Advertisement

Can a Reduction in Revision Rates Make Robotic Total Knee Arthroplasty Cost Neutral With Manual Total Knee Arthroplasty at Ten-Year Follow-Up? An Episode Cost Analysis

Published:November 19, 2021DOI:https://doi.org/10.1016/j.arth.2021.10.030

      Abstract

      Background

      Advocates of robotic total knee arthroplasty (RTKA) suggest that its greater cost may be recaptured through a reduction in revision rates. We sought to determine what reduction in revision TKA would be required for RTKA to become cost neutral with manual TKA (MTKA).

      Methods

      Episode costs were determined for 2392 RTKAs and 2392 MTKAs. Mean total cost of revision TKA in our health system was identified. Episode cost difference of the RTKA and MTKA cohorts was divided by the mean cost of revision TKA to estimate the reduction in revisions required to make RTKA cost neutral with MTKA. The National Joint Registry was consulted to determine the cumulative revision rate for the implant used in this study.

      Results

      Episode cost for the RTKA cohort was $5.7M greater than MTKA. Mean acute stay cost for revision TKA was $20,972, but post-acute costs were not available. If post-acute costs for revision TKA are conservatively estimated at 50% of episode cost (ie, episode cost = $41,944), 131 revision TKAs would need to be prevented in the RTKA cohort to make it cost neutral with MTKA. The National Joint Registry cumulative revision rate for this implant is 3.37% at 10 years, thus only 81 revisions would be expected per cohort.

      Conclusion

      Our data suggest that it is not possible for RTKA to achieve cost parity with MTKA through a reduction in revision rate alone. Future price reductions may make the cost comparison more favorable. In addition, demonstration of improved patient outcomes would undeniably add value to RTKA and change the analysis.

      Keywords

      Robotics were introduced to TKA (RTKA) intending to make the procedure more accurate and reproducible than manual TKA (MTKA). Studies have demonstrated the accuracy of bone cuts and alignment with RTKA [
      • Sires J.D.
      • Craik J.D.
      • Wilson C.J.
      Accuracy of bone resection in MAKO total knee robotic-assisted surgery.
      ,
      • Parratte S.
      • Price A.J.
      • Jeys L.M.
      • Jackson W.F.
      • Clarke H.D.
      Accuracy of a new robotically assisted technique for total knee arthroplasty: a cadaveric study.
      ], which have been better than MTKA in some cases [
      • Hampp E.L.
      • Chughai M.
      • Scholl L.Y.
      • Sodhi N.
      • Bhowmik-Stoka M.
      • Jacofsky D.J.
      Robotic-arm assisted total knee arthroplasty demonstrated greater accuracy and precision to plan compared with manual techniques.
      ,
      • Song E.K.
      • Seon J.K.
      • Yim J.H.
      • Netravali N.A.
      • Bargar W.L.
      Robotic-assisted TKA reduces postoperative alignment outliers and improves gap balance compared to conventional TKA.
      ]. Other studies have suggested that RTKA may improve ligament balance [
      • Manning W.
      • Ghosh M.
      • Wilson I.
      • Hide G.
      • Longstaff L.
      • Deehan D.
      Improved mediolateral load distribution without adverse laxity pattern in robot-assisted knee arthroplasty compared to a standard manual measured resection technique.
      ,
      • Shalhoub S.
      • Lawrence J.M.
      • Keggi J.M.
      • Randall A.L.
      • DeClaire J.H.
      • Plaskos C.
      • et al.
      Robotic-assisted total knee arthroplasty combined with a robotic tensioning system can help predict and achieve accurate postoperative ligament balance.
      ] and result in less bone or soft tissue trauma than MTKA [
      • Kayani B.
      • Konan S.
      • Pietrzak J.R.T.
      • Haddad F.
      Iatrogenic bone and soft tissue trauma in robotic-arm assisted total knee arthroplasty compared with conventional jig-based total knee arthroplasty: a prospective cohort study and validation of a new classification system.
      ,
      • Hampp E.L.
      • Sodhi N.
      • Scholl L.
      • Deren M.E.
      • Yenna Z.
      • Westrich G.
      • et al.
      Less iatrogenic soft-tissue damage utilizing robotic-assisted total knee arthroplasty when compared with a manual approach.
      ].
      Theoretical advantages notwithstanding, the addition of robotics increases the cost of TKA. Advocates of RTKA believe that the added cost will be offset in the short term by improved operating room efficiency [
      • Parsley B.S.
      Robotics in orthopedics: a brave new world.
      ,
      • Kayani B.
      • Konan S.
      • Tahmassebi J.
      • Pietrzak J.R.T.
      • Haddad F.S.
      Robotic-arm assisted total knee arthroplasty is associated with improved early functional recovery and reduced time to hospital discharge compared with conventional jig-based total knee arthroplasty.
      ,
      • Jacofsky D.J.
      • Allen M.
      Robotics in arthroplasty: a comprehensive review.
      ]. They also believe that reduced soft tissue trauma will decrease pain and expedite recovery [
      • Hampp E.L.
      • Sodhi N.
      • Scholl L.
      • Deren M.E.
      • Yenna Z.
      • Westrich G.
      • et al.
      Less iatrogenic soft-tissue damage utilizing robotic-assisted total knee arthroplasty when compared with a manual approach.
      ], shortening length of stay (LOS) [
      • Jacofsky D.J.
      • Allen M.
      Robotics in arthroplasty: a comprehensive review.
      ] and reducing complications [
      • Kayani B.
      • Konan S.
      • Pietrzak J.R.T.
      • Haddad F.
      Iatrogenic bone and soft tissue trauma in robotic-arm assisted total knee arthroplasty compared with conventional jig-based total knee arthroplasty: a prospective cohort study and validation of a new classification system.
      ,
      • Hampp E.L.
      • Sodhi N.
      • Scholl L.
      • Deren M.E.
      • Yenna Z.
      • Westrich G.
      • et al.
      Less iatrogenic soft-tissue damage utilizing robotic-assisted total knee arthroplasty when compared with a manual approach.
      ]. Finally, if improved accuracy and balance result in lower revision rates, the added costs of RTKA may also be recouped in the long term.
      We recently studied the costs and quality measures of RTKA and MTKA performed within our health system. Our data showed RTKA to be the longer and costlier procedure, without clinically significant differences in LOS or complications. Full data from the original study are shown in Appendix 1.
      Short-term benefits of RTKA have been difficult to quantify, and we observed few early benefits of RTKA to justify the additional cost. However, a reduction in revision rate would also be an objective measure of value for RTKA. The goal of this study is to estimate the reduction in revision rate necessary to offset the additional cost of RTKA and make it cost neutral with MTKA.

      Materials and Methods

      All RTKAs and MTKAs performed between January 1, 2017 and December 31, 2019 by 6 high-volume (minimum 75 TKA/y) surgeons in each cohort were identified (total 12 surgeons). All RTKAs were performed utilizing the Stryker Mako surgical robot with Stryker implants, and all MTKAs were performed with Stryker implants. The first 20 cases of each surgeon were eliminated to allow for the “learning curve” of RTKA and control for MTKA. The cohorts were one-to-one propensity score matched to eliminate demographic differences in age, body mass index, and American Society of Anesthesiologists score. After propensity score matching, 2392 RTKAs and 2392 MTKAs were studied.
      Direct cost data were obtained from the Providence Health system internal costing database for each inpatient stay and for all readmissions. Direct cost includes all costs associated with the acute stay: operating room costs, implant costs, costs associated with the surgical robot (purchase, maintenance, and per case disposables), pharmacy, nursing, and physical/occupational therapy care, where relevant.
      Because the costs of Home Healthcare (HHC) and Skilled Nursing Facility (SNF) are accrued outside of the health system, and this data was not easily obtainable for each patient, the costs of HHC and SNF episodes were estimated. The mean HHC and SNF charges reported by the Centers for Medicare and Medicaid Services for primary total knee replacement patients from Providence Health hospitals participating in bundled payment programs (Bundled Payment for Care Improvement-Advanced and Comprehensive Care for Joint Replacement) during the study period were used. These costs, multiplied by the number in each cohort requiring HHC and SNF, determined the cost of each.

      Definitions

      • “Episode Cost” = Acute stay cost + HHC cost (if applicable) + readmission cost (if applicable)
      • “Episode Cost Δ” = [Episode Cost of the RTKA cohort] − [Episode Cost for the MTKA cohort
      • Episode cost ΔMean Cost of Revision TKA=
        Number of revision TKA to be prvented for cost parity
      The mean acute stay cost for (all-cause) revision total knee for the Providence Health System during the study period was obtained from the system accounting office. Episode costs for revision TKA for the study period were not available.
      The National Joint Registry (NJR) 17th Annual Report (2020) was consulted to obtain the cumulative revision rate (first revision, any cause) for Stryker Triathlon total knees.
      The calculated number of revisions to be prevented for cost neutrality was then compared to the expected number of revisions per cohort from NJR data.

      Results

      The direct cost of the acute stays for the RTKA cohort ($27,964,333) was approximately $6.6M greater than for MTKA ($21,335,458) (Table 1).
      Table 1Acute Stay Costs (RTKA and MTKA).
      RTKA (N = 2392)MTKA (n = 2392)
      Direct acute stay cost$27,964,333$21,335,458
      RTKA, robotic total knee arthroplasty; MTKA, manual total knee arthroplasty.
      Total estimated costs of post-acute care (HHC + SNF) for RTKA ($1,787,952) were approximately $371,000 greater than MTKA ($1,416,432) (Table 2). Readmission costs for the RTKA cohort ($216,679) were approximately $1.3M less than MTKA (1,512,276) (Table 2). Post-acute costs represented 6.7% of episode cost for RTKA and 12.1% of episode costs for MTKA (Table 3).
      Table 2Post-Acute Costs.
      Home Healthcare Cost
      Mean cost$1728/patient
      MTKA (n = 694)$1,199,232
      RTKA (n = 909)$1,570,752
      Skilled Nursing Facility Cost
      Mean cost$1810/patient
      MTKA (n = 120)$217,200
      RTKA (n = 120)$217,200
      Readmission Cost
      MTKA (n = 117)$1,512,276
      RTKA (n = 29)$216,679
      RTKA, robotic total knee arthroplasty; MTKA, manual total knee arthroplasty.
      Table 3Post-Acute Care as Portion of Episode Cost.
      RTKA6.70%
      MTKA12.10%
      RTKA, robotic total knee arthroplasty; MTKA, manual total knee arthroplasty.
      Episode Costs are shown in Table 4. Episode Cost Δ was $5,704,798 greater for RTKA.
      Table 4Episode Cost.
      Episode Costs Totals and Cost ΔRTKA (N = 2392)MTKA (n = 2392)
      Direct inpatient cost$27,964,333$21,335,458
      HHC$1,570,752$1,199,232
      SNF$217,200$217,200
      Readmissions$216,679$1,512,276
      Total$29,968,964$24,264,166
      Episode Δ$5,704,798
      RTKA, robotic total knee arthroplasty; MTKA, manual total knee arthroplasty.
      The mean acute stay cost of all-cause revision TKA for the Providence Health system during the study period was $20,972.
      The calculated number of revision TKA to be prevented with RTKA to achieve cost neutrality with MTKA was
      $5,704,798$20,972/Rev TKA=272


      The NJR reports the cumulative revision rate for this implant to be 3.37% at 10 years (95% confidence interval [CI] 3.18-3.56), thus
      2392RTKA×.0337=80.6


      Eighty-one revisions would be expected by 10 years (95% CI 76-85) per cohort [
      ] (Fig. 1).
      Figure thumbnail gr1
      Fig. 1Number of revisions required for cost neutrality between MTKA and RTKA compared with number of revisions expected per cohort (NJR).

      Discussion

      With a greater episode cost of $5.7M for the RTKA cohort and a mean cost for all-cause revisions of $20,972, 272 revisions would need to be prevented to offset the greater cost of RTKA. However, the NJR data show that only 76-85 revisions are expected at 10 years (95% CI). Thus, our study predicts that the additional costs of RTKA cannot be recouped by a reduction in revisions alone (Fig. 1). However, for the calculation we used our health system’s mean cost for the revision TKA acute stay only, not the episode cost. We do not have comprehensive episode cost data for revision TKA during the study period.
      For a more meaningful comparison, we examined the portion of episode cost that post-acute expenditures represented in the 2 cohorts. Post-discharge costs (HHC, SNF, and readmissions) represented 6.7% and 12.1% of the RTKA and MTKA episode cost respectively (Table 3). However, revision TKA is a more complex procedure than primary TKA and would be expected to have greater post-discharge costs. If we were to conservatively estimate those costs as equal to the index admission costs ($20,972, or 50% of episode cost), then episode cost for revision TKA would become $41,944. In this model, 136 revisions would need be prevented to make RTKA cost neutral, still 55 more than predicted from the NJR data (Fig. 2).
      Figure thumbnail gr2
      Fig. 2The green line shows that 136 revisions would need to be prevented by RTKA for cost neutrality if post-acute care for RevTKA is estimated at 50% of episode cost.
      Robotic surgery enthusiasts cite numerous potential benefits of RTKA over MTKA. In the early postoperative period, decreased soft tissue trauma may result in lower analgesic requirements, including opioids, and shorten LOS [
      • Kayani B.
      • Konan S.
      • Tahmassebi J.
      • Pietrzak J.R.T.
      • Haddad F.S.
      Robotic-arm assisted total knee arthroplasty is associated with improved early functional recovery and reduced time to hospital discharge compared with conventional jig-based total knee arthroplasty.
      ]. In addition, the contained stereotactic boundary of the robotic arm may limit soft tissue and bony injury and reduce complications [
      • Kayani B.
      • Konan S.
      • Pietrzak J.R.T.
      • Haddad F.
      Iatrogenic bone and soft tissue trauma in robotic-arm assisted total knee arthroplasty compared with conventional jig-based total knee arthroplasty: a prospective cohort study and validation of a new classification system.
      ,
      • Hampp E.L.
      • Sodhi N.
      • Scholl L.
      • Deren M.E.
      • Yenna Z.
      • Westrich G.
      • et al.
      Less iatrogenic soft-tissue damage utilizing robotic-assisted total knee arthroplasty when compared with a manual approach.
      ]. Finally, RTKA has been demonstrated to have high accuracy and reproducibility [
      • Hampp E.L.
      • Chughai M.
      • Scholl L.Y.
      • Sodhi N.
      • Bhowmik-Stoka M.
      • Jacofsky D.J.
      Robotic-arm assisted total knee arthroplasty demonstrated greater accuracy and precision to plan compared with manual techniques.
      ,
      • Song E.K.
      • Seon J.K.
      • Yim J.H.
      • Netravali N.A.
      • Bargar W.L.
      Robotic-assisted TKA reduces postoperative alignment outliers and improves gap balance compared to conventional TKA.
      ,
      • Manning W.
      • Ghosh M.
      • Wilson I.
      • Hide G.
      • Longstaff L.
      • Deehan D.
      Improved mediolateral load distribution without adverse laxity pattern in robot-assisted knee arthroplasty compared to a standard manual measured resection technique.
      ,
      • Shalhoub S.
      • Lawrence J.M.
      • Keggi J.M.
      • Randall A.L.
      • DeClaire J.H.
      • Plaskos C.
      • et al.
      Robotic-assisted total knee arthroplasty combined with a robotic tensioning system can help predict and achieve accurate postoperative ligament balance.
      ], which advocates predict will translate into a lower rate of revision than MTKA [
      • Jacofsky D.J.
      • Allen M.
      Robotics in arthroplasty: a comprehensive review.
      ]. It is believed that the sum of these early and late benefits will justify the greater cost of RTKA.
      Following our preliminary study, which confirmed the greater cost of RTKA but did not demonstrate clinically significant improvements in LOS or complications, we focused our attention on the theory of improved revision rates. We sought to determine what reduction in revisions would be required to recapture the greater upfront cost of RTKA.
      Rather than compare the inpatient cost alone, we determined the total episode cost for each cohort. In our series the acute stay cost was greater for RTKA than MTKA. However, there were fewer readmissions in the RTKA cohort. These are costly, and we felt it would bias against RTKA to exclude this advantage from consideration. HHC and SNF costs, as part of the episode cost and therefore relevant in bundled payments, were also included. Complete cost data for all acute stays and readmissions were available through the medical record; however, costs for HHC and SNF were estimated from Centers for Medicare and Medicaid Services charges for primary total knee arthroplasty in our population of Bundled Payment for Care Improvement and Comprehensive Care for Joint Replacement patients.
      We chose the NJR as the reference for this study because of the size of its data pool and length of follow-up. The NJR provides survivorship data on a cohort of 133,729 Stryker Triathlon total knees with a maximum of 13 years of follow up [
      ]. At 10 years after surgery, the cumulative all-cause revision rate for this implant was 3.37% (95% CI 3.18-3.56). This equates to an expected 81 revisions (95% CI 76-85) per 2392 patient cohort at 10 years. In comparison, the American Joint Replacement Registry has a cohort of 79,687 Stryker Triathlon posterior-stabilized and cruciate-retaining knees, with follow-up to 7 years [
      ]. The cumulative 7-year revision rate in this registry is 1.2% (cruciate-retaining) to 1.8% (posterior-stabilized), which would equate to an expected 28-43 revisions per cohort (Fig. 3). We felt that the NJR data, with its larger pool and longer follow-up, were a more conservative estimate for this study.
      Figure thumbnail gr3
      Fig. 3The green line shows the revision rate for Stryker Triathlon TKA predicted from AJRR data is lower than the NJR projection.
      Although our study predicts that the greater cost of RTKA cannot be recaptured through a reduction in revisions, there are many variables which could change the analysis. Improved affordability of robotic technology, as well as competition from other manufacturer’s robotic platforms, will likely reduce cost. Improved utilization management of robotic patients may reduce the utilization of HHC and reduce the episode cost of RTKA.
      However, the movement of revision TKA cases to the outpatient setting will lower the episode cost. In addition, more recent registry data (ie, American Joint Replacement Registry) suggest a lower expected rate of revision for Stryker TKA than the NJR data we utilized. These factors would maintain or widen the gap between expected revisions and revisions prevented to offset cost differences.

      Strengths and Limitations

      The strengths of this study include the large sample size (2392 patients per cohort) and the access to direct costs, including costs related to the use of the surgical robot, from the Electronic Health Record. In addition, high-volume registry data with long-term follow-up provided a valid benchmark for the expected revision rate.
      There are limitations to be considered as well. The cohort study itself was retrospective and potentially subject to selection bias. Costs of the preoperative computed tomography scans required for RTKA were not included. Direct cost data included implant-and robot-associated costs, which are contractual and may be different in other health systems.
      Finally, actual cost data were available for only the acute stay and readmissions. Costs for HHC and SNF were estimated from similar bundled payment populations. However, these combined estimated costs represent only 5.8% (MTKA) to 6% (RTKA) of episode cost and are considered unlikely to have skewed the results.

      Conclusions

      This study suggests that RTKA will not become cost neutral with MTKA at 10-year follow-up through a reduction in revision rate alone. Although a reduction in revisions alone may not make RTKA cost neutral with MTKA, any reduction in revision burden achieved through robotics would be valuable and welcome. This, combined with demonstrated improvements in patient satisfaction and outcomes, would add value to RTKA and could change the analysis significantly.

      Acknowledgments

      The authors would like to acknowledge Erin Rogers for her assistance in collecting cost data and Joseph A Bosco, MD for his review of the manuscript. This study was solely funded by Providence Healthcare, no grant.

      Appendix

      Appendix 1Data From Matched Cohorts.
      MTKA (N = 2392)RTKA (N = 2392)P-Value
      Demographics of matched cohorts
       Age68.6 (8.7)68.6 (8.5).7493
       Female1363 (57%)1361 (57%).9534
       BMI30.6 (5.6)30.4 (5.7).3794
       ASA.7920
      I86 (4%)93 (4%)
      II1550 (65%)1560 (65%)
      III747 (31%)733 (31%)
      IV9 (0.4%)6 (0.3%)
      Complications
       Myocardial infarction (7 d)2 (0.1%)2 (0.1%)1.0000
       Pneumonia (7 d)3 (0.1%)1 (0.04%).3171
       OSM sepsis (7 d)3 (0.1%)1 (0.04%).3171
       Surgical site bleeding (30 d)0 (0%)0 (0%)NA
       Death0 (0%)0 (0%)NA
       Joint wound infection (90 d)8 (0.3%)10 (0.4%).6367
       Mechanical complications (90 d)0 (0%)0 (0%)NA
       Pulmonary embolism (30 d)7 (0.3%)4 (0.2%).3652
       Any complication21 (0.9%)17 (0.7%).5147
      RTKA, robotic total knee arthroplasty; MTKA, manual total knee arthroplasty; BMI, body mass index; ASA, American Society of Anesthesiologists.
      In-room/out-of-room time & procedure time.
      Median LOS (Hours)
      Mean LOS (Hours)
      Median total cost.
      Post discharge disposition and post-acute care utilization.
      Readmissions.

      Appendix A. Supplementary Data

      References

        • Sires J.D.
        • Craik J.D.
        • Wilson C.J.
        Accuracy of bone resection in MAKO total knee robotic-assisted surgery.
        J Knee Surg. 2021; 34: 745-748https://doi.org/10.1055/s-0039-1700570
        • Parratte S.
        • Price A.J.
        • Jeys L.M.
        • Jackson W.F.
        • Clarke H.D.
        Accuracy of a new robotically assisted technique for total knee arthroplasty: a cadaveric study.
        J Arthroplasty. 2019; 34: 2799-2803
        • Hampp E.L.
        • Chughai M.
        • Scholl L.Y.
        • Sodhi N.
        • Bhowmik-Stoka M.
        • Jacofsky D.J.
        Robotic-arm assisted total knee arthroplasty demonstrated greater accuracy and precision to plan compared with manual techniques.
        J Knee Surg. 2018; 32: 239-250
        • Song E.K.
        • Seon J.K.
        • Yim J.H.
        • Netravali N.A.
        • Bargar W.L.
        Robotic-assisted TKA reduces postoperative alignment outliers and improves gap balance compared to conventional TKA.
        Clin Orthop Relat Res. 2013; 471: 118-126
        • Manning W.
        • Ghosh M.
        • Wilson I.
        • Hide G.
        • Longstaff L.
        • Deehan D.
        Improved mediolateral load distribution without adverse laxity pattern in robot-assisted knee arthroplasty compared to a standard manual measured resection technique.
        Knee Surg Sports Traumatol Arthrosc. 2020; 28: 2835-2845
        • Shalhoub S.
        • Lawrence J.M.
        • Keggi J.M.
        • Randall A.L.
        • DeClaire J.H.
        • Plaskos C.
        • et al.
        Robotic-assisted total knee arthroplasty combined with a robotic tensioning system can help predict and achieve accurate postoperative ligament balance.
        J Arthroplasty. 2019; 5: 334-340
        • Kayani B.
        • Konan S.
        • Pietrzak J.R.T.
        • Haddad F.
        Iatrogenic bone and soft tissue trauma in robotic-arm assisted total knee arthroplasty compared with conventional jig-based total knee arthroplasty: a prospective cohort study and validation of a new classification system.
        J Arthroplasty. 2018; 33: 2496-2501
        • Hampp E.L.
        • Sodhi N.
        • Scholl L.
        • Deren M.E.
        • Yenna Z.
        • Westrich G.
        • et al.
        Less iatrogenic soft-tissue damage utilizing robotic-assisted total knee arthroplasty when compared with a manual approach.
        Bone Joint Res. 2019; 8: 495-501
        • Parsley B.S.
        Robotics in orthopedics: a brave new world.
        J Arthroplasty. 2018; 33: 2355-2357
        • Kayani B.
        • Konan S.
        • Tahmassebi J.
        • Pietrzak J.R.T.
        • Haddad F.S.
        Robotic-arm assisted total knee arthroplasty is associated with improved early functional recovery and reduced time to hospital discharge compared with conventional jig-based total knee arthroplasty.
        Bone Joint J. 2018; 100-B: 930-937
        • Jacofsky D.J.
        • Allen M.
        Robotics in arthroplasty: a comprehensive review.
        J Arthroplasty. 2016; 31: 2353-2363
      1. National Joint registry: 17th annual report. NJRcentre.org.uk, 2020 (Table 3.K7(a), p. 150)
      2. AAOS/AJRR: 7th Annual Report of the AJRR on Hip and Knee Arthroplasty (2020). NJRcentre.org.uk, 2020 (pp. 65-66)