Reproducibility and Accuracy of Templating Uncemented THA With Digital Radiographic and Digital TraumaCad Templating Software

Abstract

The reproducibility and accuracy of a digital software templating program on digital images was examined for primary uncemented total hip arthroplasty (THA). Forty-five patients waiting for an uncemented THA had templating performed of their digital picture archiving and communication systems (PACS) pelvic images with the TraumaCad software program (Orthocrat Ltd, Petach-Tikva, Israel). Acetabular cup size, femoral stem size, and femoral offset were noted by 2 observers, and again by 1 of the observers a week later. Through correlation coefficients and significance testing, the degree of intraobserver and interobserver variability was evaluated, as well as the level of accuracy for digital templating.

Excellent correlation was found for all data sets, with no significant difference between them in intraobserver reproducibility. Also in terms of interobserver reproducibility, we found good levels of correlation, with no significant difference between the data sets. In terms of accuracy, the correlation was at least acceptable with no significant difference between any of the data sets.

The use of the TraumaCad digital software program provides a reliable method of templating digital images for uncemented THAs. There is little intraobserver and interobserver variability, and the method produces an acceptable level of accuracy.

Preoperative templating is common practice for modern total hip arthroplasty (THA).^1-3 By anticipating intraoperative difficulties, templating has not only been shown to improve the precision of surgery in terms of reconstructing the normal hip center, regaining normal leg length, and reproducing the natural hip offset, it also shortens the length of the procedure and reduces the incidence of complications.^1-6 A further benefit is that a large number of different implants of varying sizes exist in today’s market. It is therefore impossible for any hospital to stock large inventories of implants. Templating allows specific implants and ranges of sizes to be stocked and readily available. Finally, the process of templating identifies cases of unusual anatomy, where appropriate ranges of specialized implants can be made available.⁶

Traditionally, templating was performed with acetate templates set at a standard magnification placed over hard copy analog radiographic films produced in such a fashion as to give the same degree of magnification.^1-6 However, in recent years more medical institutions are using computer-based digital imaging. This change is based on the fact that digital imaging, when compared to analog, has advantages with respect to cost, acquisition, archiving, retrieval, image manipulation, and film rejection rate.^7-10 Only in the past few years have software programs been created that template these digital images. Unfortunately, there is little in the literature that evaluates the essential function of digital orthopedic templating.

Although a few papers have assessed the accuracy of differing templating software programs, to our knowledge none have assessed reproducibility. For digital templating for uncemented primary THA, we wanted to answer 3 questions: (1) is there intraobserver reproducibility, (2) is there interobserver reproducibility, and (3) is the process accurate?

Materials and Methods

Forty-five consecutive patients scheduled to undergo a primary uncemented THA by 1 surgeon (P.L.) were selected for the prospective study. Anteroposterior (AP) images of the pelvis were taken digitally with the Centricity Picture Archiving Communication System (GE Healthcare, Slough, United Kingdom). Both legs were internally rotated as far as possible, and a 1-inch calibration metallic ball was taped to the groin between the legs, at the depth of the palpable greater trochanter. TraumaCad 1.6 templating software, where electronic overlays can be placed over these standard preoperative digital images, was used. The software has a library of vendor-specific templates for both the acetabular and femoral component. All templates in the study were of the Pinnacle acetabular cup and uncollared Corail femoral stem (Depuy Orthopedics, Warsaw, Indiana), the uncemented THA system and intended prosthesis that was to be used in each case. Having adjusted for magnification using the calibration ball, the various components were moved over the area of concern and manipulated by changing their size and orientation until satisfactory positioning was achieved, as described by Bono.¹¹

Templating Method

The acetabular cup template was digitally adjusted to get the best fill of the natural acetabulum without removing too much subchondral bone. The lower border of the cup was placed in line with the lower margin of the teardrop. The cup was also placed medially, so that the medial border reached the ilioischial line. Cups were placed at an inclination of 45° to the horizontal orientation of the pelvic axis. Once the correct orientation and size of the cup were achieved, it automatically provided the center of rotation. The standard Pinnacle system provides cup implants sized from 48 to 60 mm, increasing in 2-mm increments.

Next, an assessment of leg-length discrepancy was made using the software. It was measured as the vertical distance from the inferior point of the teardrop sign to the most inferior point of the lesser trochanter, and a comparison was made to the contralateral side. Any leg-length discrepancy was corrected by cutting the femur at the femoral neck and moving it distally to match the leg lengths on both sides. An appropriately sized Corail stem was placed into the femoral shaft. The Corail stem should be placed in the femoral canal with at least a 1-mm cancellous bone border around it. The standard Corail system provides stem implants sized from 8 to 16, increasing in 1-size increments. Finally, the standard Corail stem system only provides 2 options for the offset: standard or high. The coxa vara version of the Corail stem was not used.

A 1-inch calibration ball is visible in the center of the image field of the Figure. This image shows the placement of a 50-mm acetabular Pinnacle cup with a size 9 collarless standard offset Corail stem. This template leaves a residual leg-length discrepancy of 0.9 mm.

Figure: Example of radiographic templating image.

Data Collection

Two surgeons (P.A., H.H.) familiar with the digital templating software performed the templating for this study. Neither was the lead surgeon during the operation. When surgeon 1 preoperatively templated the patients, he made note of 3 parameters: (1) the external diameter of the acetabular cup component in millimeters, (2) the size of the femoral stem, and (3) whether a standard or high offset was to be used.

At least 1 week later, the same surgeon templated the same patients, making note of the same 3 parameters while blinded to the templating parameters he had provided previously. At a separate sitting, surgeon 2 provided the templating parameters for the same patients while blinded to the templating parameters provided by surgeon 1. Finally, a note was made of the actual sizes of the implant used. Despite the lead surgeon being aware of a set of templating values supplied by surgeon 1, he decided intraoperatively which sizes gave the best fit, soft tissue tension, and leg length.

For intraobserver evaluation, surgeon 1’s templating values were compared to the second templating values he provided 1 week later. For interobserver evaluation, the second set of surgeon 1’s templating values were compared to surgeon 2’s templating values. Finally, for accuracy, the second set of surgeon 1’s templating values were compared to the actual implant used.

Postoperative radiographs were reviewed by the senior author (P.L.) to evaluate whether the prosthetic placement and alignment were acceptable.

Statistical Method

In terms of the acetabular cup and femoral stem size parameters, these are numerical data sets following a normal distribution. Alternatively, the degree of offset was recorded as categorical data.

For both the acetabular cup sizes and the femoral stem sizes, the Pearson correlation coefficient was used to ascertain how well one data set matched another. A 0 result indicated no matching, while a +1 result indicated perfect matching between the samples. We arbitrarily felt that a value >+0.9 indicated excellent matching, a value between +0.8 and +0.89 indicated good matching, and a value between +0.7 and +0.79 indicated acceptable matching. Due to the categorical nature of the offset data, the Kappa index was used to evaluate the degree of agreement between differing sample sets. Again a value of +1 indicated perfect agreement, whereas a 0 result indicated that the agreement between the 2 sample sets was no better than chance. Here we arbitrarily felt that values ranging from 0.81 to 1 indicated excellent matching, from 0.61 to 0.8 indicated good matching, and from 0.41 to 0.6 indicated acceptable matching.

To obtain a P value, paired Student’s t test was used when comparing the acetabular and femoral sizes. Due to the categorical nature of the offset, the McNemar test provided a P value to see if there was any significant difference in the proportions between the 2 sample sets. For the purpose of this study, a P value <.05 was felt to be statistically significant.

Results

Our study revealed that for intraobserver reliability, there was excellent correlation, and therefore reproducibility, for the acetabular cup size, femoral stem size, and femoral offset (Table 1). Furthermore, there was no statistically significant difference between the data sets for all of these parameters.

In terms of interobserver variability, our results showed good correlation/reproducibility for all 3 parameters of the acetabular cup size, femoral stem size, and femoral offset (Table 2). However, the degree of correlation for all 3 parameters was not as strong as the intraobserver variability. Nonetheless, there was no significant difference between the data sets for all 3 parameters.

No significant difference was found between the templating values and the actual implants used for all 3 parameters. However, in terms of the correlation coefficients, a good level of accuracy was found for the acetabular cup size, but only an acceptable level of accuracy was found for the femoral stem size, while the offset produced an excellent level of accuracy (Table 3).

Finally, when the postoperative radiographs were reviewed by the senior author, they were all found to be acceptable in terms of fit and alignment.

Discussion

Oddy et al¹² were some of the first to assess intraobserver reliability for digital images. They placed acetate templates against the screen for digital images adjusted for 120% magnification by a calibration marker. They found correlation coefficients of >0.7. No previous studies have examined intraobserver variability for digital templating software programs. Our results found an excellent level of correlation, with coefficients of >0.9. Such variability is subjective in nature, as it is operator dependent. One would therefore expect minimal variability when the same observer is used. Furthermore, our result of excellent correlation may also reflect the ease of use of the digital method of templating when compared to Oddy et al’s¹² method of acetate templates.

In terms of interobserver reliability, Oddy et al¹² found acceptable levels of reliability with correlation coefficients >0.7. We are not aware of any previous studies that have examined interobserver variability for digital templating software programs. Despite our results showing good interobserver reproducibility with correlation coefficients of >0.8 for acetabular cup and femoral stem sizes and a coefficient >0.61 for the femoral offset, the strength of the reproducibility was not as strong as the intraobserver results. This is to be expected, as for any parameter, less variation exists when the same person makes the observations as opposed to different people. It should be noted that the levels of correlation found for digital templating were stronger than those found by Oddy et al¹² with the acetate templating method.

In terms of accuracy, our results found a good level of accuracy for the acetabular cup size (coefficient of >0.8), but only an acceptable level of accuracy for the femoral stem size (coefficient of >0.7). This can be explained by the assertion that the radiological landmarks, such as the ilioischial line and the teardrop, that are used for cup placement are obvious on radiographs. However, the decision about the amount of cancellous layer left around the femoral stem is subjective. Furthermore, the density of the cancellous bone is variable, and this in turn will dictate the size of the femoral stem it can accommodate during the operation. It is interesting to note that the Kappa index for the offset showed an excellent level of accuracy, which is likely related to the fact that only 2 options exist for the offset.

As digital imaging became established, it soon became apparent that the major difficulty for prosthetic templating was knowing the degree of magnification.^12-14 It is now generally accepted that a marker of known dimension should be placed within the image field to calibrate for the degree of magnification.^12-17 Furthermore, 2 objects should be placed at the same distance from the emission source to experience the same magnification.^16,17 In our study, we taped a 1-inch metallic ball to the inside of the groin at the level of the palpable greater trochanter to act as a calibration device. The goal was to get the calibration device at the same distance from the emission source as the pathological hip. It is impossible to place a calibration device exactly at the same level as the hip, and this leads to variability in the degree of magnification, which in turn influences the accuracy of digital templating. The radiographic image is also influenced by the degree of rotation of the femur,^18,19 which may vary due to differing degrees of arthritis. This again will influence the degree of accuracy of digital templating.

Davila et al²⁰ evaluated the accuracy of EndoMap templating software (Siemans AG; Erlangen, Germany) for uncemented primary THAs. They found that 39% of the acetabular components were estimated exactly and 86% were within 1 size, whereas 19% of the femoral components were estimated exactly and 72% were within 1 size.

A study by The et al²¹ assessed digital templating using HyperORTHO software (Rogan-Delft, Veenendaal, the Netherlands). They found it to be exactly accurate for the uncemented cup size 16% of the time and within 1 size 52% of the time. In terms of the uncemented stem size, digital templating was exactly accurate 34% and within 1 size 66% of the time.

Iorio et al²² assessed digital templating with Sectra Ortho Station software (PC Consultant Group, Inc, Miami, Florida). The digital method had a 60% accuracy of cup sizes to within 1 size, and 74% accuracy of stem sizes to within 1 size.

Our study, using an alternative program, gave better results compared to these studies. This may reflect an improved ease of use or accuracy of the software program, or our operators may be more familiar with manipulating digital images.

Conclusion

This is the first study to evaluate the variability of parameters obtained through digital templating software. We found that intraobserver reproducibility is excellent and interobserver reproducibility is good. Such variability reflects the operator subjectivity when obtaining these templating values, as well as the user-friendly nature and reliability of the digital software program and digital templating technique itself.

We found the accuracy of digital templating to be at least acceptable. Compared to previous studies on digital templating, our results appear to be slightly better in terms of percentage matching. This may be due to surgeons becoming increasingly comfortable in manipulating digital images and using the digital templating software systems, or it may reflect improvements in the software program itself.

References

1. Della Valle AG, Padgett DE, Salvati EA. Preoperative planning for primary total hip arthroplasty. J Am Acad Orthop Surg. 2005; 13(7):455-462.
2. Capello WN. Preoperative planning of total hip arthroplasty. Instr Course Lect. 1986; (35):249-257.
3. Dore DD, Rubash HE. Primary total hip arthroplasty in the older patient: optimizing the results. Instr Course Lect. 1994; (43):347-357.
4. Eggli S, Pisan M, Müller ME. The value of preoperative planning for total hip arthroplasty. J Bone Joint Surg Br. 1998; 80(3):382-390.
5. González Della Valle A, Slullitel G, Piccaluga F, Salvati EA. The precision and usefulness of preoperative planning for cemented and hybrid primary total hip arthroplasty. J Arthroplasty. 2005; 20(1):51-58.
6. Blackley HR, Howell GE, Rorabeck CH. Planning and management of the difficult primary hip replacement: preoperative planning and technical considerations. Instr Course Lect. 2000; (49):3-11.
7. Reiner BI, Siegel EL, Flagle C, Hooper FJ, Cox RE, Scanlon M. Effect of filmless imaging on the utilization of radiologic services. Radiology. 2000; 215(1):163-167.
8. Mattern CW, King BF Jr, Hangiandreou NJ, et al. Electronic imaging impact on image and report turnaround times. J Digit Imaging. 1999; 12(2 Suppl 1):155-159.
9. Strickland NH. PACS (picture archiving and communication systems): filmless radiology. Arch Dis Child. 2000; 83(1):82-86.
10. Lau SL, Mak ASH, Lam WT, Chau CK, Lau KY. Reject analysis: a comparison of conventional film—screen radiography and computed radiography with PACS. Radiography. 2004; 10(3):183-187.
11. Bono JV. Digital templating in total hip arthroplasty. J Bone Joint Surgery Am. 2004; 86(Suppl 2):118-122.
12. Oddy MJ, Jones MJ, Pendegrass CJ, Pilling JR, Wimhurst JA. Assessment of reproducibility and accuracy in templating hybrid total hip arthroplasty using digital radiographs. J Bone Joint Surg Br. 2006; 88(5):581-585.
13. White SP, Shardlow DL. Effect of introduction of digital radiographic techniques on pre-operative templating in orthopaedic practice. Ann R Coll Surg Engl. 2005; 87(1):53-54.
14. Wimsey S, Pickard R, Shaw G. Accurate scaling of digital radiographs of the pelvis. A prospective trial of two methods. J Bone Joint Surg Br. 2006; 88(11):1508-1512.
15. Conn KS, Clarke MT, Hallett JP. A simple guide to determine the magnification of radiographs and to improve the accuracy of preoperative templating. J Bone Joint Surg Br. 2002; 84(2):269-272.
16. Bayne CO, Krosin M, Barber TC. Evaluation of the accuracy and use of x-ray markers in digital templating for total hip arthroplasty. J Arthroplasty. 2009; 24(3):407-413.
17. The B, Diercks RL, Stewart RE, van Ooijen PM, van Horn JR. Digital correction of magnification in pelvic x rays for preoperative planning of hip joint replacements: theoretical development and clinical results of a new protocol. Med Phys. 2005; 32(8):2580-2589.
18. Linclau L, Dokter G, Peene P. Radiological aspects in preoperative planning and postoperative assessment of cementless total hip arthroplasty. Acta Orthop Belg. 1993; 59(2):163-167.
19. Eckrich SG, Noble PC, Tullos HS. Effects of rotation on the radiographic appearance of the femoral canal. J Arthroplasty. 1994; 9(4):419-426.
20. Davila JA, Kransdorf MJ, Duffy GP. Surgical planning of total hip arthroplasty: accuracy of computer-assisted EndoMap software in predicting component size. Skeletal Radiol. 2006; 35(6):390-393.
21. The B, Diercks RL, van Ooijen PM, van Horn JR. Comparison of analog and digital preoperative planning in total hip and knee arthroplasties. A prospective study of 173 hips and 65 total knees. Acta Orthop. 2005; 76(1):78-84.
22. Iorio R, Siegel J, Specht LM, Tilzey JF, Hartman A, Healy WL. 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.

Authors

Messrs Kumar, Kirmani, Humberg, Kavarthapu, and Li are from Kings College Hospital, London, England, United Kingdom.

Messrs Kumar, Kirmani, Humberg, Kavarthapu, and Li have no relevant financial relationships to disclose.

Correspondence should be addressed to: Sayyied J. Kirmani, MRCS, 7 Atkins Rd, Balham, London, United Kingdom, SW12 OAA (ymmij26@hotmail.com).

doi: 10.3928/01477447-20090922-08