Improvement in Total Joint Replacement Quality Metrics: Year One Versus Year Three of the Bundled Payments for Care Improvement Initiative

Dundon, John M. MD; Bosco, Joseph MD; Slover, James MD; Yu, Stephen MD; Sayeed, Yousuf MS; Iorio, Richard MD

Journal of Bone & Joint Surgery - American Volume:
doi: 10.2106/JBJS.16.00523
Scientific Articles

Background: In January 2013, a large, tertiary, urban academic medical center began participation in the Bundled Payments for Care Improvement (BPCI) initiative for total joint arthroplasty, a program implemented by the Centers for Medicare & Medicaid Services (CMS) in 2011. Medicare Severity-Diagnosis Related Groups (MS-DRGs) 469 and 470 were included. We participated in BPCI Model 2, by which an episode of care includes the inpatient and all post-acute care costs through 90 days following discharge. The goal for this initiative is to improve patient care and quality through a patient-centered approach with increased care coordination supported through payment innovation.

Methods: Length of stay (LOS), readmissions, discharge disposition, and cost per episode of care were analyzed for year 3 compared with year 1 of the initiative. Multiple programs were implemented after the first year to improve performance metrics: a surgeon-directed preoperative risk-factor optimization program, enhanced care coordination and home services, a change in venous thromboembolic disease (VTED) prophylaxis to a risk-stratified protocol, infection-prevention measures, a continued emphasis on discharge to home rather than to an inpatient facility, and a quality-dependent gain-sharing program among surgeons.

Results: There were 721 Medicare primary total joint arthroplasty patients in year 1 and 785 in year 3; their data were compared. The average hospital LOS decreased from 3.58 to 2.96 days. The rate of discharge to an inpatient facility decreased from 44% to 28%. The 30-day all-cause readmission rate decreased from 7% to 5%; the 60-day all-cause readmission rate decreased from 11% to 6%; and the 90-day all-cause readmission rate decreased from 13% to 8%. The average 90-day cost per episode decreased by 20%.

Conclusions: Mid-term results from the implementation of Medicare BPCI Model 2 for primary total joint arthroplasty demonstrated decreased LOS, decreased discharges to inpatient facilities, decreased readmissions, and decreased cost of the episode of care in year 3 compared with year 1, resulting in increased value to all stakeholders involved in this initiative and suggesting that continued improvement over initial gains is possible.

Author Information

1Department of Orthopaedic Surgery, NYU Langone Medical Center, New York, NY

E-mail address for J.M. Dundon:

E-mail address for J. Bosco:

E-mail address for J. Slover:

E-mail address for S. Yu:

E-mail address for Y. Sayeed:

E-mail address for R. Iorio:

Article Outline

In 2011, the U.S. Centers for Medicare & Medicaid Services (CMS) initiated the Bundled Payments for Care Improvement (BPCI) initiative. This CMS program was founded to “improve patient care through payment innovation that fosters improved coordination and quality through a patient centered approach”1,2. The goal of this initiative is to foster efficient, collaborative care while reducing costs and protecting or improving the overall quality of care3. This is achieved through fixed payments and can involve either retrospective or prospective payment based on the BPCI model chosen. Retrospective payment typically involves accepting a discounted payment from CMS based on the selected model, and the cost-savings from each episode can allow for incentives, such as gain-sharing between the hospital, the physician, and other stakeholders.

We implemented BPCI Model 2 for total joint arthroplasty in January 2013, and started the financial risk phase in October 2013. Included were Medicare Severity-Diagnosis Related Group (MS-DRG) 469 (total joint replacement of the lower extremity with major complications or comorbidities) and MS-DRG 470 (total joint replacement of the lower extremity without major complications or comorbidities). Total hip arthroplasty performed in the setting of femoral neck fracture was included. We chose Model 2, which involves a retrospective form of payment and requires a discounted reimbursement from CMS. The Model 2 episode of care includes the 72 hours prior to the procedure; the inpatient and post-acute care including procedural, physician, and all consultant fees; and all costs through 90 days following discharge. After the initiation of this demonstration project, our first-year results showed a decrease in length of stay (LOS) and readmissions and an increase in discharge to home, as previously reported4-6. The purpose of the current study was to compare year-1 and year-3 results of the BPCI program with respect to hospital LOS; discharge to inpatient facilities; 30, 60, and 90-day readmission rates; and the cost of the episode of care. Our hypothesis was that, after implementing the BPCI initiative, we would increase value in our total joint arthroplasty care, thus experiencing an improvement in hospital quality metrics and a positive impact on overall costs.

Back to Top | Article Outline

Materials and Methods

Year 1 of the BPCI initiative was compared with year 3 using quality and cost metrics as the basis of comparison. LOS, readmissions, discharge disposition, and cost per episode of care were analyzed. A number of programs were implemented within our institution after the first year to improve quality metrics: a surgeon-directed preoperative risk-factor optimization program7,8, enhanced care coordination and home services, a change in VTED (venous thromboembolic disease) prophylaxis to a risk-stratified protocol8, infection-prevention measures9,10, an emphasis on discharge to home rather than to an inpatient facility, and a quality-dependent gain-sharing program among surgeons. A surgeon-directed readmission risk-assessment program was also established, and no additional resources were given to this by our institution.

Back to Top | Article Outline


There were 721 consecutive Medicare primary total joint arthroplasty patients in year 1 (January 1, 2013, to December 31, 2013) and 785 consecutive total joint arthroplasty patients in year 3 (June 1, 2014, to May 31, 2015); their data were compared. No change in baseline demographic data between the 2 cohorts was noted. The average hospital LOS decreased from 3.58 days in year 1 to 2.96 days in year 3. The rate of discharge to an inpatient facility decreased from 44% to 28%. The 30-day all-cause readmission rate decreased from 7% to 5%, the 60-day all-cause readmission rate decreased from 11% to 6%, and the 90-day all-cause readmission rate decreased from 13% to 8% (Table I).

The largest CMS cost decrease was an 88% reduction in inpatient rehabilitation costs per episode of care by quarter 3 of 2014. Accordingly, there was an increase in home health-care costs by 23%; however, these costs were more than offset by the decrease in discharges to inpatient rehabilitation facilities. CMS costs related to the index admission did not experience much change, increasing only by 1% by quarter 3 of 2014. The average costs to CMS per episode of care decreased by 20% (Table II).

Back to Top | Article Outline


Porter defined health-care value as outcomes that matter to the patient divided by the cost of the health-care intervention11. In order to provide efficient care, costs need to be decreased while outcomes are maintained or improved, resulting in higher-quality care with greater value. Health-care reform efforts have mandated efforts aimed to decrease the costs associated with elective procedures, such as total hip arthroplasty and total knee arthroplasty, while maintaining or improving the quality of the care. One of the methods to encourage delivery of more value for total joint arthroplasty is through alternative payment models, such as the BPCI initiative. Unlike with fee-for-service reimbursement models, in a bundled payment model, the hospital, surgeon, insurer, and patient become stakeholders in determining the resources allocated to the procedure and the ultimate cost of the episode of care. Making the stakeholders responsible for the allocation of resources and optimization of the value equation is a change in payment paradigm that does not promote volume-only care delivery, as fee-for-service did, and allows for risk accountability for the stakeholders. The BPCI initiative aims to improve delivery of quality-based care in total joint arthroplasty, promoting value over volume.

At a large, high-volume, academic hospital, we were able to improve the quality of care by implementing evidence-based protocols and increasing care management services, leading to a decrease in overall LOS, discharges to inpatient post-acute care facilities, and readmission rates at 30, 60, and 90-day intervals over a 3-year period. There are costs associated with the creation of these committees and a substantial time commitment by the surgeons involved and team members as described in our previous report6. This effort resulted in considerable cost savings to the hospital system while improving patient outcome and the quality of care6.

Physician reimbursements beyond the standard case-based Medicare rates are determined by a gain-sharing formula agreed upon with the hospital. Individual physician participation in this gain-sharing arrangement is predicated on achieving predetermined quality metrics. Physician gain-sharing is capped at 50% above the standard Medicare reimbursement rates for the procedure or service. All physicians are required to participate in the bundle, and BPCI rules prohibit patients from receiving financial incentives4,6. We saw a substantial increase in physician reimbursement through gain-sharing over the period studied at our institution.

We employed evidence-based, standardized protocols to increase value without sacrificing quality or patient outcomes. Surgeon-directed preoperative risk-factor optimization using the Readmission Risk Assessment Tool (RRAT)12 helped to identify patients at high risk for complications and readmission. Risk factors predictive of readmission that were addressed using the RRAT include cardiac and stroke predisposing conditions, increased body mass index (BMI), poor diabetes control, VTED history or thrombophilia, smoking, behavioral/neurocognitive problems, drug and alcohol abuse, and Staphylococcus aureus colonization or infection risk; these modifiable risk factors have been linked to higher readmission rates1,5,6,9,12. A previous study has shown that increasing severity of illness is associated with not only an increased 30-day readmission rate but also an incremental increase in cost associated with those admissions and an increasing percentage of the target price of the bundled payment reconciliation consumed by those readmissions1. RRAT scores of >3 were shown to have a significantly higher odds ratio for readmission within our institution12.

In our institution, the use of the RRAT recently was combined with our perioperative orthopaedic surgical home (POSH) program in order to help optimize patients medically for surgery. The goal is to mitigate risk factors for perioperative and postoperative complications through early detection and management of patient comorbidities preoperatively. This prevents the exclusion of high-risk patients and provides a pathway to joint replacement through risk mitigation and optimization. Through standardization of the preoperative screening process and optimization protocols, we were able to lower our 90-day readmission rates substantially, from 13% to 8%. This departmental model at our institution is still in its infancy and currently is completely surgeon-directed. Currently, no additional resources are devoted to this program. In the future, we plan to expand this program to include the use of risk-stratification coordinators to further mitigate preoperative risk.

We have also changed our VTED prophylaxis regimen to a risk-stratified protocol. The use of newer, more aggressive anticoagulants such as fractionated heparins and factor Xa and direct thrombin inhibitors effectively lowers the rates of VTED following total joint arthroplasty but increases perioperative bleeding risk. Aspirin as a chemoprophylactic agent was demonstrated to significantly lower bleeding risk without changing clinical pulmonary embolism rates, although it was associated with significantly greater residual venographic clotting compared with the more potent agents13. Currently the American Academy of Orthopaedic Surgeons recommends that patients at high risk for VTED may warrant more aggressive chemoprophylaxis14. In order to resolve or avoid the deleterious side-effects of aggressive anticoagulation in addition to increased cost and monitoring, our institution implemented a risk-stratified protocol for all total joint arthroplasty patients between our year-1 and year-3 BPCI experience. Patients without risk factors (BMI of >40 kg/m2, history of VTED or thrombophilia, smoking, and/or active cancer treatment) received aspirin and use of sequential pneumatic compression devices for VTE prophylaxis, and those with any of the 4 listed risk factors underwent chemoprophylaxis with low-molecular-weight heparin. This change in protocol was deemed to be safe and efficacious, and demonstrated cost savings8.

Post-acute discharge costs were lowered by decreasing the use of inpatient post-discharge care venues. We instituted care coordination programs to preoperatively align patient expectations for going home rather than to an inpatient facility. Through focused discussion and education among the patient, the physician, and dedicated care coordinators and home services, our institution was able to decrease the rate of discharge to inpatient facilities. This enhanced care coordination resulted in a substantial reduction (88%) in such costs per episode of care during this time period. This has also improved our readmission rates, a finding supported by our previous studies demonstrating higher readmission rates and complication rates among patients discharged to an inpatient facility versus discharged to home4-6,12.

Implant costs were controlled through negotiation and containment of physician waste, and variability in cost was controlled through the establishment of ceiling pricing for acceptable implants. Implant costs account for 20% to 40% of episode-of-care hospital cost. On the basis of national data and in collaboration with both physicians and vendors, our institution was able to establish a target total implant price per procedure and decrease overall implant costs7. Increased departmental oversight regarding surgeon and case-specific implant utilization was implemented15. Specifically, implant waste was scrutinized, and through improved coordination with operating room staff and heightened implant utilization monitoring, a significant reduction (by approximately 50%) in the costs attributed to implant waste was experienced in year 115.

Blood management was also a point of emphasis. The standardization of the administration of tranexamic acid to lower transfusion rates, decreasing blood transfusions by the lowering of transfusion triggers, the elimination of reinfusion drains, and decreased use of cell savers led to a significant decrease in hospital costs16.

Each reconciliation has allowed our institution to realize a positive margin for shared savings by averaging less than the target price for the episode of care. There are substantial infrastructure costs involved for the bundled-payment program, thus making the program a financial break-even proposition for our institution. Prior to the BPCI initiative, the institution was losing $7,000 per case among Medicare total joint arthroplasty patients.

Through the implementation of strict evidence-based protocols focused on preoperative risk optimization, the appropriate utilization of hospital resources, intensive post-discharge care coordination, discharge to home rather than to an inpatient facility, and improved implant pricing, our institution improved the overall quality of care, lowered readmission rates, and decreased episode-of-care costs by year 3 of implementation of the BPCI program. These cost-reduction and quality-improvement programs, combined with increased physician and hospital alignment, led to a substantial increase in the value of the health care delivered to total joint arthroplasty patients without compromising the quality of care. Additionally, our gain-sharing program based on episode-of-care cost savings and quality improvement has improved physician and hospital alignment.

This is a mid-term report, and long-term data are needed in order to confirm our results. These results were achieved in a high-volume, tertiary, academic center with a hybrid compensation system. As part of the directives for successfully implementing a BPCI initiative and because of the need to enhance value in total joint arthroplasty, we were able to obtain considerable resources to establish an intensive care management infrastructure instrumental to the reduction in hospital resource utilization, and to transmit quality and financial performance data to our surgeons in a relevant time frame. Therefore, our results may not be applicable to low-volume, community, or less-resourced centers. More data and varied hospital settings are necessary in order to confirm the wider applicability of these results. However, our results demonstrate that value-based alternative payment models can be successful in reducing cost while maintaining or even improving quality for total joint arthroplasty. This bodes well for the more widespread implementation of total joint arthroplasty alternative payment paradigms as mandated in the Comprehensive Care for Joint Replacement (CJR) model.

Alternative payment models may have unintended consequences. There are incentives to restrict care from patients who are at higher risk of complications. Affected patients may include those who are from lower socioeconomic classes, have higher rates of comorbidities, or have less access to centers of excellence and who therefore may incur more costs and be fiscally unsound for this payment model. We believe that, through risk-factor optimization with the BPCI models available, the risk can be mitigated and optimized to provide a high quality of care. It is imperative that we focus our efforts on inclusion and optimization of these patients and not move toward a path of exclusion of care. Delaying care until patients are optimized is the right thing to do for the patients and the health-care system, but restricting access to care is unacceptable.

Investigation performed at the Department of Orthopaedic Surgery, NYU Langone Medical Center, New York, NY

Disclosure: Institutional funds were responsible for the implementation of the BPCI program. There was no external source of funding for this project. On the Disclosure of Potential Conflicts of Interest forms, which are provided with the online version of the article, one or more of the authors checked “yes” to indicate that the author had a relevant financial relationship in the biomedical arena outside the submitted work.

Disclaimer: The statements contained in this document are solely those of the authors and do not necessarily reflect the views or policies of CMS. The authors assume responsibility for the accuracy and completeness of the information contained in this document.

Back to Top | Article Outline


1. Kiridly DN, Karkenny AJ, Hutzler LH, Slover JD, Iorio R, Bosco JA 3rd. The effect of severity of disease on cost burden of 30-day readmissions following total joint arthroplasty (TJA). J Arthroplasty. 2014 ;29(8):1545–7. Epub 2014 Apr 5.
2. Centers for Medicare & Medicaid Services. Bundled Payments for Care Improvement (BPCI) initiative: general information. Accessed 2016 Sep 26.
3. Froimson MI, Rana A, White RE Jr, Marshall A, Schutzer SF, Healy WL, Naas P, Daubert G, Iorio R, Parsley B. Bundled Payments for Care Improvement initiative: the next evolution of payment formulations: AAHKS Bundled Payment Task Force. J Arthroplasty. 2013 ;28(8)(Suppl):157–65.
4. Iorio R. Strategies and tactics for successful implementation of bundled payments: bundled payment for care improvement at a large, urban, academic medical center. J Arthroplasty. 2015 ;30(3):349–50. Epub 2015 Jan 23.
5. Yu S, Garvin KL, Healy WL, Pellegrini VD Jr, Iorio R. Preventing hospital readmissions and limiting the complications associated with total joint arthroplasty. J Am Acad Orthop Surg. 2015 ;23(11):e60–71.
6. Iorio R, Clair AJ, Inneh IA, Slover JD, Bosco JA, Zuckerman JD. Early results of Medicare’s bundled payment initiative for a 90-day total joint replacement episode of care. J Arthroplasty. 2016 ;31(2):343–50.
7. Bosco JA, Alvarado CM, Slover JD, Iorio R, Hutzler LH. Decreasing total joint implant costs and physician specific cost variation through negotiation. J Arthroplasty. 2014 ;29(4):678–80. Epub 2013 Oct 14.
8. Odeh K, Doran J, Yu S, Bolz N, Bosco J, Iorio R. Risk-stratified venous thromboembolism prophylaxis after total joint arthroplasty: aspirin and sequential pneumatic compression devices vs aggressive chemoprophylaxis. J Arthroplasty. 2016 ;31(9)(Suppl):78–82. Epub 2016 Mar 16.
9. Maoz G, Phillips M, Bosco J, Slover J, Stachel A, Inneh I, Iorio R. The Otto Aufranc Award: modifiable versus nonmodifiable risk factors for infection after hip arthroplasty. Clin Orthop Relat Res. 2015 ;473(2):453–9.
10. Mehta S, Hadley S, Hutzler L, Slover J, Phillips M, Bosco JA 3rd. Impact of preoperative MRSA screening and decolonization on hospital-acquired MRSA burden. Clin Orthop Relat Res. 2013 ;471(7):2367–71. Epub 2013 Feb 20.
11. Porter ME. What is value in health care? N Engl J Med. 2010 ;363(26):2477–81. Epub 2010 Dec 8.
12. Boraiah S, Joo L, Inneh IA, Rathod P, Meftah M, Band P, Bosco JA, Iorio R. Management of modifiable risk factors prior to primary hip and knee arthroplasty: a readmission risk assessment tool. J Bone Joint Surg Am. 2015 ;97(23):1921–8.
13. Pellegrini VDJ. Prevention of pulmonary embolism after total joint arthroplasty: who knows best? The Orthopaedic Journal of Harvard Medical School. 2008.
14. American Academy of Orthopaedic Surgeons. Preventing venuous thromboembolic disease in patients undergoing elective hip and knee arthroplasty: evidence-based guideline and evidence report. 2011. Accessed 2015 Dec 1.
15. Payne A, Slover J, Inneh I, Hutzler L, Iorio R, Bosco JA 3rd. Orthopedic implant waste: analysis and quantification. Am J Orthop. 2015 ;44(12):554–60.
16. Evangelista PAMK, Brandt E, Inneh IA, Bosco JA, Iorio R. Effect of TXA on transfusion rates following TJA: a cost and comparative effectiveness analysis. J Arthroplasty. [In press].
Copyright 2016 by The Journal of Bone and Joint Surgery, Incorporated