplasties that had undergone pre-operative templating with the placement of a calibration ball. The post-operative digital radiographs were reviewed on the Picture Archiving and Communication Systems (PACS) service provided by GE Centricity. Each radiograph demonstrated at least one total hip arthroplasty with a femoral head component. A calibration ball had also been placed on the radiograph. Using the TraumaCad software hip templating program (Orthocrat Ltd, Petach-Tikva, Israel) the digital radiographs were calibrated according to the known size of the standard calibration ball, 2.54 cm (or 1 inch) in diameter placed by a radiographer in the coronal plane of the hip [5]. The radiographers supervising these radiographs had received appropriate departmental training in the placement of the ball to attempt to place it as best as possible in the coronal plane of the hip. This training involved the radiographers identifying the greater trochanter as an anatomical landmark on which to orientate them with the hip before the radiograph was performed with the patient supine with the patient’s legs in slight abduction and internal rotation. The beam source is 100 cm from the X-Ray plate and at 90° to the table [6]. Ideally the distance between the coronal planes of the hip would have been recorded and then compared to the accuracy. However, due to the variations in body mass and the difficulty in clinically determining where the plane of the hip actually is, this was not possible. The software of TraumaCad has been shown to reliably calibrate the radiograph according to calibrating ball size [6]. This is facilitated by the software, which essentially calculates the true distances along the radiograph digital image after referring to a known object on the screen of a known size, i.e. the calibration ball [5]. This allows the use of a digital ruler on the screen within the software program so that the operator is able to make linear length measurements on the radiograph that carry some accuracy. As a result a measurement of diameter of the arthroplasty femoral head size was achieved. This distance was recorded as ‘templated head size’ (THS). This process was carried out independently by two senior orthopedic trainees (VS and JW) each on two separate occasions. Their observations were compared to assess for their accuracy in templating. This provided a guide to any inter-observer and intra-observer error that the process may have incurred. The operation note of the patient was then reviewed to determine the ‘actual head size’ (AHS) used at surgery (28 mm in most cases). The THS was then compared to the AHS and any difference was recorded as a percentage of the AHS and termed the ‘percentage error’ (PE). The inter-observer error was recorded as a percentage difference between the THS for each trainee.
Results The 112 total hip arthroplasties were reviewed in 92 patients, 60 female and 32 male. The mean age of the patients was 72.5 years (range 52–98 years). The range of femoral head component sizes used is described in Table 1. The mean PE between THS and AHS was 6.8% (range 0–26%) with a median value of 5.7%.
Table 1 Distribution of Femoral Head Sizes Used with Their Percentage Errors (PE). Actual Prosthetic Head Size (mm) 22.24 28 32 36
Number of Patients
Mean PE (%)
Median PE (%)
Range of PE (%)
25 80 4 3
8.2 6.4 10.2 5.0
7.2 6.3 9.8 4.3
0.9–21.6 0–17.1 9.4–11.3 3.3–7.1
The variation between observers was shown to be a median interobserver error of 0.92% and a mean of 0.88% (range 0–3.57%). The variation seen for intra-observer error was a median of 0.36% and a mean of 0.51% (range 0–2.86%). The trend was to under-estimate the AHS with the THS, with an undersized THS being suggested in 63.4% of cases. No trend was observed relating to PE and AHS used. Discussion Pre-operative templating for total hip arthroplasty has gained popularity as part of the pre-operative planning process for the surgeon [1–3]. Most of the literature concerning this has thus far investigated the accuracy of templating in terms of the size of the components introduced at the time of surgery. The accuracy of templating is often described as being accurate to within one component size [5,7]. With our study, we have demonstrated a different methodology to that used by most templating studies [5,7,8], although still having been previously described [4]. We have demonstrated that the positioning of the calibration ball is responsible for a mean 6.8% error in accuracy in terms of size of the femoral head. Such inaccuracy may also suggest inaccuracy with the acetabulum size and femoral component size. Using the example of a 50 mm acetabular uncemented component a 6.8% inaccuracy would result in a potential variation in templated size between 47 mm (minus 6.8% error) and 53 mm (plus 6.8% error). This represents a 6 mm size difference in the potential acetabular component required compared with the potential size templated. Assuming that the error is an under sizing of the component then the error becomes 3 mm. However, this still represents 3 potential component size differences, which is greater than the usual one size described in the literature. Such differences, if heavily relied upon by the surgeon may have disastrous consequences. These include equipment and operative technique consequences. With streamlining and cost-cutting becoming more prevalent in United Kingdom hospitals, templating may become more relied upon in order to determine the appropriate component sizes required for a procedure. Inadequate accuracy of the templating technique requires the full inventory of component sizes to be available to confidently provide the suitable size for the procedure. Failure to have the full selection available may result in a patient safety issue. The templating findings from this study highlight the importance of having the full range of component sizes available for every individual operation and demonstrate the importance of an accurate inventory of stock. Regarding operative technique, if the acetabular component is oversized then intra-operative fracture may occur. If it is undersized then component loosening may be seen. The same principle of fracture in oversized and loosening in undersized components concerns a cementless femoral stem size. Such errors in component size also carry implications to the positioning of the prosthesis and therefore such factors as leg length and off-set thereby also influencing the stability of the joint [9]. In reality, templating forms merely part of the pre-operative plan and as a result is used merely as a guide rather than an actuality. One would hope that common sense would prevail and that the surgeon would not risk fracture in pursuit of achieving the templated size, rather the surgeon would use tactile feedback to appreciate when to limit their reaming process. Nevertheless, in the stressful environment of theater an inexperienced surgeon may not have such an appreciation for any imminent fractures and thereby relies upon their templating to a greater degree, highlighting the importance of our findings. Ultimately our findings suggest that while pre-operative templating may well be useful as part of formal pre-operative planning for total hip arthroplasty surgery, the inconsistent calibration ball placement contributes to the prediction of prosthesis size being inaccurate to a point where it cannot be relied upon in the surgery. The incorrect positioning of the calibration ball is highlighted by the finding that most
Please cite this article as: Sinclair VF, et al, Assessment of Accuracy of Marker Ball Placement in Pre-operative Templating for Total Hip Arthroplasty, J Arthroplasty (2014), http://dx.doi.org/10.1016/j.arth.2014.03.013
V.F. Sinclair et al. / The Journal of Arthroplasty xxx (2014) xxx–xxx
of the THS were undersized compared to the AHS which suggests that the calibration ball is measured relatively too big to the femoral head size. This occurs because the calibration ball is effectively anterior to the femoral head and therefore closer to the beam, ultimately not being in the same plane as the hip joint which remains the prerequisite to providing accurate templating. The process of templating is reproducible and has a low inter-observer and intra-observer percentage error from our observations, suggesting that the errors in THS and AHS are as a result of the calibration ball position rather than the templating process itself. Therefore a technique that provided a consistent form of calibration may provide a consistently accurate form of templating for component size [10]. This may be achieved by regulatory calibration ball placement. However, this is not as straightforward a prospect as one may initially believe. The body mass of every patient is different and as a result, surface landmarks of the hip such as the greater trochanter will be more easily palpable in some individuals than others. Similarly, femoral neck anteversion is variable between patients and this would have an impact upon where to place any calibration ball in order for it to be in the plane of the hip joint. By placing the calibration ball to the side of the hip, either between the patient’s legs or lateral to the thigh, a small but unavoidable magnification error is introduced, hence the importance of not placing the ball in front of or behind the coronal plane of the hip further increasing this error. In our study the beam source of the radiograph was 100 cm from the X-Ray plate and at 90° to the table. As mentioned, the distance between the X-Ray plate and the determined coronal plane of the hip would have been recorded and then compared to the accuracy to determine the effect of changes in this distance. This was not deemed possible due to the variations in body mass and the difficulty in clinically determining where the plane of the hip truly is, rather than the assumption made by the positioning of the legs and referencing the greater trochanter as performed in this study. A trigonometry calculation (Tan (angle) = opposite/adjacent) can offer some guidance to the effect of varied positioning of the calibration ball. Assuming the source beam is 100 cm from the X-Ray plate and that the true femoral head center lies 10 cm in front of the plate, and 90 cm from the source beam, and that each hip is 10 cm lateral of the midline, moving the calibration ball between the legs (i.e. 10 cm medially) while remaining in the coronal plane of the hip would result in a magnification error of − 0.61%. Moving the ball 10 cm laterally in the coronal plane of the hip would generate a 1.81% magnification error. However, moving the ball just 1 cm towards the beam (now 89 cm from the beam) would result in a magnification error of 1.12%, suggesting that greater error is caused by anteroposterior malposition and thereby emphasizing the importance of obtaining a calibration ball in the coronal plane of the hip if that is to be used for pre-operative templating purposes in digital radiographs. Our findings suggest that a mean error of 6.8% likely represents error induced from a combination of placement of the calibration ball laterally and incorrectly in the antero-posterior. Such error is likely representative of the difficulty in estimating the coronal plane of the hip for the radiographer when placing the calibration ball. This is not aided in patients that have a large amount of subcutaneous fat, which not only will make it difficult to palpate the greater trochanter but also by definition will force the position of the ball to even more lateral
3
than in a patient of less fat and be of difficulty in placement in the coronal plane of the hip. This fact highlights the necessity for vigilance by the radiographer performing the X-Ray. Ultimately, it may be that radiographs are not a suitable method for templating due to such difficulties to place any calibration ball in the plane of the hip. It is possible that Computer Tomography (CT) scanning would provide greater accuracy, however this would involve greater cost and greater radiation exposure [10]. Such a contrast in the benefits and risks highlights the issues surrounding the process of accurate pre-operative templating in hip arthroplasty and the difficulty in placing the calibration ball in the plane of the hip in radiographs. The strengths or our study are that we have highlighted the importance of the correct placement of the calibration ball for templating and in particular have confirmed that templating should not be used as an absolute guide to a surgeon as part of their preoperative planning and rather be used as a relative guide. We also highlight the suggestion that 3-dimensional imaging (such as with a CT Scan) may have a greater accuracy with regard to preoperative templating rather than the traditional 2-dimensional imaging (plain radiographs) [10]. Our weaknesses are the limits in the patient numbers, the limited number of observers and the fact that we could not compare the radiograph templating to any CT scan templating to confirm an increase in accuracy. Future studies could focus on a comparison of these templating modalities or modify the radiographic templating technique to have specific points for the X-Ray source, plate and calibration device, although, as has already been mentioned, this would be difficult to achieve practically due to the variations in human morphology. Ultimately, until the point where a more accurate method of templating is fully established and demonstrated to be reproducible, current radiographic templating using a calibration ball should only be used as a guide and not an absolute when the surgeon is determining correct implant size for the patient. References 1. Jassim SS, Ingham C, Keeling M, et al. Digital templating facilitates accurate leg length correction in total hip arthroplasty. Acta Orthop Belg 2012;78(3):344. 2. Efe T, El Zayat BF, Heyse TJ, et al. Precision of preoperative digital templating in total hip arthroplasty. Acta Orthop Belg 2011;77(5):616. 3. Levine B, Fabi D, Deirmengian C. Digital templating in primary total hip and knee arthroplasty. Orthopedics 2010;33(11):797. 4. 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. 5. Steinberg EL, Shasha N, Menahem A, et al. Preoperative planning of total hip replacement using the TraumaCad™ system. Arch Orthop Trauma Surg 2010;130 (12):1429. 6. Westacott DJ, McArthur J, King RJ, et al. Assessment of cup orientation in hip resurfacing: comparison of TraumaCad and computed tomography. J Orthop Surg Res 2013;11(8):8. 7. Gamble P, de Beer J, Petruccelli D, et al. The accuracy of digital templating in uncemented total hip arthroplasty. J Arthroplasty 2010;25(4):529. 8. Unnanuntana A, Wagner D, Goodman SB. The accuracy of preoperative templating in cementless total hip arthroplasty. J Arthroplasty 2009;24(2):180. 9. Padgett DE, Warashina H. The unstable total hip replacement. Clin Orthop Relat Res 2004;420:72. 10. Sariali E, Mauprivez R, Khiami F, et al. Accuracy of the preoperative planning for cementless total hip arthroplasty. A randomised comparison between threedimensional computerised planning and conventional templating. Orthop Traumatol Surg Res 2012;98(2):151.
Please cite this article as: Sinclair VF, et al, Assessment of Accuracy of Marker Ball Placement in Pre-operative Templating for Total Hip Arthroplasty, J Arthroplasty (2014), http://dx.doi.org/10.1016/j.arth.2014.03.013
