The Journal of Arthroplasty 29 (2014) 2065–2069
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
Patient-Specific Versus Conventional Instrumentation for Total Knee Arthroplasty: Peri-Operative and Cost Differences Alexander M. DeHaan, MD, Jacob R. Adams, MD, Matthew L. DeHart, BS, Thomas W. Huff, MD Department of Orthopaedics and Rehabilitation, Oregon Health & Science University, Portland, Oregon 97239-3098
a r t i c l e
i n f o
Article history: Received 17 February 2014 Accepted 24 June 2014 Keywords: total knee arthroplasty custom cutting blocks patient-specific instrumentation patient-specific total knee arthroplasty cost-effectiveness
a b s t r a c t The role of patient-specific instrumentation in total knee arthroplasty (TKA) is yet to be clearly defined. Current evidence evaluating peri-operative and cost differences against conventional TKA is unclear. We reviewed 356 TKAs between July 2008 and April 2013; 306 TKAs used patient-specific instrumentation while 50 had conventional instrumentation. The patient-specific instrumentation cohort averaged 20.4 min less surgical time (P b 0.01) and had a 42% decrease in operating room turnover time (P = 0.022). At our institution, the money saved through increased operating room efficiency offset the cost of the custom cutting blocks and pre-operative advanced imaging. Routine use of patient-specific TKA can be performed with less surgical time, no increase in peri-operative morbidity, and at no increased cost when compared to conventional TKA. © 2014 Elsevier Inc. All rights reserved.
The demand for total knee arthroplasty (TKA) has increased dramatically over the past twenty years, and is expected to increase by more than 600% by the year 2030 [1]. As a result, there is pressure on surgeons and implant companies to increase operating room efficiency and improve patient outcomes, while lowering the cost to the healthcare system. These ideas led to the development of patient-specific instrumentation in TKA, in which custom alignment guides for the femur and tibia are created from pre-operative three-dimensional magnetic resonance imaging (MRI) or computed tomography (CT) scans. Historically, TKA has used intramedullary or extramedullary femoral and tibial guides for instrumentation. As these guides are not customized to each individual patient, the component size, rotation, position, and orientation are based upon preset valgus angles for the femur and external landmarks that can vary from patient to patient. With patient-specific TKA, the custom instruments are designed to fit to each patient’s unique anatomy, with the precise orientation and component rotation built into the guide. Consequently, it was theorized that patient-specific TKA could decrease surgical time and provide improved component alignment in TKA while also streamlining the number of instrument trays required for a given case. As with computer navigation, patient specific instrumentation does not rely on instrumentation of the intramedullary canal, and as such may result in decreased peri-operative blood loss and a lower risk of fat embolism compared to conventional instrumentation [2,3]. Due to these proposed advantages, there has been a rapid increase in the use and interest in this
new technology for routine TKA; specifically, its global use increased by a factor of 1.5 from 2011 to 2012 [4]. Evidence of these proposed advantages has lagged behind the growth in popularity. Preliminary results of the patient-specific instrumentation TKAs have been inconsistent across the orthopedic literature with regard to operative time [5–16], component alignment [5–7,9–14,17–22], and blood loss [5,7,8,12,14,15]. Some studies report decreased operative time and blood loss with patient-specific instrumentation, while others show the opposite. In addition, some studies have brought concern over the cost of the custom cutting jigs and pre-operative imaging, added expenditures that may not be justified for routine TKA [6,16,23]. However, many of these studies have small patient numbers and may be underpowered for what they are trying to assess. Another concern is whether or not the learning curve involved with the custom cutting guides, as with any new surgical procedure, is taken into account when comparing patientspecific and conventional instrumentation methods. We set out to evaluate whether or not patient-specific instrumentation for TKA leads to decreased peri-operative morbidity when compared to conventional TKA through a large single surgeon case series. Additionally, we evaluated the sizing accuracy of the predicted cutting block templates in the cases utilizing patient-specific instrumentation. Finally, we evaluated the cost of this new technology, and whether any increases in operating room efficiency would justify its routine use. Methods
The Conflict of Interest statement associated with this article can be found at http:// dx.doi.org/10.1016/j.arth.2014.06.019. Reprint requests: Thomas W. Huff, MD, Oregon Health and Science University, Department of Orthopaedics and Rehabilitation, 3181 S.W. Sam Jackson Park Rd, Portland, Oregon 97239-3098. http://dx.doi.org/10.1016/j.arth.2014.06.019 0883-5403/© 2014 Elsevier Inc. All rights reserved.
Study Design Institutional review board approval was obtained prior to beginning this retrospective review. A case list was obtained through
2066
A.M. DeHaan et al. / The Journal of Arthroplasty 29 (2014) 2065–2069
a database search using the CPT code 27447 (primary TKA) between the dates July 2008 and April 2013. Subsequently, patient information was obtained through a review of each patient’s clinical chart and operative record. All procedures were performed at a single institution by the senior author with the assistance of an orthopedic resident, medical student and/or physician assistant. Exclusion criteria consisted of patients who underwent bilateral TKA, unicompartmental knee arthroplasty, TKA with concurrent hardware removal from a previous operation, or revision TKA. Surgical Indications & Arthroplasty Components The indication for TKA was tri-compartmental degenerative joint disease refractory to conservative treatment modalities including weight loss, low-impact aerobic exercise, activity modification, antiinflammatory medications, bracing, and injections. Patient-specific instrumentation was used as the standard for all patients for the majority of the study period. The decision to proceed with conventional instrumentation was based on the patient’s preference against further advanced imaging, logistical issues obtaining the advanced imaging, or if the surgery date the patient desired was to be sooner than the time required for the advanced imaging and development of the custom cutting block. Advanced imaging for the patient-specific instrumentation cases was based on three-dimensional MRI. There were 13 cases which were excluded from the study; these cases utilized a CT scan due to previous knee hardware or a cardiac pacemaker that prevented use of an MRI. From the computerized three-dimensional model, a disposable custom cutting block was created that would fit onto the arthritic knee. The patientspecific instrumentations system used was the Smith and Nephew Visionaire (Memphis, TN, USA) in 306 cases. The TKA components utilized were the Smith and Nephew Legion Primary Knee (Smith and Nephew Inc, Memphis, TN, USA) or the Smith and Nephew Journey Primary Knee (Smith and Nephew Inc, Memphis, TN, USA). Surgical Procedure and Protocol Primary TKA was performed in a standard fashion utilizing a medial parapatellar approach with the use of a tourniquet in all cases. Patient-specific instrumentation utilized pre-fabricated custom cutting blocks, while conventional arthroplasty utilized intramedullary femoral and extramedullary tibial referencing. The components were cemented in position for all cases. A hemovac drain was placed at the end of the surgery and removed on post-operative day 1 for all cases. Procedures were performed in a clean-air laminar-flow environment using body exhaust suits. Pre-operative antibiotics (routinely Cefazolin, unless otherwise indicated) were given within 1 h of incision and continued for 24 h post-operatively. For venous thomboembolic prophylaxis, all patients received mechanical compression devices while in the hospital along with pharmacologic prophylaxis, routinely consisting of Aspirin 325 mg twice daily for 4 weeks unless the patient was deemed high risk or had a history of a previous thromboembolic event. While in the hospital, patients worked daily with a physical therapist to encourage early mobilization. The patients were discharged when they were medically stable, their pain controlled on oral medication, and were ambulatory with the use of an assistive device. Outcome Measures This study set out to answer three specific outcome measures, all of which could be answered through review of the operative data and the patient’s medical record. First, we determined whether patientspecific instrumentation resulted in decreased peri-operative morbidity when compared to conventional TKA. Variables used to assess for differences between the two cohorts included tourniquet
and total operative time, operating room turnover time, estimated intra-operative blood loss, change in post-operative hematocrit and hemoglobin, need for a post-operative blood transfusion, drain output, date of hospital discharge, and intra-operative complications. Second, we evaluated the sizing accuracy of the predicted patientspecific femoral and tibial MRI-based templates. This was done through a retrospective review of the predicted template size for each case, with comparison to the component size used at the time of implantation. Third, we assessed for the potential of any cost savings through use of patient-specific instrumentation. The cost of the pre-operative imaging ranged from $430 to $1360, dependent upon which local imaging center was used and the patient’s insurance type. The tibial and femoral custom cutting blocks was a fixed cost at $500. This represents a negotiated cost at our institution, which is also bundled into the overall implant cost. The cost of operating room time was $129 per minute for the first 30 min and $65 per minute for every minute thereafter, and accounts for the personnel, nursing, equipment, and fixed overhead costs. The cost of the implant tray sterilization was $60 for each tray and includes the cost of the utilities, personnel, equipment, and sterilization process. The difficulty in quantifying the true costs of the above variables comes from the inherent differences between the price that is charged versus collected, what the patient pays versus what the insurance carrier pays, along with the variable costs between institutions and imaging centers. In addition, this cost could be different for each individual based on their deductible and insurance type. Therefore to simplify the cost analysis and make it more generalizable to all patients, we used the cost that was billed to the patient based on their insurance type, with the assumption of full re-imbursement by the payer. Statistical Analysis Statistical analyses were performed using the R language and environment [24]. With the exception of ‘hematocrit change 1’, all selected continuous variables demonstrated a linear relationship and were normally distributed, thus meeting the assumptions necessary for parametric analysis. We utilized classic 2 × 2 chi square design [25] accompanied by the phi Cramer’s V post hoc testing design [26] when analyzing the associations between post-operative blood transfusion and intra-operative complications in patients receiving conventional TKA versus patient-specific TKA. When analyzing the difference in means between levels of tourniquet and total operative time, estimated blood loss, post-operative hematocrit, drain output, length of stay, and cost statistics, we utilized Levene’s test for equality of variance [27] leading to a Student’s t-test design [28]. For the purpose of analyzing ‘hematocrit change 1’ in patients receiving conventional TKA vs patient-specific TKA, a significance statistic was calculated utilizing a Mann–Whitney U testing design due to abnormal distribution [29]. All reported frequencies and significance statistics were calculated utilizing one of these three methods. Results Patient Demographics 356 cases of primary TKAs were performed in 303 patients during the study period. The mean patient age was 62.8 years (standard deviation 10.3 years) with a mean body mass index of 32.2 kg/m2 (standard deviation 6.6 kg/m2). 63.7% of cases were performed in female patients, and 51% involved the right knee. Of the 356 total cases, 306 were performed using patient-specific instrumentation and 50 utilized conventional methods of intramedullary femoral and extramedullary tibial referencing. The two cohorts were similar with regard to patient age, body mass index, and sidedness of surgery (P N 0.05, Table 1); however, there were gender differences between the two
A.M. DeHaan et al. / The Journal of Arthroplasty 29 (2014) 2065–2069
groups, with more females in the patient-specific cohort and more males in the conventional instrumentation cohort (P b 0.01, Table 1).
Table 2 Peri-Operative Results.
Peri-Operative Morbidity Peri-operative factors are shown in Table 2. Cases that utilized patient-specific instrumentation TKA had significantly shorter tourniquet and operative times than conventional TKA (19.9 and 20.4 min, respectively; P b 0.03). The operating room turnover time between cases was 42% shorter for the patient-specific TKA cohort as well (6.4 min; P = 0.022). There were no statistically significant differences in post-operative hematocrit or hemoglobin change, estimated intra-operative blood loss, drain output, or transfusion rate between the two cohorts. There was no difference in length of hospital stay, with two-thirds of patients in both groups leaving the hospital on post-operative day 3. Importantly, no intra-operative complications occurred in either group. Sizing Accuracy of the Patient-Specific Instrumentation Templates The image based sizing template for the femoral and tibial components was available for 272 of the 306 patient-specific instrumentation TKAs (Table 3). Analyzed independently, the predicted component size was correct in 81.3% (221/272 cases) of femoral templates and 90.4% (246/272 cases) of tibial templates. When analyzed by the accuracy of both femoral and tibial components in the same surgical case, the predicted size was correct in 74.3% of cases (202/272 cases). However, this sizing template was incorrect for one of the two components in 22.1% of cases (femur 15.1%, tibia 7.0%) and both components in 3.6% of cases. When a change was required, the femoral component had to be downsized by one 96.1% of the time (49/51 cases), while the tibial component required downsizing by one in 51.7% of cases (15/29 cases) and upsizing by one in 48.3% of cases (14/29 cases). In no case was the predicted cutting block size off by two or more sizes. Cost Analysis Of the variables assessed for patient-specific TKAs, the two that added cost were the price of the pre-operative advanced imaging ($430–$1360) and the cost of the custom cutting blocks ($500). The other variables measured all served to save money and increase efficiency for the patient-specific instrumentation group: shorter operating room time, fewer instrument trays requiring sterilization, and a shorter operating room turnover time. The 20.4 min shorter operating room time for the patient-specific TKA group provided an average cost savings of $1326 (20.4 min × $65/min) per case when compared to conventional
Tourniqet Time (mean, min) Operative Time (mean, min) Room Turnover Time (mean, min) Change in Hematocrit (mean, volume %) Preop to POD1 Preop to POD2 Change in Hemoglobin (mean, g/dL) Preop to POD1 Preop to POD2 Intra-operative EBL (mean, ml) Drain Output (mean, ml) PRBC Transfusion (n cases) Date of Discharge (n, %) POD2 or less POD3 POD4 or more
Patient Specific (306 Cases)
Conventional (50 Cases)
P Value
63.3 ± 16.8 86.8 ± 16.9 15.2 ± 7.8
83.2 ± 22.2 107.2 ± 29.5 21.6 ± 6.5
P = 0.024 P = b0.01 P = 0.022
7.4 ± 2.6 8.4 ± 2.9
7.8 ± 2.9 9 ± 3.2
P = 0.326 P = 0.245
2.6 ± 2.0 2.9 ± 2.1 120.8 ± 68.4 544.58 ± 238 4
2.5 ± 0.9 3 ± 1.0 116.7 ± 60.5 563.2 ± 363.9 1
73 (23.9%) 209 (68.3%) 24 (7.8%)
9 (18%) 34 (68%) 7 (14%)
P P P P P
= 0.909 = 0.738 = 0.699 = 0.743 = 0.461
P = 0.285
POD: post-operative day; g: grams; dL: deciliters; EBL: estimated blood loss; ml: milliliters; PRBC: packed red blood cell; ±includes standard deviation if applicable.
TKA. Similarly, arthroplasty done with patient-specific instrumentation used 4 fewer instrument trays than with conventional instrumentation, and would thus provide an additional cost savings of $240 (4 trays × $60/tray) per case. Unfortunately, the financial impact of a 42% shorter operating room turnover time could not be quantified; however, this increased efficiency would provide additional cost savings to the hospital through money saved in fewer nursing and surgical staff hours, along with the potential of having additional time for more surgical cases. Thus, we found that the routine use of patient-specific instrumentation in TKA added between $830 and $1860 per case based on the cost of the advanced imaging and the prefabricated cutting blocks. However, shorter operating room time and the use of less surgical trays per case saved $1566 per case, along with the additional variable of increased turnover time efficiency. Thus, routine use of patientspecific instrumentation does not appear to add to the overall cost to primary TKA, but can actually result in significant cost savings dependent upon which imaging center is used. Discussion With the increasing demand for TKA over the next twenty years [1], the orthopedic community is in search of a means to provide high quality and efficient patient care at a reasonable cost. Recently, there have been mixed reports on whether patient-specific instrumentation is the means to approach this issue [4–23,30,31]. In the largest retrospective case series to date, Ng et al measured post-operative
Table 3 Accuracy of the Predicted Implant Size.
Table 1 Patient Demographics.
Total n Cases Total n Patients Gender male patients, n (%) female patients, n (%) Mean Age (years) Mean BMI (kg/m2) Side right, n cases (%) left, n cases (%)
2067
Patient Specific
Conventional
306 258
50 45
82 (31.8%) 176 (68.2%) 62.8 ± 10.9 31.9 ± 6.6
28 (62.2%) 17 (37.8%) 62.2 ± 9.6 33.1 ± 6.6
155 (50.1%) 151 (49.9%)
25 (50%) 25 (50%)
P Value
P b 0.01 P = 0.902 P = 0.084 P = 0.932
n: number; BMI: body mass index; kg: kilograms; m: meters; ±includes standard deviation if applicable.
Correct Femur Template Correct Tibia Template Correct Size of Femur AND Tibia Incorrect Size of Femur OR Tibia Incorrect Size of Femur AND Tibia Incorrect Femur Template Required Upsize by 1 Required Downsize by 1 Incorrect Tibia Template Required Upsize by 1 Required Downsize by 1
n
%
221 246 202 60 10 51 2 49 29 14 15
81.3 90.4 74.3 22.1 3.6 3.9 96.1 48.3 51.7
306 total patient specific TKAs, of which 272 had the predicted templated size available for analysis (89% of cases); n: number of cases.
2068
A.M. DeHaan et al. / The Journal of Arthroplasty 29 (2014) 2065–2069
radiographic alignment in 569 TKAs carried out using patient-specific arthroplasty and 155 TKAs carried out using conventional methods, but did not evaluate for peri-operative differences between these two groups [20]. Our series is the second largest to date, that we know of, and is the largest to evaluate sizing accuracy as well as peri-operative and cost differences with regard to these two groups. We found that patient-specific instrumentation in TKA was 20.4 min shorter and resulted in no increase in peri-operative morbidity as opposed to conventional TKA. There was also a 42% reduction in turnover time between cases, which may be a result of fewer trays requiring set-up and removal during room turnover. Through our cost analysis, the money saved in decreased surgical time, decreased operating room turnover time, and the cost of fewer trays requiring sterilization was able to offset the added costs of the pre-operative advanced imaging and custom cutting blocks. In cases in which the cost of the MRI was only $430, patient-specific instrumentation could actually result in a cost savings of $736 per case. A more pertinent cost analysis may focus specifically on the hospital cost, which weighs the implant upcharge of $500 for custom instrumentation against the $1566 savings in tray processing and operating room time reduction. With this analysis, the financial benefit to the hospital is clear. The cost of imaging may actually be further incentive for the hospital if the imaging center used is a part of the hospital system. Previous costs analyses had found that routine use of patientspecific instrumentation was not cost effective [6,16,22]. This discrepancy seems to stem from the differences in the cost of operating room time and implant tray sterilization. Barrack et al reported that each tray cost $30.96 for sterilization, which was half the price of our institution [6]. Similarly, they reported a savings of 11 min in operating time per case, which led to a cost savings of $201.37 per case, or $18.31 per minute [6]. Watters et al reported that although patient-specific instrumentation required 13 min less per case than conventional arthroplasty, this only amounted to $101.01 savings per case with operating time valued at $7.77 per minute [16]. These values are much less than what is reported from our institution, where the charge is $129 per minute for the first 30 min and $65 per minute for each subsequent minute thereafter. Our series reports costs at a regional, academic, tertiary care hospital and level one trauma center, which is likely the reason for relatively high operating cost compared to a smaller community hospital or surgery center. It is well known that hospital costs can vary based on volume, geographic location, and payor contracts; however, one or two orders of magnitude in cost difference are surprising, and may speak to the lack of transparency of the true cost of health care delivery. Accurate numbers are crucial to the cost analysis for a particular institution. The cost of the custom instruments is a negotiated cost and may vary among institutions as well. It may be that patient specific instruments are more cost effective at larger institutions and may be harder to justify in more focused, high volume joint arthroplasty centers where efficiencies have been optimized, implant costs have been negotiated down to a minimum, and a 2 room model is in place which can effectively obviate the turnover time savings. The reported effect of patient-specific instrumentation on surgical time and blood loss is extremely variable in the literature. With regard to surgical time, two studies have found conventional instrumentation to be faster [10,11], several have found no difference between the two groups [9,15], while multiple studies have found that patient-specific instrumentation resulted in shorter surgical times [5–8,12–14,16]. The mean time difference between conventional and patient-specific instrumentation was often within 10 min. However, the total number of patients in each group was often less than 60 [5,9–14], which brings into question whether the studies were appropriately powered to determine if one method differed from the other. Similarly, differences in operative time can be extremely variable based on surgeon training and the number of cases that surgeon has done with the technique, due
to the learning curve associated with any procedure. Thus, if the reported cases in these studies with small patient numbers included the first cases performed with this technique, the results may have been inadvertently skewed towards the technique that the surgeon was more comfortable and familiar with. With respect to blood loss, one may hypothesize that by avoiding intramedullary guides used in conventional instrumentation, there may be less post-operative blood loss. Two studies have shown that patient-specific instrumentation results in less postoperative blood loss [5,7], while others reported no differences between the two groups [8,12,14,15]. In our study, the patientspecific cohort did have a lower change in the post-operative hematocrit levels, but this was not statistically or clinically significant. Similarly, there was no difference in blood transfusion rates between the two groups. Our analysis of the accuracy of the predicted implant size from the pre-operative three-dimensional imaging revealed that the templated size was correct in 81.3% of femoral templates and 90.4% of tibial templates. The predicted implant size was never off by more than one size. For the femur, it nearly always had to be downsized by one (96.1%), while the tibial prediction had to be upsized (48.3%) or downsized (51.7%) nearly equally. This trend and accuracy in implant size, along with the rates of needing to be upsized or downsized by one component size from the templated prediction, are useful for surgeons and implant companies to be aware of. Previous reports of cutting jig accuracy vary greatly from 100% accurate to less than 25% accurate [5,15,30,31]. Bali et al reviewed 32 TKAs carried out using patient-specific instrumentation, and found that size of the femoral component matched the plan size in all cases, while the planned tibial component size had to be downsized in two cases [5]. Howell et al reviewed 48 TKAs carried out using custom cutting guides, and found that the predicted size of both the femoral and tibial components matched the implanted size in all cases [30]. Issa et al reviewed 89 TKA cases and found that the implant size was correctly predicted in 93% of tibial and 95.5% of femoral components [31]. However, Stronach et al reviewed 66 TKA cases performed with the use of patient-specific instrumentation, and found that the preoperative predicted size was correct in only 23% of femurs and 47% of tibias [15]. All studies utilized MRI-based patient-specific instrumentation, except for a minority of cases that required CT imaging due to pacemakers or other issues that prevented use of an MRI. The cause of the large discrepancies in sizing accuracy in the literature is uncertain, but is likely in part the result of subjective intra-operative assessment amongst surgeons. In this series, we found a trend toward increased accuracy as the study period progressed. This is due to a combination of comfort level with the pre-operative plan and improved communication with the assigned engineer from the manufacturer. There are several limitations to our study. The first is its retrospective nature and the bias that comes along with this. This study is a comparative study of patient-specific versus conventional instrumentation, however it is not randomized, and there are nearly five times the number of patients in the patient-specific instrumentation cohort. All studies were done by a fellowship trained arthroplasty surgeon who is beyond the learning curve and readily capable of both instrumentation methods; however, as this is a university hospital, there are residents and medical students who assist in various parts of the case. We were unable to correlate patient outcomes with validated patient reported outcome measures for all patients. We do not have a full set of pre-operative or post-operative long leg radiographs to measure the radiographic accuracy and correction of the two arthroplasty instrumentation techniques for all patients. However, the focus of this study was on peri-operative factors and primarily cost. Safety and efficacy have been previously validated [5,12,18,20,30]. Lastly, we were not able to find the pre-operative templated implant size in 34 of the 306 patientspecific TKAs (11%).
A.M. DeHaan et al. / The Journal of Arthroplasty 29 (2014) 2065–2069
Our study has several strengths. It is the second largest review of patient-specific instrumentation reported in the literature, and the largest to report on peri-operative outcomes, surgical time, and the accuracy of the predicted component sizes. Being that all cases are done by a single surgeon, the learning curve for the new instrumentation method is not an issue to skew the data. Although a formal power analysis was not performed, the large number of cases allows for a reliable estimate of component sizing accuracy. And perhaps most importantly, we were able to demonstrate considerable cost savings for our hospital through the routine use of patient-specific instrumentation for primary TKA. In conclusion, this retrospective comparative study is the largest study to date, as far as we are aware, to measure peri-operative and cost differences between patient-specific and conventional instrumentation in TKA. In our patient cohort, the use of custom cutting guides resulted in accurately sized implants, a significantly shorter surgical time, and no increase in peri-operative morbidity when compared to the conventional instrumentation group. Additionally, patient-specific TKA was cost neutral to cost-effective for the hospital, dependent on the imaging center used, as a result of shorter surgical time, fewer instrument trays requiring sterilization, and the benefit of increased turnover time efficiency. As such, our routine utilization of patient-specific TKA was done at no increased cost to the healthcare system as a whole, and led to a considerable cost savings for the hospital, when compared to conventional TKA. Acknowledgment We would like to acknowledge and thank Marie Kane for her assistance reviewing and editing this manuscript. References 1. Kurtz S, Ong K, Lau E, et al. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am 2007;89A (4):780. 2. Kalairajah Y, Cossey AJ, Verrall GM, et al. Are systemic emboli reduced in computerassisted knee surgery? A prospective, randomised, clinical trial. J Bone Joint Surg Br 2006;88B(2):198. 3. Kalairajah Y, Simpson D, Cossey AJ, et al. Blood loss after total knee replacement: effects of computer-assisted surgery. J Bone Joint Surg Br 2005;87B(11):1480. 4. Thienpont E, Bellemans J, Delport H, et al. Patient-specific instruments: industry's innovation with a surgeon's interest. Knee Surg Sports Traumatol Arthrosc 2013;21 (10):2227. 5. Bali K, Walker P, Bruce W. Custom-fit total knee arthroplasty: our initial experience in 32 knees. J Arthroplasty 2012;27(6):1149. 6. Barrack RL, Ruh EL, Williams BM, et al. Patient specific cutting blocks are currently of no proven value. J Bone Joint Surg Br 2012;94B(11, Supplement A):95. 7. Boonen B, Schotanus MGM, Kerens B, et al. Intra-operative results and radiological outcome of conventional and patient-specific surgery in total knee arthroplasty: a multicentre, randomised controlled trial. Knee Surg Sports Traumatol Arthrosc 2013;21(10):2206.
2069
8. Chareancholvanich K, Narkbunnam R, Pornrattanameneegwong C. A prospective randomised controlled study of patient-specific cutting guides compared with conventional instrumentation in total knee replacement. J Bone Joint Surg Br 2013;95B(3):354. 9. Chen JY, Yeo SJ, Yew AK, et al. The radiological outcomes of patient-specific instrumentation versus conventional total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc 2014;22(3):630. 10. Hamadouche M, Abane L, Boisgard S, et al. RCT multicenter comparison of primary TKA using patient specific versus conventional instrumentation. Paper #26. AAHKS 2012 Annual Meeting 11. Hamilton WG, Parks NL, Saxena A. Patient-specific instrumentation does not shorten surgical time: a prospective, randomised trial. J Arthroplasty 2013;28(8 Supplement):96. 12. Noble Jr JW, Moore CA, Liu N. The value of patient-matched instrumentation in total knee arthroplasty. J Arthroplasty 2012;27(1):153. 13. Nunley RM, Ellison BS, Ruh EL, et al. Are patient-specific cutting blocks costeffective for total knee arthroplasty? Clin Orthop Relat Res 2012;470(3):889. 14. Spencer BA, Mont MA, McGrath MS, et al. Initial experience with custom-fit total knee replacement: intra-operative events and long-leg coronal alignment. Int Orthop 2009;33(6):1571. 15. Stronach BM, Pelt CE, Erickson J, et al. Patient-specific total knee arthroplasty required frequent surgeon-directed changes. Clin Orthop Relat Res 2013;471 (1):169. 16. Watters TS, Mather III RC, Browne JA, et al. Analysis of procedure-related costs and proposed benefits of using patient-specific approach in total knee arthroplasty. J Surg Orthop Adv 2011;20(2):112. 17. Conteduca F, Iorio R, Mazza D, et al. Are MRI-based, patient matched cutting jigs as accurate as the tibial guides? Int Orthop 2012;36(8):1589. 18. Daniilidis KTC, Tibesku CO. A comparison of conventional and patient-specific instruments in total knee arthroplasty. Int Orthop 2014;38(3):503. 19. Klatt BA, Goyal N, Austin MS, et al. Custom-fit total knee arthroplasty (OtisKnee) results in malalignment. J Arthroplasty 2008;23(1):26. 20. Ng VY, DeClaire JH, Berend KR, et al. Improved accuracy of alignment with patientspecific positioning guides compared with manual instrumentation in TKA. Clin Orthop Relat Res 2012;470(1):99. 21. Parratte S, Blanc G, Boussemart T, et al. Rotation in total knee arthroplasty: no difference between patient-specific and conventional instrumentation. Knee Surg Sports Traumatol Arthrosc 2013;21(10):2213. 22. Silva A, Sampaio R, Pinto E. Patient-specific instrumentation improves tibial component rotation in TKA. Knee Surg Sports Traumatol Arthrosc 2014;22(3):636. 23. Slover JD, Rubash HE, Malchau H, et al. Cost-effectiveness analysis of custom total knee cutting blocks. J Arthroplasty 2012;27(2):180. 24. R: a language and environment for statistical computing. Vienna Austria: R Core Team, R Foundation for Statistical Computing; 2013 [http://www.R-project.org]. 25. Pearson Karl. On the criterion that a given system of deviation from the probable in the case of a correlated system of variables is such that it can be reasonably supposed to have arisen from random sampling. Philosophical Magazine Series 5, 50 (302); 1900. p. 157. 26. Cramer H. Mathematical methods of statistics. Princeton: Princeton University Press; 1946. 27. Levene Howard. In: Olkin Ingram, Hotelling Harold, et al, editors. Contributions to probability and statistics: essays in honor of Harold Hotelling. Stanford University Press; 1960. p. 278. 28. Hazewinkel Michiel, editor. Students test. , Encyclopedia of mathematicsSpringer; 2001. 29. Mann Henry B, Whitney Donald R. On a test of whether one of two random variables is stochastically larger than the other. Ann Math Stat 1947;18(1):50. 30. Howell SM, Kuznik K, Hull ML, et al. Results of an initial experience with custom-fit positioning total knee arthroplasty in a series of 48 patients. Orthopaedics 2008;31 (9):857. 31. Issa K, Rifai A, McGrath MS, et al. Reliability of templating with patient-specific instrumentation in total knee arthroplasty. J Knee Surg 2013;26(6):429.
Contents lists available at ScienceDirect
The Journal of Arthroplasty journal homepage: www.arthroplastyjournal.org
Patient-Specific Versus Conventional Instrumentation for Total Knee Arthroplasty: Peri-Operative and Cost Differences Alexander M. DeHaan, MD, Jacob R. Adams, MD, Matthew L. DeHart, BS, Thomas W. Huff, MD Department of Orthopaedics and Rehabilitation, Oregon Health & Science University, Portland, Oregon 97239-3098
a r t i c l e
i n f o
Article history: Received 17 February 2014 Accepted 24 June 2014 Keywords: total knee arthroplasty custom cutting blocks patient-specific instrumentation patient-specific total knee arthroplasty cost-effectiveness
a b s t r a c t The role of patient-specific instrumentation in total knee arthroplasty (TKA) is yet to be clearly defined. Current evidence evaluating peri-operative and cost differences against conventional TKA is unclear. We reviewed 356 TKAs between July 2008 and April 2013; 306 TKAs used patient-specific instrumentation while 50 had conventional instrumentation. The patient-specific instrumentation cohort averaged 20.4 min less surgical time (P b 0.01) and had a 42% decrease in operating room turnover time (P = 0.022). At our institution, the money saved through increased operating room efficiency offset the cost of the custom cutting blocks and pre-operative advanced imaging. Routine use of patient-specific TKA can be performed with less surgical time, no increase in peri-operative morbidity, and at no increased cost when compared to conventional TKA. © 2014 Elsevier Inc. All rights reserved.
The demand for total knee arthroplasty (TKA) has increased dramatically over the past twenty years, and is expected to increase by more than 600% by the year 2030 [1]. As a result, there is pressure on surgeons and implant companies to increase operating room efficiency and improve patient outcomes, while lowering the cost to the healthcare system. These ideas led to the development of patient-specific instrumentation in TKA, in which custom alignment guides for the femur and tibia are created from pre-operative three-dimensional magnetic resonance imaging (MRI) or computed tomography (CT) scans. Historically, TKA has used intramedullary or extramedullary femoral and tibial guides for instrumentation. As these guides are not customized to each individual patient, the component size, rotation, position, and orientation are based upon preset valgus angles for the femur and external landmarks that can vary from patient to patient. With patient-specific TKA, the custom instruments are designed to fit to each patient’s unique anatomy, with the precise orientation and component rotation built into the guide. Consequently, it was theorized that patient-specific TKA could decrease surgical time and provide improved component alignment in TKA while also streamlining the number of instrument trays required for a given case. As with computer navigation, patient specific instrumentation does not rely on instrumentation of the intramedullary canal, and as such may result in decreased peri-operative blood loss and a lower risk of fat embolism compared to conventional instrumentation [2,3]. Due to these proposed advantages, there has been a rapid increase in the use and interest in this
new technology for routine TKA; specifically, its global use increased by a factor of 1.5 from 2011 to 2012 [4]. Evidence of these proposed advantages has lagged behind the growth in popularity. Preliminary results of the patient-specific instrumentation TKAs have been inconsistent across the orthopedic literature with regard to operative time [5–16], component alignment [5–7,9–14,17–22], and blood loss [5,7,8,12,14,15]. Some studies report decreased operative time and blood loss with patient-specific instrumentation, while others show the opposite. In addition, some studies have brought concern over the cost of the custom cutting jigs and pre-operative imaging, added expenditures that may not be justified for routine TKA [6,16,23]. However, many of these studies have small patient numbers and may be underpowered for what they are trying to assess. Another concern is whether or not the learning curve involved with the custom cutting guides, as with any new surgical procedure, is taken into account when comparing patientspecific and conventional instrumentation methods. We set out to evaluate whether or not patient-specific instrumentation for TKA leads to decreased peri-operative morbidity when compared to conventional TKA through a large single surgeon case series. Additionally, we evaluated the sizing accuracy of the predicted cutting block templates in the cases utilizing patient-specific instrumentation. Finally, we evaluated the cost of this new technology, and whether any increases in operating room efficiency would justify its routine use. Methods
The Conflict of Interest statement associated with this article can be found at http:// dx.doi.org/10.1016/j.arth.2014.06.019. Reprint requests: Thomas W. Huff, MD, Oregon Health and Science University, Department of Orthopaedics and Rehabilitation, 3181 S.W. Sam Jackson Park Rd, Portland, Oregon 97239-3098. http://dx.doi.org/10.1016/j.arth.2014.06.019 0883-5403/© 2014 Elsevier Inc. All rights reserved.
Study Design Institutional review board approval was obtained prior to beginning this retrospective review. A case list was obtained through
2066
A.M. DeHaan et al. / The Journal of Arthroplasty 29 (2014) 2065–2069
a database search using the CPT code 27447 (primary TKA) between the dates July 2008 and April 2013. Subsequently, patient information was obtained through a review of each patient’s clinical chart and operative record. All procedures were performed at a single institution by the senior author with the assistance of an orthopedic resident, medical student and/or physician assistant. Exclusion criteria consisted of patients who underwent bilateral TKA, unicompartmental knee arthroplasty, TKA with concurrent hardware removal from a previous operation, or revision TKA. Surgical Indications & Arthroplasty Components The indication for TKA was tri-compartmental degenerative joint disease refractory to conservative treatment modalities including weight loss, low-impact aerobic exercise, activity modification, antiinflammatory medications, bracing, and injections. Patient-specific instrumentation was used as the standard for all patients for the majority of the study period. The decision to proceed with conventional instrumentation was based on the patient’s preference against further advanced imaging, logistical issues obtaining the advanced imaging, or if the surgery date the patient desired was to be sooner than the time required for the advanced imaging and development of the custom cutting block. Advanced imaging for the patient-specific instrumentation cases was based on three-dimensional MRI. There were 13 cases which were excluded from the study; these cases utilized a CT scan due to previous knee hardware or a cardiac pacemaker that prevented use of an MRI. From the computerized three-dimensional model, a disposable custom cutting block was created that would fit onto the arthritic knee. The patientspecific instrumentations system used was the Smith and Nephew Visionaire (Memphis, TN, USA) in 306 cases. The TKA components utilized were the Smith and Nephew Legion Primary Knee (Smith and Nephew Inc, Memphis, TN, USA) or the Smith and Nephew Journey Primary Knee (Smith and Nephew Inc, Memphis, TN, USA). Surgical Procedure and Protocol Primary TKA was performed in a standard fashion utilizing a medial parapatellar approach with the use of a tourniquet in all cases. Patient-specific instrumentation utilized pre-fabricated custom cutting blocks, while conventional arthroplasty utilized intramedullary femoral and extramedullary tibial referencing. The components were cemented in position for all cases. A hemovac drain was placed at the end of the surgery and removed on post-operative day 1 for all cases. Procedures were performed in a clean-air laminar-flow environment using body exhaust suits. Pre-operative antibiotics (routinely Cefazolin, unless otherwise indicated) were given within 1 h of incision and continued for 24 h post-operatively. For venous thomboembolic prophylaxis, all patients received mechanical compression devices while in the hospital along with pharmacologic prophylaxis, routinely consisting of Aspirin 325 mg twice daily for 4 weeks unless the patient was deemed high risk or had a history of a previous thromboembolic event. While in the hospital, patients worked daily with a physical therapist to encourage early mobilization. The patients were discharged when they were medically stable, their pain controlled on oral medication, and were ambulatory with the use of an assistive device. Outcome Measures This study set out to answer three specific outcome measures, all of which could be answered through review of the operative data and the patient’s medical record. First, we determined whether patientspecific instrumentation resulted in decreased peri-operative morbidity when compared to conventional TKA. Variables used to assess for differences between the two cohorts included tourniquet
and total operative time, operating room turnover time, estimated intra-operative blood loss, change in post-operative hematocrit and hemoglobin, need for a post-operative blood transfusion, drain output, date of hospital discharge, and intra-operative complications. Second, we evaluated the sizing accuracy of the predicted patientspecific femoral and tibial MRI-based templates. This was done through a retrospective review of the predicted template size for each case, with comparison to the component size used at the time of implantation. Third, we assessed for the potential of any cost savings through use of patient-specific instrumentation. The cost of the pre-operative imaging ranged from $430 to $1360, dependent upon which local imaging center was used and the patient’s insurance type. The tibial and femoral custom cutting blocks was a fixed cost at $500. This represents a negotiated cost at our institution, which is also bundled into the overall implant cost. The cost of operating room time was $129 per minute for the first 30 min and $65 per minute for every minute thereafter, and accounts for the personnel, nursing, equipment, and fixed overhead costs. The cost of the implant tray sterilization was $60 for each tray and includes the cost of the utilities, personnel, equipment, and sterilization process. The difficulty in quantifying the true costs of the above variables comes from the inherent differences between the price that is charged versus collected, what the patient pays versus what the insurance carrier pays, along with the variable costs between institutions and imaging centers. In addition, this cost could be different for each individual based on their deductible and insurance type. Therefore to simplify the cost analysis and make it more generalizable to all patients, we used the cost that was billed to the patient based on their insurance type, with the assumption of full re-imbursement by the payer. Statistical Analysis Statistical analyses were performed using the R language and environment [24]. With the exception of ‘hematocrit change 1’, all selected continuous variables demonstrated a linear relationship and were normally distributed, thus meeting the assumptions necessary for parametric analysis. We utilized classic 2 × 2 chi square design [25] accompanied by the phi Cramer’s V post hoc testing design [26] when analyzing the associations between post-operative blood transfusion and intra-operative complications in patients receiving conventional TKA versus patient-specific TKA. When analyzing the difference in means between levels of tourniquet and total operative time, estimated blood loss, post-operative hematocrit, drain output, length of stay, and cost statistics, we utilized Levene’s test for equality of variance [27] leading to a Student’s t-test design [28]. For the purpose of analyzing ‘hematocrit change 1’ in patients receiving conventional TKA vs patient-specific TKA, a significance statistic was calculated utilizing a Mann–Whitney U testing design due to abnormal distribution [29]. All reported frequencies and significance statistics were calculated utilizing one of these three methods. Results Patient Demographics 356 cases of primary TKAs were performed in 303 patients during the study period. The mean patient age was 62.8 years (standard deviation 10.3 years) with a mean body mass index of 32.2 kg/m2 (standard deviation 6.6 kg/m2). 63.7% of cases were performed in female patients, and 51% involved the right knee. Of the 356 total cases, 306 were performed using patient-specific instrumentation and 50 utilized conventional methods of intramedullary femoral and extramedullary tibial referencing. The two cohorts were similar with regard to patient age, body mass index, and sidedness of surgery (P N 0.05, Table 1); however, there were gender differences between the two
A.M. DeHaan et al. / The Journal of Arthroplasty 29 (2014) 2065–2069
groups, with more females in the patient-specific cohort and more males in the conventional instrumentation cohort (P b 0.01, Table 1).
Table 2 Peri-Operative Results.
Peri-Operative Morbidity Peri-operative factors are shown in Table 2. Cases that utilized patient-specific instrumentation TKA had significantly shorter tourniquet and operative times than conventional TKA (19.9 and 20.4 min, respectively; P b 0.03). The operating room turnover time between cases was 42% shorter for the patient-specific TKA cohort as well (6.4 min; P = 0.022). There were no statistically significant differences in post-operative hematocrit or hemoglobin change, estimated intra-operative blood loss, drain output, or transfusion rate between the two cohorts. There was no difference in length of hospital stay, with two-thirds of patients in both groups leaving the hospital on post-operative day 3. Importantly, no intra-operative complications occurred in either group. Sizing Accuracy of the Patient-Specific Instrumentation Templates The image based sizing template for the femoral and tibial components was available for 272 of the 306 patient-specific instrumentation TKAs (Table 3). Analyzed independently, the predicted component size was correct in 81.3% (221/272 cases) of femoral templates and 90.4% (246/272 cases) of tibial templates. When analyzed by the accuracy of both femoral and tibial components in the same surgical case, the predicted size was correct in 74.3% of cases (202/272 cases). However, this sizing template was incorrect for one of the two components in 22.1% of cases (femur 15.1%, tibia 7.0%) and both components in 3.6% of cases. When a change was required, the femoral component had to be downsized by one 96.1% of the time (49/51 cases), while the tibial component required downsizing by one in 51.7% of cases (15/29 cases) and upsizing by one in 48.3% of cases (14/29 cases). In no case was the predicted cutting block size off by two or more sizes. Cost Analysis Of the variables assessed for patient-specific TKAs, the two that added cost were the price of the pre-operative advanced imaging ($430–$1360) and the cost of the custom cutting blocks ($500). The other variables measured all served to save money and increase efficiency for the patient-specific instrumentation group: shorter operating room time, fewer instrument trays requiring sterilization, and a shorter operating room turnover time. The 20.4 min shorter operating room time for the patient-specific TKA group provided an average cost savings of $1326 (20.4 min × $65/min) per case when compared to conventional
Tourniqet Time (mean, min) Operative Time (mean, min) Room Turnover Time (mean, min) Change in Hematocrit (mean, volume %) Preop to POD1 Preop to POD2 Change in Hemoglobin (mean, g/dL) Preop to POD1 Preop to POD2 Intra-operative EBL (mean, ml) Drain Output (mean, ml) PRBC Transfusion (n cases) Date of Discharge (n, %) POD2 or less POD3 POD4 or more
Patient Specific (306 Cases)
Conventional (50 Cases)
P Value
63.3 ± 16.8 86.8 ± 16.9 15.2 ± 7.8
83.2 ± 22.2 107.2 ± 29.5 21.6 ± 6.5
P = 0.024 P = b0.01 P = 0.022
7.4 ± 2.6 8.4 ± 2.9
7.8 ± 2.9 9 ± 3.2
P = 0.326 P = 0.245
2.6 ± 2.0 2.9 ± 2.1 120.8 ± 68.4 544.58 ± 238 4
2.5 ± 0.9 3 ± 1.0 116.7 ± 60.5 563.2 ± 363.9 1
73 (23.9%) 209 (68.3%) 24 (7.8%)
9 (18%) 34 (68%) 7 (14%)
P P P P P
= 0.909 = 0.738 = 0.699 = 0.743 = 0.461
P = 0.285
POD: post-operative day; g: grams; dL: deciliters; EBL: estimated blood loss; ml: milliliters; PRBC: packed red blood cell; ±includes standard deviation if applicable.
TKA. Similarly, arthroplasty done with patient-specific instrumentation used 4 fewer instrument trays than with conventional instrumentation, and would thus provide an additional cost savings of $240 (4 trays × $60/tray) per case. Unfortunately, the financial impact of a 42% shorter operating room turnover time could not be quantified; however, this increased efficiency would provide additional cost savings to the hospital through money saved in fewer nursing and surgical staff hours, along with the potential of having additional time for more surgical cases. Thus, we found that the routine use of patient-specific instrumentation in TKA added between $830 and $1860 per case based on the cost of the advanced imaging and the prefabricated cutting blocks. However, shorter operating room time and the use of less surgical trays per case saved $1566 per case, along with the additional variable of increased turnover time efficiency. Thus, routine use of patientspecific instrumentation does not appear to add to the overall cost to primary TKA, but can actually result in significant cost savings dependent upon which imaging center is used. Discussion With the increasing demand for TKA over the next twenty years [1], the orthopedic community is in search of a means to provide high quality and efficient patient care at a reasonable cost. Recently, there have been mixed reports on whether patient-specific instrumentation is the means to approach this issue [4–23,30,31]. In the largest retrospective case series to date, Ng et al measured post-operative
Table 3 Accuracy of the Predicted Implant Size.
Table 1 Patient Demographics.
Total n Cases Total n Patients Gender male patients, n (%) female patients, n (%) Mean Age (years) Mean BMI (kg/m2) Side right, n cases (%) left, n cases (%)
2067
Patient Specific
Conventional
306 258
50 45
82 (31.8%) 176 (68.2%) 62.8 ± 10.9 31.9 ± 6.6
28 (62.2%) 17 (37.8%) 62.2 ± 9.6 33.1 ± 6.6
155 (50.1%) 151 (49.9%)
25 (50%) 25 (50%)
P Value
P b 0.01 P = 0.902 P = 0.084 P = 0.932
n: number; BMI: body mass index; kg: kilograms; m: meters; ±includes standard deviation if applicable.
Correct Femur Template Correct Tibia Template Correct Size of Femur AND Tibia Incorrect Size of Femur OR Tibia Incorrect Size of Femur AND Tibia Incorrect Femur Template Required Upsize by 1 Required Downsize by 1 Incorrect Tibia Template Required Upsize by 1 Required Downsize by 1
n
%
221 246 202 60 10 51 2 49 29 14 15
81.3 90.4 74.3 22.1 3.6 3.9 96.1 48.3 51.7
306 total patient specific TKAs, of which 272 had the predicted templated size available for analysis (89% of cases); n: number of cases.
2068
A.M. DeHaan et al. / The Journal of Arthroplasty 29 (2014) 2065–2069
radiographic alignment in 569 TKAs carried out using patient-specific arthroplasty and 155 TKAs carried out using conventional methods, but did not evaluate for peri-operative differences between these two groups [20]. Our series is the second largest to date, that we know of, and is the largest to evaluate sizing accuracy as well as peri-operative and cost differences with regard to these two groups. We found that patient-specific instrumentation in TKA was 20.4 min shorter and resulted in no increase in peri-operative morbidity as opposed to conventional TKA. There was also a 42% reduction in turnover time between cases, which may be a result of fewer trays requiring set-up and removal during room turnover. Through our cost analysis, the money saved in decreased surgical time, decreased operating room turnover time, and the cost of fewer trays requiring sterilization was able to offset the added costs of the pre-operative advanced imaging and custom cutting blocks. In cases in which the cost of the MRI was only $430, patient-specific instrumentation could actually result in a cost savings of $736 per case. A more pertinent cost analysis may focus specifically on the hospital cost, which weighs the implant upcharge of $500 for custom instrumentation against the $1566 savings in tray processing and operating room time reduction. With this analysis, the financial benefit to the hospital is clear. The cost of imaging may actually be further incentive for the hospital if the imaging center used is a part of the hospital system. Previous costs analyses had found that routine use of patientspecific instrumentation was not cost effective [6,16,22]. This discrepancy seems to stem from the differences in the cost of operating room time and implant tray sterilization. Barrack et al reported that each tray cost $30.96 for sterilization, which was half the price of our institution [6]. Similarly, they reported a savings of 11 min in operating time per case, which led to a cost savings of $201.37 per case, or $18.31 per minute [6]. Watters et al reported that although patient-specific instrumentation required 13 min less per case than conventional arthroplasty, this only amounted to $101.01 savings per case with operating time valued at $7.77 per minute [16]. These values are much less than what is reported from our institution, where the charge is $129 per minute for the first 30 min and $65 per minute for each subsequent minute thereafter. Our series reports costs at a regional, academic, tertiary care hospital and level one trauma center, which is likely the reason for relatively high operating cost compared to a smaller community hospital or surgery center. It is well known that hospital costs can vary based on volume, geographic location, and payor contracts; however, one or two orders of magnitude in cost difference are surprising, and may speak to the lack of transparency of the true cost of health care delivery. Accurate numbers are crucial to the cost analysis for a particular institution. The cost of the custom instruments is a negotiated cost and may vary among institutions as well. It may be that patient specific instruments are more cost effective at larger institutions and may be harder to justify in more focused, high volume joint arthroplasty centers where efficiencies have been optimized, implant costs have been negotiated down to a minimum, and a 2 room model is in place which can effectively obviate the turnover time savings. The reported effect of patient-specific instrumentation on surgical time and blood loss is extremely variable in the literature. With regard to surgical time, two studies have found conventional instrumentation to be faster [10,11], several have found no difference between the two groups [9,15], while multiple studies have found that patient-specific instrumentation resulted in shorter surgical times [5–8,12–14,16]. The mean time difference between conventional and patient-specific instrumentation was often within 10 min. However, the total number of patients in each group was often less than 60 [5,9–14], which brings into question whether the studies were appropriately powered to determine if one method differed from the other. Similarly, differences in operative time can be extremely variable based on surgeon training and the number of cases that surgeon has done with the technique, due
to the learning curve associated with any procedure. Thus, if the reported cases in these studies with small patient numbers included the first cases performed with this technique, the results may have been inadvertently skewed towards the technique that the surgeon was more comfortable and familiar with. With respect to blood loss, one may hypothesize that by avoiding intramedullary guides used in conventional instrumentation, there may be less post-operative blood loss. Two studies have shown that patient-specific instrumentation results in less postoperative blood loss [5,7], while others reported no differences between the two groups [8,12,14,15]. In our study, the patientspecific cohort did have a lower change in the post-operative hematocrit levels, but this was not statistically or clinically significant. Similarly, there was no difference in blood transfusion rates between the two groups. Our analysis of the accuracy of the predicted implant size from the pre-operative three-dimensional imaging revealed that the templated size was correct in 81.3% of femoral templates and 90.4% of tibial templates. The predicted implant size was never off by more than one size. For the femur, it nearly always had to be downsized by one (96.1%), while the tibial prediction had to be upsized (48.3%) or downsized (51.7%) nearly equally. This trend and accuracy in implant size, along with the rates of needing to be upsized or downsized by one component size from the templated prediction, are useful for surgeons and implant companies to be aware of. Previous reports of cutting jig accuracy vary greatly from 100% accurate to less than 25% accurate [5,15,30,31]. Bali et al reviewed 32 TKAs carried out using patient-specific instrumentation, and found that size of the femoral component matched the plan size in all cases, while the planned tibial component size had to be downsized in two cases [5]. Howell et al reviewed 48 TKAs carried out using custom cutting guides, and found that the predicted size of both the femoral and tibial components matched the implanted size in all cases [30]. Issa et al reviewed 89 TKA cases and found that the implant size was correctly predicted in 93% of tibial and 95.5% of femoral components [31]. However, Stronach et al reviewed 66 TKA cases performed with the use of patient-specific instrumentation, and found that the preoperative predicted size was correct in only 23% of femurs and 47% of tibias [15]. All studies utilized MRI-based patient-specific instrumentation, except for a minority of cases that required CT imaging due to pacemakers or other issues that prevented use of an MRI. The cause of the large discrepancies in sizing accuracy in the literature is uncertain, but is likely in part the result of subjective intra-operative assessment amongst surgeons. In this series, we found a trend toward increased accuracy as the study period progressed. This is due to a combination of comfort level with the pre-operative plan and improved communication with the assigned engineer from the manufacturer. There are several limitations to our study. The first is its retrospective nature and the bias that comes along with this. This study is a comparative study of patient-specific versus conventional instrumentation, however it is not randomized, and there are nearly five times the number of patients in the patient-specific instrumentation cohort. All studies were done by a fellowship trained arthroplasty surgeon who is beyond the learning curve and readily capable of both instrumentation methods; however, as this is a university hospital, there are residents and medical students who assist in various parts of the case. We were unable to correlate patient outcomes with validated patient reported outcome measures for all patients. We do not have a full set of pre-operative or post-operative long leg radiographs to measure the radiographic accuracy and correction of the two arthroplasty instrumentation techniques for all patients. However, the focus of this study was on peri-operative factors and primarily cost. Safety and efficacy have been previously validated [5,12,18,20,30]. Lastly, we were not able to find the pre-operative templated implant size in 34 of the 306 patientspecific TKAs (11%).
A.M. DeHaan et al. / The Journal of Arthroplasty 29 (2014) 2065–2069
Our study has several strengths. It is the second largest review of patient-specific instrumentation reported in the literature, and the largest to report on peri-operative outcomes, surgical time, and the accuracy of the predicted component sizes. Being that all cases are done by a single surgeon, the learning curve for the new instrumentation method is not an issue to skew the data. Although a formal power analysis was not performed, the large number of cases allows for a reliable estimate of component sizing accuracy. And perhaps most importantly, we were able to demonstrate considerable cost savings for our hospital through the routine use of patient-specific instrumentation for primary TKA. In conclusion, this retrospective comparative study is the largest study to date, as far as we are aware, to measure peri-operative and cost differences between patient-specific and conventional instrumentation in TKA. In our patient cohort, the use of custom cutting guides resulted in accurately sized implants, a significantly shorter surgical time, and no increase in peri-operative morbidity when compared to the conventional instrumentation group. Additionally, patient-specific TKA was cost neutral to cost-effective for the hospital, dependent on the imaging center used, as a result of shorter surgical time, fewer instrument trays requiring sterilization, and the benefit of increased turnover time efficiency. As such, our routine utilization of patient-specific TKA was done at no increased cost to the healthcare system as a whole, and led to a considerable cost savings for the hospital, when compared to conventional TKA. Acknowledgment We would like to acknowledge and thank Marie Kane for her assistance reviewing and editing this manuscript. References 1. Kurtz S, Ong K, Lau E, et al. Projections of primary and revision hip and knee arthroplasty in the United States from 2005 to 2030. J Bone Joint Surg Am 2007;89A (4):780. 2. Kalairajah Y, Cossey AJ, Verrall GM, et al. Are systemic emboli reduced in computerassisted knee surgery? A prospective, randomised, clinical trial. J Bone Joint Surg Br 2006;88B(2):198. 3. Kalairajah Y, Simpson D, Cossey AJ, et al. Blood loss after total knee replacement: effects of computer-assisted surgery. J Bone Joint Surg Br 2005;87B(11):1480. 4. Thienpont E, Bellemans J, Delport H, et al. Patient-specific instruments: industry's innovation with a surgeon's interest. Knee Surg Sports Traumatol Arthrosc 2013;21 (10):2227. 5. Bali K, Walker P, Bruce W. Custom-fit total knee arthroplasty: our initial experience in 32 knees. J Arthroplasty 2012;27(6):1149. 6. Barrack RL, Ruh EL, Williams BM, et al. Patient specific cutting blocks are currently of no proven value. J Bone Joint Surg Br 2012;94B(11, Supplement A):95. 7. Boonen B, Schotanus MGM, Kerens B, et al. Intra-operative results and radiological outcome of conventional and patient-specific surgery in total knee arthroplasty: a multicentre, randomised controlled trial. Knee Surg Sports Traumatol Arthrosc 2013;21(10):2206.
2069
8. Chareancholvanich K, Narkbunnam R, Pornrattanameneegwong C. A prospective randomised controlled study of patient-specific cutting guides compared with conventional instrumentation in total knee replacement. J Bone Joint Surg Br 2013;95B(3):354. 9. Chen JY, Yeo SJ, Yew AK, et al. The radiological outcomes of patient-specific instrumentation versus conventional total knee arthroplasty. Knee Surg Sports Traumatol Arthrosc 2014;22(3):630. 10. Hamadouche M, Abane L, Boisgard S, et al. RCT multicenter comparison of primary TKA using patient specific versus conventional instrumentation. Paper #26. AAHKS 2012 Annual Meeting 11. Hamilton WG, Parks NL, Saxena A. Patient-specific instrumentation does not shorten surgical time: a prospective, randomised trial. J Arthroplasty 2013;28(8 Supplement):96. 12. Noble Jr JW, Moore CA, Liu N. The value of patient-matched instrumentation in total knee arthroplasty. J Arthroplasty 2012;27(1):153. 13. Nunley RM, Ellison BS, Ruh EL, et al. Are patient-specific cutting blocks costeffective for total knee arthroplasty? Clin Orthop Relat Res 2012;470(3):889. 14. Spencer BA, Mont MA, McGrath MS, et al. Initial experience with custom-fit total knee replacement: intra-operative events and long-leg coronal alignment. Int Orthop 2009;33(6):1571. 15. Stronach BM, Pelt CE, Erickson J, et al. Patient-specific total knee arthroplasty required frequent surgeon-directed changes. Clin Orthop Relat Res 2013;471 (1):169. 16. Watters TS, Mather III RC, Browne JA, et al. Analysis of procedure-related costs and proposed benefits of using patient-specific approach in total knee arthroplasty. J Surg Orthop Adv 2011;20(2):112. 17. Conteduca F, Iorio R, Mazza D, et al. Are MRI-based, patient matched cutting jigs as accurate as the tibial guides? Int Orthop 2012;36(8):1589. 18. Daniilidis KTC, Tibesku CO. A comparison of conventional and patient-specific instruments in total knee arthroplasty. Int Orthop 2014;38(3):503. 19. Klatt BA, Goyal N, Austin MS, et al. Custom-fit total knee arthroplasty (OtisKnee) results in malalignment. J Arthroplasty 2008;23(1):26. 20. Ng VY, DeClaire JH, Berend KR, et al. Improved accuracy of alignment with patientspecific positioning guides compared with manual instrumentation in TKA. Clin Orthop Relat Res 2012;470(1):99. 21. Parratte S, Blanc G, Boussemart T, et al. Rotation in total knee arthroplasty: no difference between patient-specific and conventional instrumentation. Knee Surg Sports Traumatol Arthrosc 2013;21(10):2213. 22. Silva A, Sampaio R, Pinto E. Patient-specific instrumentation improves tibial component rotation in TKA. Knee Surg Sports Traumatol Arthrosc 2014;22(3):636. 23. Slover JD, Rubash HE, Malchau H, et al. Cost-effectiveness analysis of custom total knee cutting blocks. J Arthroplasty 2012;27(2):180. 24. R: a language and environment for statistical computing. Vienna Austria: R Core Team, R Foundation for Statistical Computing; 2013 [http://www.R-project.org]. 25. Pearson Karl. On the criterion that a given system of deviation from the probable in the case of a correlated system of variables is such that it can be reasonably supposed to have arisen from random sampling. Philosophical Magazine Series 5, 50 (302); 1900. p. 157. 26. Cramer H. Mathematical methods of statistics. Princeton: Princeton University Press; 1946. 27. Levene Howard. In: Olkin Ingram, Hotelling Harold, et al, editors. Contributions to probability and statistics: essays in honor of Harold Hotelling. Stanford University Press; 1960. p. 278. 28. Hazewinkel Michiel, editor. Students test. , Encyclopedia of mathematicsSpringer; 2001. 29. Mann Henry B, Whitney Donald R. On a test of whether one of two random variables is stochastically larger than the other. Ann Math Stat 1947;18(1):50. 30. Howell SM, Kuznik K, Hull ML, et al. Results of an initial experience with custom-fit positioning total knee arthroplasty in a series of 48 patients. Orthopaedics 2008;31 (9):857. 31. Issa K, Rifai A, McGrath MS, et al. Reliability of templating with patient-specific instrumentation in total knee arthroplasty. J Knee Surg 2013;26(6):429.
