- Email: [email protected]
A straightforward method of assessing the accuracy of implantation of knee prostheses
The Knee 8 Ž2001. 139᎐144
A straightforward method of assessing the accuracy of implantation of knee prostheses Benjamin Hewitt, David ShakespeareU Department of Orthopaedic Surgery, Warwick Hospital, Warwickshire, CV34 5BW, UK Received 31 January 2000; received in revised form 30 June 2000; accepted 27 September 2000
Abstract Accurate component placement in knee replacement surgery is important. The precision with which the implants are placed directly affects patient outcome as implant position and alignment influence stability, durability and patellar tracking. The ability to measure the accuracy of implantation of knee replacement components is valuable in assessing not only ones own technique but also in evaluating new instruments or implants and in teaching. The standard AP and lateral radiographs employed by most surgeons give inadequate information to assess alignment of each component accurately. We present a straightforward way of assessing femoral and tibial component alignment by using a series of three radiographs. This technique is reproducible and can be performed using standard equipment in any radiology department. This technique was applied to 160 total knee replacements performed using newly developed instrumentation. It was shown to be simple and the measurements were reproducible, with very little inter observer bias. We believe this technique has a role in audit, teaching, training and assessing new techniques and instruments. 䊚 2001 Elsevier Science B.V. All rights reserved. Keywords: Knee arthroplasty; Accuracy; Alignment
1. Introduction Accurate placement of knee components is important in the durability of the implant. Malalignment leads to polythene overload, bone overload, soft tissue imbalance and patellar maltracking. A number of authors have confirmed the association of varus alignment, tibial lucency and pain w1᎐3,5x. The number of possible malalignments of components is large. Rotational malalignment of the tibia and femur can occur in each of the three axes Ž x, y and z .. These give rise to errors in flexionrextension, valgusrvarus and rotation around the long axis of the U
Corresponding author. Kimberley House, 3 Lillington Avenue, Leamington Spa, Warwickshire, CV32 5UF, UK. Tel.: q44-1926422119; fax: q44-1926-422119. E-mail address: davidshakespeare@uk y consultants.co.uk ŽD. Shakespeare.
bone. In addition there are three translational malalignments leading to errors in medialrlateral, anteriorrposterior and proximalrdistal placement of the components on the bone. There is considerable controversy over the best radiological technique for the assessment of knee replacement. Most surgeons assess their implants with a short AP and lateral film from which accurate measurements cannot be derived. Not only is overall alignment difficult to measure, but individual component position cannot be assessed. Doubt has been cast on the validity of long leg standing films in the early post-operative period as both rotation of the limb and a fixed flexion deformity affect the result. We present a straight forward method of assessing the alignment of both femoral and tibial components in knee replacement, by using three X-rays taken in a normal X-Ray Department without the need for so-
0968-0160r01r$ - see front matter 䊚 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 6 8 - 0 1 6 0 Ž 0 0 . 0 0 0 7 8 - 8
140
B. Hewitt, D. Shakespeare r The Knee 8 (2001) 139᎐144
phisticated equipment. The technique is reproducible and accurate. We believe it has a place in any surgeon’s assessment of their own technique or in teaching and training knee surgeons. It has the advantage of being applicable in the early post-operative period, so that feedback is available quickly allowing surgical technique to be adjusted. It has a particular place in the assessment of new instrumentation for knee replacement. In our centre we have used the technique to assess our accuracy during the development of instruments for the 913 arthroplasty system Ždesigned by the Hospital for Special Surgery, New York; manufactured by Cremascoli, Turin.. It has allowed us to refine both the instruments and our technique in 160 knee replacements, and now forms part of our routine follow up examination of all implants.
Fig. 2. Long lateral view. The posterior slope of the tibial component and the flexion of the femoral component can be assessed. In the 913 arthroscopy the profile of the intercondylar box of the femur subtends an angle of 10⬚ with the true position of the femoral component.
2. Materials and methods 2.1. Subjects The subjects were 137 patients undergoing 160 primary total knee arthroplasties. There were 71 males and 66 females, ranging in age from 45 to 85 years with a mean of 73 years. The primary diagnosis was osteoarthritis or rheumatoid arthritis in all patients. All operations were performed at our institution between July 1998 and April 2000. 2.2. Materials The prosthesis used was the 913 arthroplasty. A posterior stabilised system developed by the Hospital for Special Surgery to improve the function of the Insall Burstein Arthroplasty. The instrumentation was customised specifically for the 913. It included a new femoral cutting jig which enabled all five femoral cuts to be made with a single jig placement, and was based around a intramedullary alignment system which is generally accepted to be the superior method w6,7x. The tibial jig was extra medullary. All saw cuts were made through captive slots using 1.2-mm saw blades ŽMicro Aire large oscillating saw blade 25.4= 90 = 1.2 mm..
Fig. 1. Long lateral position. The heel is propped so that the knee is clear of the table and the X-ray recorded onto a 44-cm cassette.
Our aim was to place the femoral component in 5⬚ of valgus and with no flexion of the femoral component relative to the long axis of the femur. The tibial placement was to be perpendicular to the long axis in both antero-posterior and lateral planes. 2.3. Method of measurement Our method of measuring the alignment of components is based on a series of three radiographs performed 5 days after surgery. These were as follows. 2.3.1. Long lateral of knee This was a true lateral of the knee performed using a standard 44-cm cassette with the film centred on the knee. The ankle was supported with the posterior thigh clear of the X-ray table ŽFig. 1.. We were able to exclude rotated films by the prosthesis silhouette and thus obtain true laterals. On this film the femoral flexion angle and posterior tibial slope were measured. Incidentally an accurate assessment of maximum knee extension can be obtained from the films for comparison with films taken at subsequent outpatient attendance. The posterior tibial slope was defined as the angle between the tibial base plate and the long axis of the tibia. The femoral flexion angle is that angle between the long axis of the femur and the edge of the femoral component silhouette. The long axis was defined as a line joining the femoral isthmus to the point of entry of the intramedullary rod into the canal. In the 913 arthroplasty this corresponds to a point on the prosthesis silhouette at the junction of the anterior third and posterior two thirds, and can usually be recognised by the bone block used to fill the intramedullary drill hole ŽFig. 2..
B. Hewitt, D. Shakespeare r The Knee 8 (2001) 139᎐144
141
Fig. 3. PA of distal femur. The patient lies prone with the femur parallel to the X-ray plate.
2.3.2. PA of distal femur This was performed using a standard 44-cm cassette with the knee at the lower end of the film. The film was taken with the patient prone with the knee slightly flexed to ensure the femur was parallel to the plate ŽFig. 3.. We were able to exclude rotated films by the prosthesis silhouette which on a true PA demonstrated a symmetrical prosthetic notch ŽFigs. 4 and 5.. On this film the femoral alignment was defined as the angle between the long axis of the femur, drawn from the isthmus to the point of entry to the medullary canal, and a line drawn across the distal edge of the femoral condyles ŽFig. 6.. Although most of our patients were elderly and some had bilateral procedures, our radiographers did not encounter significant problems lying the patients prone by the 5th day after surgery.
Fig. 5. Close up of the femoral notch. Poor rotation. The notch is asymmetrical.
tred on the knee and then moving the whole table under the X-ray tube a second film could be taken on the same cassette but now centred on the ankle ŽFig. 7.. The resultant X-ray can then be analysed ŽFig. 8.. Rotated or tilted films were excluded by examination of the tibial component silhouette. On this film tibial alignment was measured and defined as the angle between the tibial base plate and a line drawn from the centre of the cut surface of the tibia and the
2.3.3. Whole tibial film This was performed using a standard 44-cm cassette arranged diagonally. The X-ray table was constrained to move only longitudinally. The knee and ankle were both supported at the same level to ensure the plate was parallel to the tibia. By taking a coned AP cen-
Fig. 4. Close up of the distal femoral notch. Perfect rotation. The notch is symmetrical.
Fig. 6. PA of distal femur. The long axis of the femur is drawn from the isthmus to the point of entry into the femoral canal. The angle between this and the distal edge of the prosthesis is assessed.
142
B. Hewitt, D. Shakespeare r The Knee 8 (2001) 139᎐144
Fig. 7. Whole tibial position. Coned views are taken of the knee and then the ankle on the same cassette. The X-ray table is locked to prevent lateral movement as it is moved from knee to ankle.
centre of the talus. In almost all cases this line passed through the centre of the tibial base plate unless the component was not sitting centrally on the cut surface. From a practical standpoint it is convenient to display the X-ray on the screen as a right knee in all cases. This avoids confusion over varus and valgus position. All measurements were performed by one of the two authors using a standard goniometer and recorded onto a data sheet. 2.4. E¨ aluation of reproducibility To evaluate the reproducibility and quantify inter observer variation we collected nine sequential sets of X-rays and subjected them to measurement by eight orthopaedic medical staff Žtwo seniors and six juniors. who were blinded to previous measurements. In order to investigate variability in X-ray examination for any given individual we subjected 12 patients to repeat X-rays at 12 months after surgery. These films were measured and compared with the readings from the original post-operative films.
doctors, therefore 288 measurements were taken. We present them in table form listing the average range and standard deviation of each of the four measurements performed on each X-ray set ŽTable 1.. These results show that the most variability occurred when measuring the femoral angle where the average range was 2.88⬚ and the least occurred with measuring the tibial angle where the average range was only 2⬚. Overall there was very little variability with most measurements being within 1᎐2⬚ of each other. The results of the X-rays taken at 5 days and 12 months were analysed using paired Student’s t-tests and there was no significant difference between them Ž P) 0.5.. 3.2. Accuracy measurements All four measurements were graphed as histograms ŽFigs. 9 and 10. and summarised in Table 2. Tibial alignment showed a cluster at 90⬚ with 57% being recorded as the desired result. The mean was 90.3% with a standard deviation of 1.1 ŽFig. 9.. Excellent tibial alignment was achieved in 93% of cases ᎏ excellent being defined as target " 2⬚. This correlates well with the published precision of extramedullary tibial guides by Dennis et al. w4x who achieved 88%. In spite of great care we placed more tibial components in varus than valgus.
2.5. Statistical analysis of results 2.5.1. Reproducibility The range and standard deviation of the measurements performed by the eight staff on each X-ray were calculated. These values were then averaged to provide an indication of measurement variability. The measurements of the films taken at 5 days and 12 months were subjected to paired Student’s t-tests. 2.5.2. Accuracy of measurements The average and standard deviation were calculated for each of the four measurements. 3. Results 3.1. Reproducibility All nine sets of X-rays were reviewed by the eight
Fig. 8. Whole tibia view. The film is always displayed as right. The angle is measured between the tibial base plate and a line drawn between the centres of the tibial cut surface and the talus.
B. Hewitt, D. Shakespeare r The Knee 8 (2001) 139᎐144
143
Table 1 Inter-observer variability a Femoral flexion
Avg a
Posterior tib slope
Femoral angle
Tibial angle
Range
S.D.
Range
S.D.
Range
S.D.
Range
S.D.
2.88
1.9
2.55
1.07
3.22
1.29
2
0.8
The average of the ranges and standard deviations of each measurement performed on each X-ray set.
Table 2 Summarised results of the 160 consecutive knee arthroplasties
Average S.D.
Femoral angle
Tibial angle
Calculated valgus angle
Femoral flexion angle
Lateral tibial angle
5.45 1.72
90.27 1.13
5.1 2.1
9.6 1.75
89.9 1.45
Fig. 9. Accuracy of tibial alignment in 160 knee arthroplasties.
Posterior tibial slope showed a very strong cluster at 90⬚. The mean was 89.9⬚ with a standard deviation of 1.45. The angles of flexion for the femoral component were not so strongly clustered at 8⬚, the spread here was larger with the mean being 9.6⬚ and the standard deviation 1.75. The 913 arthroplasty unlike the IB11 has a 10⬚ slope on the lateral profile of the intercondylar box. Hence a measured flexion angle of 10⬚ equates to correct alignment.
Fig. 10. Accuracy of femoral alignment in 160 knee arthroplasties.
Distal femoral angles were strongly clustered at 5⬚ with a mean of 5.45 and a standard deviation of 1.72 ŽFig. 10.. Early in the series we placed some components in excessive valgus. Judging from the signs left on the X-ray by the intramedullary drill, we feel this was probably due to valgus positioning of the intramedullary rod. Soft bone under the lateral femoral condyle may also have contributed to the valgus attitude. Combined tibial and femoral alignment were also calculated ŽFig. 11.. The mean is 5.1⬚ and the standard deviation 2.1⬚. The histogram is shifted somewhat towards varus under the influence of varus positioning of some tibial components.
4. Discussion There is real need to establish satisfactory methods of assessing surgical performance. We believe that the X-ray techniques we have described accurately assess most elements of the positioning of both femoral and tibial components in knee arthroplasty. It was our
Fig. 11. Combined femoral and tibial alignment in 160 knees.
144
B. Hewitt, D. Shakespeare r The Knee 8 (2001) 139᎐144
intention to assess patellar implantation with a skyline X-ray, but this proved unreliable in the early post-operative period. There is no doubt that our radiographers needed practice in setting up the X-rays and initially used screening to set up the whole tibial films. They now produce films of a consistently high standard. Although assessment of varus and valgus malposition is straightforward, rotational alignment of both femur and tibia is very difficult to judge radiologically, except by sophisticated techniques such as CT scanning. We employed a combination of, the transepicondylar axis, and Whitesides line w8x to judge femoral rotation during surgery. Tibial placement was determined relative to the tibial tubercle with a freefloating rod as an additional guide. We know of no way which this can be assessed from straightforward radiographs. A small number of our X-rays revealed slight errors in placement of the components mediorlaterally. As this does not influence the overall alignment of the limb and has only a small influence on load transmission, we have not specifically reported it in our study. We saw no errors in anterorposterior placement of the tibial components. Our results show that the 160 patients scrutinised in this series had the components of their arthroplasty implanted with satisfactory accuracy. Currently in the literature there is no previous reports of what is acceptable, but the majority of these patients had implants within 1᎐2⬚ of what was deemed to be ideal. As a corollary to this we have shown that the 913 knee prosthesis is an implant which can be accurately implanted, the instruments and jigs appear to work well and have validated our choice of an intramedullary femoral system and extra medullary tibial jig. We believe that these straightforward X-rays have a place in assessment of accuracy in knee joint replacement. Each component can be studied independently
and the rapid feedback allows technique to be turned in the light of the findings. It is of particular value in teaching knee surgery and in assessment of new instrumentation. It is also an interesting exercise for the established surgeon to subject his or herself to the rigid discipline of the ruler and the goniometer.
Acknowledgements We would like to thank the radiographers at Warwick Hospital and the Warwickshire Nuffield Hospital without whose skill this study could not have been carried out. References w1x Ewald FC, Jacobs MA, Walker PS, Thomas WH, Scott RD, Sledge CB. Accuracy of total knee replacement: component position and relation to bonercement interface reaction. In: Dorrid, editor. The knee. Baltimore: University Park Press, 1990. w2x Dorrid, Conaty JP, Schreiber R, Mehne DK, Hull D. Technical factors that influence mechanical loosening of total knee arthroplasty. In: Dorrid, editor. The knee. Baltimore: University Park Press, 1985:117᎐120. w3x Kristenson O, Nafei A, Kjaersgaard-Anderson P, Hid I, Jensen J. Long term results of total condylar knee arthroplasty in rheumatoid arthritis. J Bone Joint Surg Br 1992;74:803᎐806. w4x Dennis DA, Channer M, Susman MH, Stringer EA. Intramedullary versus extra medullary alignment systems in total knee arthroplasty. J Arthroplasty 1993;8:43᎐47. w5x Ritter MA, Faris PM, Keating EM, Meding JB. Post operative alignment of total knee replacement. Its effect on survival. Clin Orthop 1994;299:153᎐156. w6x Cates HE, Ritter MA, Keating EM, Faris PM. Intramedullary vs. extra medullary femoral alignment: systems in total knee replacement. Clin Orthop 1993;286:32᎐39. w7x Engh GA, Pedersen TL. Comparative experience with intramedullary and extra medullary alignment in total knee arthroplasty. J Arthroplasty 1990;5:1. w8x Whiteside LA, Arima J, McCarthy DS. Rotational alignment of the femoral component in valgus total knee arthroplasty based upon the anterior᎐posterior axis. Presented at the AAOS Knee Society Meeting, 1995.
A straightforward method of assessing the accuracy of implantation of knee prostheses Benjamin Hewitt, David ShakespeareU Department of Orthopaedic Surgery, Warwick Hospital, Warwickshire, CV34 5BW, UK Received 31 January 2000; received in revised form 30 June 2000; accepted 27 September 2000
Abstract Accurate component placement in knee replacement surgery is important. The precision with which the implants are placed directly affects patient outcome as implant position and alignment influence stability, durability and patellar tracking. The ability to measure the accuracy of implantation of knee replacement components is valuable in assessing not only ones own technique but also in evaluating new instruments or implants and in teaching. The standard AP and lateral radiographs employed by most surgeons give inadequate information to assess alignment of each component accurately. We present a straightforward way of assessing femoral and tibial component alignment by using a series of three radiographs. This technique is reproducible and can be performed using standard equipment in any radiology department. This technique was applied to 160 total knee replacements performed using newly developed instrumentation. It was shown to be simple and the measurements were reproducible, with very little inter observer bias. We believe this technique has a role in audit, teaching, training and assessing new techniques and instruments. 䊚 2001 Elsevier Science B.V. All rights reserved. Keywords: Knee arthroplasty; Accuracy; Alignment
1. Introduction Accurate placement of knee components is important in the durability of the implant. Malalignment leads to polythene overload, bone overload, soft tissue imbalance and patellar maltracking. A number of authors have confirmed the association of varus alignment, tibial lucency and pain w1᎐3,5x. The number of possible malalignments of components is large. Rotational malalignment of the tibia and femur can occur in each of the three axes Ž x, y and z .. These give rise to errors in flexionrextension, valgusrvarus and rotation around the long axis of the U
Corresponding author. Kimberley House, 3 Lillington Avenue, Leamington Spa, Warwickshire, CV32 5UF, UK. Tel.: q44-1926422119; fax: q44-1926-422119. E-mail address: davidshakespeare@uk y consultants.co.uk ŽD. Shakespeare.
bone. In addition there are three translational malalignments leading to errors in medialrlateral, anteriorrposterior and proximalrdistal placement of the components on the bone. There is considerable controversy over the best radiological technique for the assessment of knee replacement. Most surgeons assess their implants with a short AP and lateral film from which accurate measurements cannot be derived. Not only is overall alignment difficult to measure, but individual component position cannot be assessed. Doubt has been cast on the validity of long leg standing films in the early post-operative period as both rotation of the limb and a fixed flexion deformity affect the result. We present a straight forward method of assessing the alignment of both femoral and tibial components in knee replacement, by using three X-rays taken in a normal X-Ray Department without the need for so-
0968-0160r01r$ - see front matter 䊚 2001 Elsevier Science B.V. All rights reserved. PII: S 0 9 6 8 - 0 1 6 0 Ž 0 0 . 0 0 0 7 8 - 8
140
B. Hewitt, D. Shakespeare r The Knee 8 (2001) 139᎐144
phisticated equipment. The technique is reproducible and accurate. We believe it has a place in any surgeon’s assessment of their own technique or in teaching and training knee surgeons. It has the advantage of being applicable in the early post-operative period, so that feedback is available quickly allowing surgical technique to be adjusted. It has a particular place in the assessment of new instrumentation for knee replacement. In our centre we have used the technique to assess our accuracy during the development of instruments for the 913 arthroplasty system Ždesigned by the Hospital for Special Surgery, New York; manufactured by Cremascoli, Turin.. It has allowed us to refine both the instruments and our technique in 160 knee replacements, and now forms part of our routine follow up examination of all implants.
Fig. 2. Long lateral view. The posterior slope of the tibial component and the flexion of the femoral component can be assessed. In the 913 arthroscopy the profile of the intercondylar box of the femur subtends an angle of 10⬚ with the true position of the femoral component.
2. Materials and methods 2.1. Subjects The subjects were 137 patients undergoing 160 primary total knee arthroplasties. There were 71 males and 66 females, ranging in age from 45 to 85 years with a mean of 73 years. The primary diagnosis was osteoarthritis or rheumatoid arthritis in all patients. All operations were performed at our institution between July 1998 and April 2000. 2.2. Materials The prosthesis used was the 913 arthroplasty. A posterior stabilised system developed by the Hospital for Special Surgery to improve the function of the Insall Burstein Arthroplasty. The instrumentation was customised specifically for the 913. It included a new femoral cutting jig which enabled all five femoral cuts to be made with a single jig placement, and was based around a intramedullary alignment system which is generally accepted to be the superior method w6,7x. The tibial jig was extra medullary. All saw cuts were made through captive slots using 1.2-mm saw blades ŽMicro Aire large oscillating saw blade 25.4= 90 = 1.2 mm..
Fig. 1. Long lateral position. The heel is propped so that the knee is clear of the table and the X-ray recorded onto a 44-cm cassette.
Our aim was to place the femoral component in 5⬚ of valgus and with no flexion of the femoral component relative to the long axis of the femur. The tibial placement was to be perpendicular to the long axis in both antero-posterior and lateral planes. 2.3. Method of measurement Our method of measuring the alignment of components is based on a series of three radiographs performed 5 days after surgery. These were as follows. 2.3.1. Long lateral of knee This was a true lateral of the knee performed using a standard 44-cm cassette with the film centred on the knee. The ankle was supported with the posterior thigh clear of the X-ray table ŽFig. 1.. We were able to exclude rotated films by the prosthesis silhouette and thus obtain true laterals. On this film the femoral flexion angle and posterior tibial slope were measured. Incidentally an accurate assessment of maximum knee extension can be obtained from the films for comparison with films taken at subsequent outpatient attendance. The posterior tibial slope was defined as the angle between the tibial base plate and the long axis of the tibia. The femoral flexion angle is that angle between the long axis of the femur and the edge of the femoral component silhouette. The long axis was defined as a line joining the femoral isthmus to the point of entry of the intramedullary rod into the canal. In the 913 arthroplasty this corresponds to a point on the prosthesis silhouette at the junction of the anterior third and posterior two thirds, and can usually be recognised by the bone block used to fill the intramedullary drill hole ŽFig. 2..
B. Hewitt, D. Shakespeare r The Knee 8 (2001) 139᎐144
141
Fig. 3. PA of distal femur. The patient lies prone with the femur parallel to the X-ray plate.
2.3.2. PA of distal femur This was performed using a standard 44-cm cassette with the knee at the lower end of the film. The film was taken with the patient prone with the knee slightly flexed to ensure the femur was parallel to the plate ŽFig. 3.. We were able to exclude rotated films by the prosthesis silhouette which on a true PA demonstrated a symmetrical prosthetic notch ŽFigs. 4 and 5.. On this film the femoral alignment was defined as the angle between the long axis of the femur, drawn from the isthmus to the point of entry to the medullary canal, and a line drawn across the distal edge of the femoral condyles ŽFig. 6.. Although most of our patients were elderly and some had bilateral procedures, our radiographers did not encounter significant problems lying the patients prone by the 5th day after surgery.
Fig. 5. Close up of the femoral notch. Poor rotation. The notch is asymmetrical.
tred on the knee and then moving the whole table under the X-ray tube a second film could be taken on the same cassette but now centred on the ankle ŽFig. 7.. The resultant X-ray can then be analysed ŽFig. 8.. Rotated or tilted films were excluded by examination of the tibial component silhouette. On this film tibial alignment was measured and defined as the angle between the tibial base plate and a line drawn from the centre of the cut surface of the tibia and the
2.3.3. Whole tibial film This was performed using a standard 44-cm cassette arranged diagonally. The X-ray table was constrained to move only longitudinally. The knee and ankle were both supported at the same level to ensure the plate was parallel to the tibia. By taking a coned AP cen-
Fig. 4. Close up of the distal femoral notch. Perfect rotation. The notch is symmetrical.
Fig. 6. PA of distal femur. The long axis of the femur is drawn from the isthmus to the point of entry into the femoral canal. The angle between this and the distal edge of the prosthesis is assessed.
142
B. Hewitt, D. Shakespeare r The Knee 8 (2001) 139᎐144
Fig. 7. Whole tibial position. Coned views are taken of the knee and then the ankle on the same cassette. The X-ray table is locked to prevent lateral movement as it is moved from knee to ankle.
centre of the talus. In almost all cases this line passed through the centre of the tibial base plate unless the component was not sitting centrally on the cut surface. From a practical standpoint it is convenient to display the X-ray on the screen as a right knee in all cases. This avoids confusion over varus and valgus position. All measurements were performed by one of the two authors using a standard goniometer and recorded onto a data sheet. 2.4. E¨ aluation of reproducibility To evaluate the reproducibility and quantify inter observer variation we collected nine sequential sets of X-rays and subjected them to measurement by eight orthopaedic medical staff Žtwo seniors and six juniors. who were blinded to previous measurements. In order to investigate variability in X-ray examination for any given individual we subjected 12 patients to repeat X-rays at 12 months after surgery. These films were measured and compared with the readings from the original post-operative films.
doctors, therefore 288 measurements were taken. We present them in table form listing the average range and standard deviation of each of the four measurements performed on each X-ray set ŽTable 1.. These results show that the most variability occurred when measuring the femoral angle where the average range was 2.88⬚ and the least occurred with measuring the tibial angle where the average range was only 2⬚. Overall there was very little variability with most measurements being within 1᎐2⬚ of each other. The results of the X-rays taken at 5 days and 12 months were analysed using paired Student’s t-tests and there was no significant difference between them Ž P) 0.5.. 3.2. Accuracy measurements All four measurements were graphed as histograms ŽFigs. 9 and 10. and summarised in Table 2. Tibial alignment showed a cluster at 90⬚ with 57% being recorded as the desired result. The mean was 90.3% with a standard deviation of 1.1 ŽFig. 9.. Excellent tibial alignment was achieved in 93% of cases ᎏ excellent being defined as target " 2⬚. This correlates well with the published precision of extramedullary tibial guides by Dennis et al. w4x who achieved 88%. In spite of great care we placed more tibial components in varus than valgus.
2.5. Statistical analysis of results 2.5.1. Reproducibility The range and standard deviation of the measurements performed by the eight staff on each X-ray were calculated. These values were then averaged to provide an indication of measurement variability. The measurements of the films taken at 5 days and 12 months were subjected to paired Student’s t-tests. 2.5.2. Accuracy of measurements The average and standard deviation were calculated for each of the four measurements. 3. Results 3.1. Reproducibility All nine sets of X-rays were reviewed by the eight
Fig. 8. Whole tibia view. The film is always displayed as right. The angle is measured between the tibial base plate and a line drawn between the centres of the tibial cut surface and the talus.
B. Hewitt, D. Shakespeare r The Knee 8 (2001) 139᎐144
143
Table 1 Inter-observer variability a Femoral flexion
Avg a
Posterior tib slope
Femoral angle
Tibial angle
Range
S.D.
Range
S.D.
Range
S.D.
Range
S.D.
2.88
1.9
2.55
1.07
3.22
1.29
2
0.8
The average of the ranges and standard deviations of each measurement performed on each X-ray set.
Table 2 Summarised results of the 160 consecutive knee arthroplasties
Average S.D.
Femoral angle
Tibial angle
Calculated valgus angle
Femoral flexion angle
Lateral tibial angle
5.45 1.72
90.27 1.13
5.1 2.1
9.6 1.75
89.9 1.45
Fig. 9. Accuracy of tibial alignment in 160 knee arthroplasties.
Posterior tibial slope showed a very strong cluster at 90⬚. The mean was 89.9⬚ with a standard deviation of 1.45. The angles of flexion for the femoral component were not so strongly clustered at 8⬚, the spread here was larger with the mean being 9.6⬚ and the standard deviation 1.75. The 913 arthroplasty unlike the IB11 has a 10⬚ slope on the lateral profile of the intercondylar box. Hence a measured flexion angle of 10⬚ equates to correct alignment.
Fig. 10. Accuracy of femoral alignment in 160 knee arthroplasties.
Distal femoral angles were strongly clustered at 5⬚ with a mean of 5.45 and a standard deviation of 1.72 ŽFig. 10.. Early in the series we placed some components in excessive valgus. Judging from the signs left on the X-ray by the intramedullary drill, we feel this was probably due to valgus positioning of the intramedullary rod. Soft bone under the lateral femoral condyle may also have contributed to the valgus attitude. Combined tibial and femoral alignment were also calculated ŽFig. 11.. The mean is 5.1⬚ and the standard deviation 2.1⬚. The histogram is shifted somewhat towards varus under the influence of varus positioning of some tibial components.
4. Discussion There is real need to establish satisfactory methods of assessing surgical performance. We believe that the X-ray techniques we have described accurately assess most elements of the positioning of both femoral and tibial components in knee arthroplasty. It was our
Fig. 11. Combined femoral and tibial alignment in 160 knees.
144
B. Hewitt, D. Shakespeare r The Knee 8 (2001) 139᎐144
intention to assess patellar implantation with a skyline X-ray, but this proved unreliable in the early post-operative period. There is no doubt that our radiographers needed practice in setting up the X-rays and initially used screening to set up the whole tibial films. They now produce films of a consistently high standard. Although assessment of varus and valgus malposition is straightforward, rotational alignment of both femur and tibia is very difficult to judge radiologically, except by sophisticated techniques such as CT scanning. We employed a combination of, the transepicondylar axis, and Whitesides line w8x to judge femoral rotation during surgery. Tibial placement was determined relative to the tibial tubercle with a freefloating rod as an additional guide. We know of no way which this can be assessed from straightforward radiographs. A small number of our X-rays revealed slight errors in placement of the components mediorlaterally. As this does not influence the overall alignment of the limb and has only a small influence on load transmission, we have not specifically reported it in our study. We saw no errors in anterorposterior placement of the tibial components. Our results show that the 160 patients scrutinised in this series had the components of their arthroplasty implanted with satisfactory accuracy. Currently in the literature there is no previous reports of what is acceptable, but the majority of these patients had implants within 1᎐2⬚ of what was deemed to be ideal. As a corollary to this we have shown that the 913 knee prosthesis is an implant which can be accurately implanted, the instruments and jigs appear to work well and have validated our choice of an intramedullary femoral system and extra medullary tibial jig. We believe that these straightforward X-rays have a place in assessment of accuracy in knee joint replacement. Each component can be studied independently
and the rapid feedback allows technique to be turned in the light of the findings. It is of particular value in teaching knee surgery and in assessment of new instrumentation. It is also an interesting exercise for the established surgeon to subject his or herself to the rigid discipline of the ruler and the goniometer.
Acknowledgements We would like to thank the radiographers at Warwick Hospital and the Warwickshire Nuffield Hospital without whose skill this study could not have been carried out. References w1x Ewald FC, Jacobs MA, Walker PS, Thomas WH, Scott RD, Sledge CB. Accuracy of total knee replacement: component position and relation to bonercement interface reaction. In: Dorrid, editor. The knee. Baltimore: University Park Press, 1990. w2x Dorrid, Conaty JP, Schreiber R, Mehne DK, Hull D. Technical factors that influence mechanical loosening of total knee arthroplasty. In: Dorrid, editor. The knee. Baltimore: University Park Press, 1985:117᎐120. w3x Kristenson O, Nafei A, Kjaersgaard-Anderson P, Hid I, Jensen J. Long term results of total condylar knee arthroplasty in rheumatoid arthritis. J Bone Joint Surg Br 1992;74:803᎐806. w4x Dennis DA, Channer M, Susman MH, Stringer EA. Intramedullary versus extra medullary alignment systems in total knee arthroplasty. J Arthroplasty 1993;8:43᎐47. w5x Ritter MA, Faris PM, Keating EM, Meding JB. Post operative alignment of total knee replacement. Its effect on survival. Clin Orthop 1994;299:153᎐156. w6x Cates HE, Ritter MA, Keating EM, Faris PM. Intramedullary vs. extra medullary femoral alignment: systems in total knee replacement. Clin Orthop 1993;286:32᎐39. w7x Engh GA, Pedersen TL. Comparative experience with intramedullary and extra medullary alignment in total knee arthroplasty. J Arthroplasty 1990;5:1. w8x Whiteside LA, Arima J, McCarthy DS. Rotational alignment of the femoral component in valgus total knee arthroplasty based upon the anterior᎐posterior axis. Presented at the AAOS Knee Society Meeting, 1995.