Aims. There is concern that aggressive target pricing in the new Bundled Payment for Care Improvement Advanced (BPCI-A) penalizes high-performing groups that had achieved low costs through prior experience in bundled payments. We hypothesize that this methodology incorporates unsustainable downward trends on Target Prices and will lead to groups opting out of BPCI Advanced in favour of a traditional fee for service. Methods. Using the Centers for Medicare and Medicaid Services (CMS) data, we compared the Target Price factors for hospitals and physician groups that participated in both BPCI Classic and BPCI Advanced (legacy groups), with groups that only participated in BPCI Advanced (non-legacy). With rebasing of Target Prices in 2020 and opportunity for participants to drop out, we compared retention rates of hospitals and physician groups enrolled at the onset of BPCI Advanced with current enrolment in 2020. Results. At its peak in July 2015, 342 acute care hospitals and physician groups participated in Lower Extremity Joint Replacement (LEJR) in BPCI Classic. At its peak in March 2019, 534 acute care hospitals and physician groups participated in LEJR in BPCI Advanced. In January 2020, only 14.5% of legacy hospitals and physician groups opted to stay in BPCI Advanced for LEJR. Analysis of Target Price factors by legacy hospitals during both programmes demonstrates that participants in BPCI Classic received larger negative adjustments on the Target Price than non-legacy hospitals. Conclusion. BPCI Advanced provides little opportunity for a reduction in cost to offset a reduced Target Price for efficient providers, as made evident by the 85.5% withdrawal rate for BPCI Advanced. Efficient providers in BPCI Advanced are challenged by the programme’s application of trend and efficiency factors that presumes their cost reduction can continue to decline at the same rate as non-efficient providers. It remains to be seen if reverting back to Medicare fee for service will support the same level of care and quality achieved in historical bundled payment programmes. Cite this article: Bone Joint J 2021;103-B(6 Supple A):119–125

Introduction. Pressure to control health care costs may limit the ability of new implants to enter the market. Customized individually made (CIM) knee implants are produced from CT scans of each patient and may result in improved clinical outcomes based on early data showing less blood loss, reduced bone resection, and better implant function and alignment. Limited economic evidence suggests that the use of CIM technology may result in cost savings, particularly when post-discharge expenses are included. The purpose of this study was to evaluate real-world cost data to determine episode spending in a Medicare population receiving either CIM or off-the-shelf (OTS) implants. Methods. The Yale Center for Musculoskeletal Care and Baker Tilly Healthcare Management reviewed episode expenditures among Medicare beneficiaries who received CIM and OTS implants for TKA between 01/01/2015 and 12/31/2015. Episode costs included the pre-operative CT scan, index TKA procedure, and 12-month post-index spending for inpatient (IP), outpatient (OP), emergency room (ER), skilled nursing facility (SNF), and home health (HH) services. CIM patients were identified through a matching process utilizing de-identified patient demographic and procedural information and the presence of a CT scan 28–365 days before index. OTS patients included those without a CT scan within one year of index. CIM and OTS cohorts were propensity matched to produce comparable cohorts at a one-to-five ratio based on age range, gender, race, geographical location, and comorbidities. Average expenditure was used to calculate one-year costs of care differences between the CIM and the OTS technologies. A Generalized Linear Model (GLM) and two-part model (logistics and GLM) were used to test statistical significance. Results. The study included 4,434 patients (CIM: 739, OTS: 3,695). CIM index hospital expenses were 6.5% less than OTS (CIM: $11,579; OTS: $12,386, p<0.0001). CIM patients were 37.5% less likely to incur SNF costs than OTS patients (CIM: 3.0% vs. OTS: 4.8%; p=0.0241) and had 45.1% lower average SNF expenditures (CIM: $8,882 vs. OTS: $16,183; p=0.0236). There was no difference in the probability of incurring post-index IP costs (CIM: 15.7% vs. OTS: 15.4%; p=0.9437) but average post-index IP expenditures were 27.2% lower among the CIM cohort than OTS (CIM: $12,817 vs. OTS: $17,605; p=0.0008). However, CIM patients were 10.5% more likely to incur OP costs (CIM: 90.9% vs. OTS: 85.7%; p=0.0005) and had 6.1% higher average OP expenditures (CIM: $2,328 vs. OTS: $2,106; p=0.0377). Average 12-month episode spending for the CIM cohort was 8.4% less than the OTS cohort ($1,697 Difference; CIM: $18,585 vs. OTS: $20,280; p<0.0001). Conclusion. Study findings demonstrate that the use of CIM technology can result in significant 12-month episode savings among a Medicare population. The average expenditure differences noted for post-index IP, SNF, and OP cost categories suggest there are differences in required post-operative treatment across the two cohorts. Savings calculated in the analysis is a meaningful reduction for payers. Providers may also benefit from index and episode cost savings through shared savings arrangements. CIM technology should be considered as a method to reduce TKA episode spending

Introduction. Bundled Payments (BP) were a revolutionary new experiment for CMS that tested whether risk sharing for an episode of care would improve quality and reduce costs. The initial success of BP accelerated their growth as evidence by the launch of both mandatory and commercial bundles. Success in BP is dependent on the target price and the opportunity to reduce avoidable costs during the episode of care. There is concern that the aggressive target pricing methodology in the new model (BPCI-Advanced) penalizes high performing groups that already achieved low episode costs through prior experience and investment in BP. We hypothesize that this methodology incorporates unsustainable downward trends on target prices to a point beyond reasonableness for efficient groups to reduce additional costs and will lead to a large percentage of groups opting out of BPCI-A in favor of a return to fee for service (FFS) reimbursement. Methods. Using CMS data, we compared the target price factors for hospitals that participated in both BPCI classic (2013 –2018) and BPCI Advanced (beginning 10/2018), referred to as “legacy hospitals”, with hospitals that only participated in BPCI Advanced (beginning 10/2018). With the rebasing of BPCI-A target prices in Jan 2020 and the opportunity for participants to drop out of individual episode types or the program all together, we compared the retention of episode types that hospitals initially enrolled at the onset of BPCI-A with the current enrollment in 2020. Locally, we analyzed the BPCI-A target price factors across hospitals for a large orthopaedic practice that participated in BPCI Classic and the impact it had on the financial incentive/disincentive to remain in the lower extremity joint replacement episode type in 2020. Results. At its peak in July 2015, 423 acute care hospitals participated in one or more episode type in BPCI Classic. At its peak in March 2019, 715 acute care hospitals participated in one or more episode types in BPCI-Advanced. 130 (18%) of the hospitals in BPCI Advanced were also legacy participants in BPCI Classic, enrolling in 414 of the same episode types during both programs. In 2020, 251 (61%) of the episode types that hospitals were in enrolled in for both BPCI Classic and BPCI Advanced were dropped, suggesting prior experience in BPCI influences a participant's opportunity for success in BPCI Advanced. Furthermore, an analysis of the target price factors for episode types enrolled in by legacy hospitals during both programs suggests that prior participation in BPCI Classic is correlated with more aggressive target prices. A comparison of target price factors of similar hospitals reveals that legacy BPCI Classic hospitals that participated in lower extremity joint replacement (LEJR) BPCI Advanced received a larger negative adjustment on the target price (0.11 lower on average as a product of the Peer Adjusted Trend factor and ACH Efficiency factor) than non-legacy hospitals that participated in BPCI Advanced. Furthermore, analysis of the hospital targets for a large, high-performing legacy Physician Group Practice in BPCI Classic for LEJR revealed even greater negative adjustment on the target price than non-legacy participants. Comparing participants of similar peer groups on the Peer Adjusted Trend and ACH Efficiency factors suggests that CMS expects costs to decline more for legacy hospitals that have achieved efficiency than hospitals with no prior BP experience and higher baseline spend. Conclusions. BPCI Advanced provides little to no opportunity for a reduction in cost for already efficient TJA providers, as evident by the 55% dropout rate for BPCI-A participants in LEJR between model years 1 and 3. Efficient TJA providers in BPCI Advanced are challenged by the program's utilization of a peer adjusted trend factor and efficiency factor that presumes their costs will decline at the same aggressive rate or more than nonefficient TJA providers. It remains to be seen if reverting to Medicare fee for service will support the same level of care coordination, cost and quality achieved in historical TJA bundled payment programs

Aims. The purpose of this study was to determine the impact of the removal of total knee arthroplasty (TKA) from the Medicare Inpatient Only (IPO) list on our Bundled Payments for Care Improvement (BPCI) Initiative in 2018. Methods. We examined our institutional database to identify all Medicare patients who underwent primary TKA from 2017 to 2018. Hospital inpatient or outpatient status was cross-referenced with Centers for Medicare & Medicaid Services (CMS) claims data. Demographics, comorbidities, and outcomes were compared between patients classified as ‘outpatient’ and ‘inpatient’ TKA. Episode-of-care BPCI costs were then compared from 2017 to 2018. Results. Of the 2,135 primary TKA patients in 2018, 908 (43%) were classified as an outpatient and were excluded from BPCI. Inpatient classified patients had longer mean length of stay (1.9 (SD 1.4) vs 1.4 (SD 1.7) days, p < 0.001) and higher rates of discharge to rehabilitation (17% vs 3%, p < 0.001). Post-acute care costs increased when comparing the BPCI patients from 2017 to 2018, ($5,037 (SD $7,792) vs $5793 (SD $8,311), p = 0.010). The removal of TKA from the IPO list turned a net savings of $53,805 in 2017 into a loss of $219,747 in 2018 for our BPCI programme. Conclusions. Following the removal of TKA from the IPO list, nearly half of the patients at our institution were inappropriately classified as an outpatient. Our target price was increased and our institution realized a substantial loss in 2018 BPCI despite strong quality metrics. CMS should address its negative implications on bundled payment programmes. Cite this article: Bone Joint J 2020;102-B(6 Supple A):19–23

Introduction. The Centers for Medicare and Medicaid Services (CMS) removed total knee arthroplasty (TKA) from inpatient-only status in 2018. Our goal was to measure the adoption of outpatient TKAs, the impact on re-treatment rates, and the economic implications for hospitals. Methods. We utilized 100% national Medicare Part A fee-for-service (FFS) patient-level claims data for 2017–2018. We excluded DRG 469 TKAs since they are unlikely to be outpatient candidates, which left 257,107 primary TKAs in 2017 and 264,393 in 2018. We examined the time trend in monthly case volume and 30-day retreatment rate (defined as percent of patients having a second TKA within 30 days of the first. We calculated the loss in revenue for a hospital by multiplying the decrease in payment rate between inpatient and outpatient by the outpatient and total 2018 TKA volume. Results. In 2017 0.2% of primary TKAs were performed outpatient. Following the rule change, 25% of cases were performed outpatient in Q1 2018. This stayed at 25% in Q2, increased to 27% in Q3, and then increased to 30% in Q4 2018. The 30-day re-treatment rate was 0.16% in 2017 and 0.15% in 2018. Across hospitals there was the following distribution in the decrease in payment rate from inpatient to outpatient TKAs in 2018: 10. th. percentile: $1,994, 25. th. : $2,612, 50. th. : $3,487, 75. th. : $4,918, 90. th. : $7,231. In 2018 outpatient TKA coding cost hospitals (saved CMS) $243M in Medicare FFS payments (an average of $89,000 per hospital). If all TKAs were performed outpatient hospital Medicare FFS payments would have been $965M lower ($353,000 per hospital) in 2018. Conclusion. Outpatient TKA volumes grew through 2018. This did not impact 30-day retreatment rates. Medicare FFS payment rates declined by a median of $3,487 per outpatient case. As more TKAs are performed outpatient, total Medicare payments will further decline. For figures, tables, or references, please contact authors directly

Introduction. In November 2017, the Center for Medicare and Medicaid Services (CMS) finalized the 2018 Medicare Outpatient Prospective Payment System rule that removed total knee arthroplasty (TKA) procedures from the Medicare inpatient-only (IPO) list of procedures. This action had significant and unexpected consequences. For several years, CMS has utilized a rule called the “Two-Midnight Rule” to define outpatient status for all procedures not on the IPO list. CMS made TKA subject to the “Two-Midnight Rule” in conjunction with the decision to move TKA off the IPO list. According to the “Two-Midnight Rule,” a hospital admission should be expected to span at least two midnights in order to be covered as an inpatient procedure. If it can be reliably expected that the patient will not require at least two midnights in the hospital, the “Two-Midnight Rule” suggests that the patient is considered an outpatient and is therefore subject to outpatient payment policies. Under prior guidance related to the “Two-Midnight Rule;” however, CMS also states that Medicare may treat some admissions spanning less than two midnights as inpatient procedures if the patient record contains documentation of medical need. The final rule was clear in stating CMS's expectation was that the great majority of TKAs would continue to be provided in an inpatient setting. Methods. We looked at 3 different levels of the IPO rule impact on TKA for Medicare beneficiaries: a national comparison of fee for service (FFS) inpatient and outpatient classification for 2017 vs 2018; a survey of AAHKS surgeons completed in April of 2019; and an in-depth analysis of a large academic medical center experience. An analysis of change in inpatient classification of TKA patients over time, number of Quality Improvement Organization (QIO) audits, compliance solutions of organizations for the new rule and cost implications of those compliance solutions were evaluated. Results. Hospital reimbursement averages $10,122 in an outpatient facility (includes implant, other supplies, ancillary staff, etc.) but does NOT include the physician payment. Average hospital reimbursement in the inpatient setting is $11,760. The difference in reimbursement to hospitals varies widely however due to nuances in the CMS reimbursement formula (90. th. percentile decrease, $6,725 vs 10th percentile $2,048). Physician payments are the same in both settings (avg $1,403). TKA patients not designated for in-patient admissions are not eligible for bundle payment programs thus removing the healthiest, most predictable patients from the program. Patients designated as outpatients are subject to higher out of pocket expenses. Patients may have an annual Medicare Part B Deductible ($185) and a 20% copay as well as prescription and durable medical equipment costs. A survey of AAHKS surgeons demonstrated that 44.74% were doing inpatient designation only, 17.89% were doing outpatient designation only, 25.53 % designated patients as necessary, and 10% were designated by the hospital. This survey showed that 66/374 (17.65%) of AAHKS responders had undergone a QIO audit as a result of issues with the IPO rule. An evaluation of a large academic medical center demonstrated that since January 1, 2018, 470/690 (68.1%) of CMS TKA patients left the institution in less than 2 midnights. During this time period the institution was subjected to 2 QIO audits. All CMS patients had been designated as inpatients prior to May 2019. Conclusions. There are many unintended consequences to the IPO rule application to TKA. Clearly, more study is needed to provide better guidelines to knee replacement surgeons. A well-defined outpatient TKA bundle would allow CMS and TKA surgeons to better serve their patients. For figures, tables, or references, please contact authors directly

Introduction. Bundled reimbursement models for total knee arthroplasty (TKA) by the Center for Medicare and Medicaid Services (CMS) have resulted in an effort to decrease the cost of care. However, these models may incentivize bias in patient selection to avoid excess cost of care. We sought to determine the impact of the Comprehensive Care for Joint Replacement (CJR) model at a single center. Methods. This is a retrospective review of primary TKA patients from July 2015-December 2017. Patients were stratified by whether or not their surgery was performed before or after implementation of the CJR bundle. Patient demographic data including age, sex, and BMI were collected in addition to Elixhauser comorbidities and ASA score. In-hospital outcomes were then examined including surgery duration, length of stay, discharge disposition, and direct cost of care. Results. 1,248 TKA patients (546 Medicare and 702 Commercial Insurance) were evaluated, with 27.0% undergoing surgery prior to the start of the bundle. Compared to patients following implementation of the bundle, there was no significant difference in age, gender, or BMI. However, pre-CJR Medicare patients were more likely to have fewer Elixhauser comorbidities (p<0.001), prolonged length of stay (p<0.001), and greater discharges to inpatient facilities (p=0.019). There was no significant difference in direct hospital costs or operative service time comparing pre and post bundle patients. Conclusions. Implementation of the bundled reimbursement model did not result in biased patient selection at our institution; importantly, it also did not result in decreased hospital costs despite apparent improvement in value-based outcome metrics. This should be taken into consideration as future adaptations to reimbursement are made by CMS. For figures, tables, or references, please contact authors directly

Introduction. Patient-reported outcome (PRO) data are variably collected before and after total hip/knee arthroplasty (THA/TKA). We assessed the generalizability of incentivized, prospectively collected PRO data for THA/TKA patient-reported outcome performance measure (PRO-PM) development. Methods. The Centers for Medicare & Medicaid Services (CMS) received PRO data voluntarily submitted by hospitals in a bundled payment model for THA/TKA procedures. Participating hospitals who collected and successfully submitted these data received an increase in their overall quality score, possibly resulting in a positive impact on model reconciliation payments. PRO data were collected from Medicare Fee-For-Service beneficiaries >= 65 years undergoing elective primary THA/TKA procedures from July 1 to August 31, 2016 at hospitals participating in the model. Pre-operative PRO and risk variable data were collected 0 – 90 days prior to surgery, while post-operative PRO data were collected 270 – 365 days following elective THA/TKA. PRO pre-op and post-op data were matched to Medicare claims data for determination of clinically eligible procedures and clinical comorbidities. We compared the characteristics of patients submitting PRO data to other elective primary THA/TKA recipients in the US. Results. Four patient characteristics were associated with HOOS Jr. mean change scores (sex, narcotic use in past 90 days, other joint pain, and back pain) and four with KOOS Jr. mean change scores (sex, Hispanic ethnicity, other joint pain, and back pain). The frequency of simultaneous bilateral procedures, dementia, trauma, and dialysis were statistically significantly lower in patients submitting PRO data compared to other US Medicare beneficiaries undergoing elective primary THA/TKA, but no difference was greater than 1.5% absolute percentage points between groups. Conclusions. Offering financial incentives in alternative payment models to encourage PRO data collection and submission can produce generalizable data for PRO measure development. The possibility of non-respondent biases will need to be specifically considered in measure development

Abstract. Introduction. Periprosthetic joint infection (PJI) is a common cause of revision total knee surgery. Although debridement and implant retention (DAIR) has lower success rates in the chronic setting, it is an accepted treatment for acute PJI. There are two broad DAIR strategies: single debridement or a planned double debridement performed days apart. The purpose of this study is to evaluate the cost-effectiveness of single versus double DAIR with antibiotic beads for acute PJI in total knee arthroplasty (TKA). Methodology. A decision tree using single or double DAIR as treatment strategies for acute PJI was constructed. Quality Adjusted Life Years (QALYs) and costs associated with the two treatment arms were calculated. Treatment success rates, failure rates, and mortality rates were derived from the literature. Medical costs were derived from both the literature and Medicare data. A cost-effectiveness plane was constructed from multiple Monte Carlo trials. A sensitivity analysis identified parameters most influencing the optimal strategy decision. Results. Double DAIR with antibiotic beads was the optimal treatment strategy both in terms of the health utility state (82% of trials), and medical cost (97% of trials). Strategy tables demonstrated that as long as the success rate of double debridement is 10% or greater than the success rate of a single debridement, the two-stage protocol is cost-effective. Conclusions. This Markov analysis demonstrates that in the setting of acute PJI following TKA, a double DAIR with antibiotic beads is more cost effective than single DAIR from a societal perspective

Aims. The practice of alternating operating theatres has long been used to reduce surgeon idle time between cases. However, concerns have been raised as to the safety of this practice. We assessed the payments and outcomes of total knee arthroplasty (TKA) performed during overlapping and nonoverlapping days, also comparing the total number of the surgeon’s cases and the total time spent in the operating theatre per day. Materials and Methods. A retrospective analysis was performed on the Centers for Medicare & Medicaid Services (CMS) Limited Data Set (LDS) on all primary elective TKAs performed at the New England Baptist Hospital between January 2013 and June 2016. Using theatre records, episodes were categorized into days where a surgeon performed overlapping and nonoverlapping lists. Clinical outcomes, economic outcomes, and demographic factors were calculated. A regression model controlling for the patient-specific factors was used to compare groups. Total orthopaedic cases and aggregate time spent operating (time between skin incision and closure) were also compared. Results. A total of 3633 TKAs were performed (1782 on nonoverlapping days; 1851 on overlapping days). There were no differences between the two groups for length of inpatient stay, payments, mortality, emergency room visits, or readmission during the 90-day postoperative period. The overlapping group had 0.74 fewer skilled nursing days (95% confidence interval (CI) -0.26 to -1.22; p < 0.01), and 0.66 more home health visits (95% CI 0.14 to 1.18; p = 0.01) than the nonoverlapping group. On overlapping days, surgeons performed more cases per day (5.01 vs 3.76; p < 0.001) and spent more time operating (484.55 minutes vs 357.17 minutes; p < 0.001) than on nonoverlapping days. Conclusion. The study shows that the practice of alternating operating theatres for TKA has no adverse effect on the clinical outcome or economic utilization variables measured. Furthermore, there is opportunity to increase productivity with alternating theatres as surgeons with overlapping cases perform more cases and spend more time operating per day. Cite this article: Bone Joint J 2019;101-B:1081–1086

Aims. The Bundled Payments for Care Improvement (BPCI) initiative has identified pathways for improving the value of care. However, patient-specific modifiable and non-modifiable risk factors may increase costs beyond the target payment. We sought to identify risk factors for exceeding our institution’s target payment, the so-called ‘bundle busters’. Patients and Methods. Using our data warehouse and Centers for Medicare and Medicaid Services (CMS) data we identified all 412 patients who underwent total joint arthroplasty and qualified for our institution’s BPCI model, between July 2015 and May 2017. Episodes where CMS payments exceeded the target payment were considered ‘busters’ (n = 123). Risk ratios (RRs) were calculated using a modified Poisson regression analysis. Results. An increased risk of exceeding the target payment was significantly associated with increasing age (adjusted RR 1.04, 95% confidence interval (CI) 1.01 to 1.06) and body mass index (adjusted RR 1.03, 95% CI 1.003 to 1.06). Eight comorbid risk factors were also identified (all p < 0.05), only two of which were considered to be potentially modifiable (diabetes with complications and preoperative anaemia). An American Society of Anesthesiologist physical status classification system (ASA) score ≥ 3 (adjusted RR 2.3, 95% CI 1.67 to 3.18) and Charlson Comorbidity Index (CCI) ≥ 3 (adjusted RR 1.94, 95% CI 1.45 to 2.60) were risk factors for bundle busting. Conclusion. Non-modifiable preoperative risk factors can increase costs and exceed the target payment. Future bundled payment models should incorporate the stratification of risk. Cite this article: Bone Joint J 2019;101-B(7 Supple C):64–69

Introduction. The purpose of bundled payment programs is to reduce cost via risk sharing, while still maintaining quality. If savings are achieved under a historic target price, the orthopedic surgeon will receive a monetary bonus. If costs are higher, a portion is deducted from payment to the orthopedic surgeon. The purpose of this study was to evaluate our experience with the Bundled Payments for Care Improvement Program (BPCI) when run by an orthopedic surgeon group to determine patient safety and who benefited the most financially. Methods. This program ran from January 2015 through September 2018. 3,186 Medicare total hip and knee replacements, elective (DRG 470) and for fracture (DRG 469), performed by our group were included. 90 day hospital and all postoperative expenditures were reconciled against our historic cost. All patients were medically optimized with discharge plans established preoperatively. We developed preferred skilled nursing facilities and home health care agencies with synergistic medical providers so that discharges were recommended as soon as appropriate. We hired two full-time case managers to have direct contact with patients pre-and post-operatively. Waiver assistance such as house and pet sitters were used if necessary at our expense. 35% of savings went to the convener, who acted as a liaison between our group and CMS. Expenditures for the 90-day period for all patients were calculated to determine where savings occurred and which entity benefitted financially. Results. There was an average 9.2% reduction in hospital readmissions. An estimated total savings of $5,100,000 occurred. There was a 17% reduction in hospital costs, a 12.1% reduction in admissions to skilled nursing facilities with a 34% reduction in length of stay, and a 5% reduction in admissions to inpatient rehabilitation facilities. There was a 35% reduction in home health visits, but no change in outpatient physical therapy visits. After group expenses, final bonus to the orthopedic provider was on average $262 per patient. Conclusion. The physician managed program was very successful from Medicare's standpoint, achieving significant monetary savings without reducing quality of care. However, the bonus to the providing and managing physicians was nominal. It also does not take into consideration the 50 plus hours spent in meetings to develop this program. Participation could be considered a defensive posture so as not to lose more reimbursement. However, experience was gained which will be valuable for future gain sharing programs. Physicians and physician organizations need to sit at the head of the table to manage future payment bundles and perhaps also act as the convener. We deserve this, as a result of demonstrating high safety and cost savings. For figures, tables, or references, please contact authors directly

Introduction. As total knee arthroplasty incidence in the United States continues to increase, health care entities are looking to reform policy to decrease costs while improving efficiency and quality of care. The allocation of hospital and surgeon charges and payments is an important aspect of health care economics, but the trends and relationship between surgeon and hospital charges and payments for knee arthroplasty have not been well examined. The goal of this study is to report trends and variation in hospital charges and payments compared to surgeon charges and payments for total knee arthroplasty in a Medicare population. Methods. The 5% Medicare sample was used to capture hospital and surgeon charges and payments for total knee arthroplasty from 2005–2014. Two important values were calculated: (1) the charge multiplier (CM) which is the ratio of hospital to surgeon charges, and (2) the payment multiplier (PM), which is the ratio of hospital to surgeon payments. The year to year variation and regional trends in patient demographics, Charlson Comorbidity Index (CCI), length of stay (LOS), CM and PM were evaluated for all patients. Statistical significance of trends was evaluated using student's t-tests. Correlations between the financial multipliers and LOS were evaluated using a Pearson correlation coefficient (r). Results. 117,698 patients were included. Hospital charges were significantly higher than surgeon charges throughout the study period and increased substantially (CM increased from 7.9 to 11.4, p <0.0001) [Fig 1,2]. Hospital payments relative to surgeon payments ratio (PM) followed a similar trend, increasing from 10.0 to 14.6 (p < 0.0001). [Fig 1,2]. Similar trends were noted in all four regions of the US. LOS decreased significantly throughout the study from 3.8 to 2.8 days (p < 0.0001). CCI remained stable over the study period. Both the CM (r2 = −0.90) and PM (r2 = −0.84) were strongly negatively associated with LOS, meaning that as LOS decreased, the ratio of hospital to surgeon charges and payments (CM and PM) paradoxically significantly increased [Fig 3]. Conclusions. Hospital charges and payments relative to surgeon charges and payments have significantly increased for total knee arthroplasty despite stable patient complexity as measured by CCI and decreasing LOS. These results encourage the need for future studies with detailed cost analysis to identify the causes of hospital and surgeon financial malalignment. As healthcare shifts toward value-based care with shared responsibility for outcomes and cost, more closely aligned incentives between hospitals and providers is needed. For figures, tables, or references, please contact authors directly

Introduction. Recent studies of novel healthcare episode payment models, such as the Bundled Payments for Care Improvement (BPCI) initiative, have demonstrated pathways for improving value. However, these models may not provide appropriate payments for patients with significant medical comorbidities or complications. The objective of this study was to identify risk factors for exceeding our institution's target payment, the so-called “bundle busters.”. Methods. After receiving an exemption from the Institutional Review Board, we queried our institutional data warehouse for all patients (n=412) that underwent total joint arthroplasty (TJA) of the hip (n=192), knee (n=207), or ankle (n=13), and qualified for our institution's bundled payments model during the study time period (July 2015 – May 2017). Patients with medical conditions that were not well controlled or were potentially optimizable were all sent for preoperative medical optimization prior to surgery. For each 90-day episode, patient characteristics, medical comorbidities, perioperative data, and payments from the Centers for Medicare and Medicaid Services (CMS) were obtained. Episodes where Medicare payments exceeded the target payment were considered “busters”. The busters were older, and had higher comorbidity scores (all, p<0.01). Variables were summarized using descriptive statistics and risk ratios were calculated using a modified Poisson regression analysis. Results. Of the 412 patients, 123 were bundle busters (30%). There was a median institutional loss of $11,797 (IQR, $4,312 – $26,771) for the bundle busters and a median gain of $7,402 ($5,657 – $9,206) for the non-busters. Of the 32 risk factors evaluated, 11 were identified as Independent risk factors for busting the bundle (all, p<0.05). Nine of the 11 (82%) are non-modifiable risk factors and include age, disease specific diagnoses (fracture and avascular necrosis), and medical comorbidities (congestive heart failure, pulmonary circulation disorders, renal disease, cardiac arrhythmia, chronic pulmonary disease, and neurological disorder). The remaining two medical comorbidities are potentially modifiable and include diabetes with complications, and preoperative anemia. Conclusion. Though modifiable risk factors should continue to be optimized prior to TJA, as they were in this population, there are still many non-modifiable preoperative risk factors that can lead to costs exceeding the BPCI established institutional payment goal. As such, further work with payors may be needed to help fairly and appropriately consider these non-modifiable factors which result in increased costs

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. Background. The Bundled Payments for Care Improvement (BPCI) initiative was introduced to reduce healthcare costs while maintaining quality. We examined data from a healthcare system comprised of five hospitals that elected to participate in the BPCI Major Joint Replacement of the Lower Extremity Model 2 initiative beginning July 1, 2015. We compared one hospital that did 507 BPCI knee cases to the four other hospitals that did 566 cases. Stratifying the data by hospital volume, we sought to determine if costs decreased during the BPCI period, how the savings were achieved, and if savings resulted in financial rewards for participation. Methods. The Medicare data included the target cost for each episode (based on historical data from 2009–2012 for each hospital that was adjusted quarterly) and actual Part A and Part B spending for 90 days. Using 1,836 primary knee replacements, we analyzed the costs associated with the anchor hospitalization, inpatient rehabilitation, skilled nursing facilities, home health, outpatient physical therapy and readmission to compare the 1,073 knees done during the 16-month BPCI initiative period with the 763 knees done during the 1-year period preceding BPCI participation. Owing to the nonparametric distribution of the cost data, a Mann-Whitney U test was used to compare the higher volume hospital with the four lower volume hospitals. Results. Compared to the preceding year, the mean episode of care cost during BPCI participation decreased by 8.5% (from $20,853 to $19,087, p=0.24) at the higher volume hospital while remaining virtually unchanged (going from $20,383 to $20,380, p=0.10) at the lower volume hospitals. During the BPCI period, the mean Medicare 90-day target cost was $18,307 at the higher volume hospital and $22,287 at the lower volume hospitals (p<0.001). At the higher volume hospital, the major components of the savings included $367,290 from reduced readmission rates (5.7% versus 8.7%, p=0.11), $207,608 primarily due to a reduction in the length of stay at skilled nursing facilities (mean 15 days versus 25 days, p=0.005), and $130,894 associated with a decreased percentage of patients using inpatient rehabilitation (3.2% versus 4.9%, p=0.22). Although offset by other cost increases, the largest component of the savings at the lower volume hospitals was $262,548 due to a decrease in the percentage of patients (2.3% versus 4.8%, p=0.04) using inpatient rehabilitation. Despite its savings, the mean reconciliation penalty was $851 per case at the higher volume hospital while the lower volume hospitals received a mean reward of $2,165 per case. Conclusion. Based on the reduction in costs and decreased readmission rates, the BPCI initiative is achieving its objectives. Despite an 8.5% decrease in costs, the $18,307 target based on historical data resulted in an $851 penalty per case at the higher volume hospital. In contrast, as a result of a $3,980 higher target, the lower volume hospitals were rewarded even though they did not achieve cost savings. As structured, there is no incentive for centers with historically low costs to participate in BPCI

Introduction. The Center for Medicare and Medicaid Services (CMS) is faced with a challenge of decreasing the cost of care for total knee arthroplasty (TKA), but must make efforts to prevent patient selection bias in the process. Currently, no appropriate modifier codes exist for primary TKA based on case complexity. We sought to determine differences in perioperative parameters for patients with “complex” primary TKA with the hypothesis that they would require increased cost of care, prolonged care times, and have worse postoperative outcome metrics. Methods. We performed a single center retrospective review from 2015 to 2018 of all primary TKA. Patient demographics, medial proximal tibial angle (mPTA), lateral distal femoral angle (lDFA), flexion contracture, cost of care, and early postoperative outcomes were collected. ‘Complex’ patients were defined as those requiring stems or augments, and multivariable logistic regression analysis and propensity score matching were performed to evaluate perioperative outcomes. Results. 1046 primary TKA were studied and 84 patients (8.3%) were classified as “complex”. For this cohort, surgery duration was greater (117 vs 82 minutes; p<0.001), cost of care excessive (p<0.001), and patients had a greater likelihood for 90-day hospital return. Deviation of mPTA and lDFA was significantly greater preoperatively before and after propensity score matching. Cutpoint analysis demonstrated that preoperative mPTA <83o or >91o, lDFA <84o or >90o, flexion contracture >10o, and BMI > 35.7 were associated with ‘Complex’ procedures. Conclusions. Complex primary TKA may be identifiable preoperatively and are associated prolonged operative time, excess hospital cost of care and increased 90-day hospital returns. This should be considered in future reimbursement models to prevent patient selection bias, and a complexity modifier is warranted

Introduction. Previous studies on Medicare populations have shown improved outcomes and decreased 90-day episode-of-care costs with robotic assisted total knee arthroplasty (rTKA). The purpose of this study was to evaluate the expenditures and utilization following rTKA in the under 65 y/o population. Methods. TKA procedures were identified using the OptumInsights Inc. database. A two-year window was studied. The procedures were stratified in two groups: the rTKA or manual (mTKA) cohorts. Propensity score matching (PSM) was performed at 1:5. Utilization and associated costs were analyzed for 90 days following the index procedure. 357 rTKA and 1785 mTKA were included in this analysis. Results. Within the 90 days following the surgery, patients who had robotic assisted procedures were less likely to utilize inpatient services (2.24 vs. 4.37%; p=0.0444) and skilled nursing visits (SNF) (1.68 vs. 6.05%; p<0.0001). No patients in the robotic TKA group went to inpatient rehab while 0.90% of the manual cases went to an inpatient rehabilitation facility. Patients who utilized a home health aide in the rTKA arm utilized significantly fewer days of home health (5.33 vs. 6.36 days; p=0.0037). Cost associated with the utilization of these services was lower in the rTKA arm; the overall post-surgery expenditures were $1,332 less in the rTKA arm ($6,857 vs. $8,189; p=0.0018). The 90-day global expenditures (index plus post-surgery) were $4,049 less in the rTKA arm ($28,204 vs. $32,253; p<0.0001). Lastly, length of stay (LOS) after surgery was nearly a day less for the rTKA arm (1.80 vs. 2.72 days; p<0.0001). Conclusion. Robotic assisted TKA was associated with shorter LOS, reduced utilization of services and reduced 90-day payer costs when compared to the manual TKA. For figures, tables, or references, please contact authors directly

Introduction. There is insufficient data on the trends of anticoagulation after total knee arthroplasty (TKA) in the USA, and the efficacy and safety of rivaroxaban, beyond randomized clinical trials and small cohort studies. Patients and Methods. Using the Truven Health MarketScan database, we retrospectively evaluated new anticoagulation prescriptions after elective TKA from 2010 to 2015. The frequency of deep vein thrombosis (DVT), pulmonary embolism (PE), and adverse events, within 90 days, were then evaluated in 24,856 new users of warfarin and 21,398 new users of rivaroxaban in commercially insured patients (COM), and 15,483 new users of warfarin and 8,997 new users of rivaroxaban in Medicare supplement patients (MED). Data was analyzed by odds ratios using logistic regression models with stabilized inverse probability treatment weighting. Results. Warfarin use decreased from approximately 50% to 17% in COM patients and 60% to 25% in MED patients. Rivaroxaban use increased from 0 to 35% in COM patients and from 0 to 39% in MED patients. Older patients, females, a history of DVT, renal impairment, use of antiplatelet agents or surgery performed as an outpatient had lower odds of getting rivaroxaban. Patients in Western region and having surgery in 2015 had higher odds of getting rivaroxaban. COM patients with capitated insurance plans and a history of PE had lower odds of rivaroxaban initiation. MED patients with atrial fibrillation, cardiovascular disease or hyperlipidemia had lower odds of rivaroxaban initiation. Warfarin users had significantly higher odds ratio of DVT (OR 2.06 in COM patients and OR 2.21 in MED patients) and PE (OR 2.03 in COM patients and OR 2.16 in MED patients) than rivaroxaban users. There were no statistically significant differences in the bleeding risk between the two agents, but warfarin users had a significantly higher odds ratio of periprosthetic infection in both COM (1.57) and MED (1.79) patients. Conclusions. There has been an increase in prophylaxis with rivaroxaban, and a decrease in both warfarin and LMWH use after elective TKA over four years. Rivaroxaban had lower odds ratio of both DVT and PE than warfarin, and bleeding risks were similar. For figures, tables, or references, please contact authors directly

Introduction. Dialysis has been associated with increased complication rates following total knee arthroplasty (TKA). However, the current literature on this topic is limited and does not distinguish between hemodialysis (HD) and peritoneal dialysis (PD). It is unknown if the type of dialysis influences the risks of adverse outcome. The purpose of this study was to determine if infection and complication rates after TKA differ based on the mode of dialysis. Methods. Patients undergoing primary TKA from 2005 to 2014 were identified in the Medicare database. Patients with PD were matched 1:1 to patients with HD and 1:3 with patients who were not receiving either form of dialysis. Multivariate regression analysis was performed to examine several adverse events including the incidence of infection at 1 year and readmission to the hospital at 30 days. Results. PD was identified in 531 patients undergoing TKA. Infection rates were significantly lower in patients receiving PD compared to HD (3.39% versus 6.03%; p=0.026; odds ratio, 0.67; 95% CI 0.49 – 0.93). PD appeared to result in a similar infection rates when compared to matched controls who were not receiving dialysis. Other assessed outcomes (mortality rates, emergency room visits, hospital readmission rates) were very similar between PD and HD but noted to all be significantly higher than non-dialysis dependent controls. Discussion and Conclusions. The increased risk of complications in dialysis-dependent patients following TKA depends upon the type of dialysis. Whereas patients with HD have a significantly higher risk of infection, patients with PD do not appear to have this same risk when compared to non-dialysis dependent patients. To our knowledge, this is the only study to date to distinguish patients by the type of dialysis and suggests the type of dialysis should be used to assess risk when considering TKA

Introduction. Robotic assisted Total Knee Arthroplasty (rTKA), provides surgeons with preoperative planning and real-time data allowing for continuous assessment of ligamentous tension and range-of-motion. Using this technology, soft tissue protection, reduced early post-operative pain and improved patient satisfaction have been shown. These advances have the potential to enhance surgical outcomes and may also reduce episode-of-care (EOC) costs for patients, payers, and hospitals. The purpose of this study was to compare robotic assisted vs. manual total knee arthroplasty: 1) 90-day episode-of-care (EOC) costs; 2) index costs; 3) lengths-of-stay (LOS); 4) discharge disposition; and 5) readmission rates. Methods. TKA procedures were identified using the Medicare 100% Standard Analytic Files including; Inpatient, Outpatient, Skilled Nursing (SNF) and Home Health. Members included patients with rTKA or manual TKA (mTKA) between 1/1/2016-3/31/2017. To account for potential baseline differences, propensity score matching (PSM) was performed in a 1-to-5 ratio, robotic to manual based on age, sex, race, geographic division, and comorbidities. After PSM, 519 rTKA and 2,595 mTKA were identified and included for analysis. Ninety-day episode-of-care cost, index cost, LOS, discharge disposition and readmission rates were assessed. Results. Overall 90-day EOC costs were $2,391 less for rTKA patients ($18,568 vs. $20,960; p<.0001). Index facility cost and LOS were also less for rTKA patients by $640 ($12,384 vs. $13,024; p=.0001) and 0.7 days (p<.0001). Additionally, rTKA patients were discharged to SNF less frequently (12.52% vs. 21.70%; p<.0001) and home with health aid (56.65% vs. 46.67%; p<.0001) or self-care (27.55% vs. 23.62%; p=.0566) more frequently and had a 90-day readmission reduction of 33% (p=.0423). Conclusion. Robotic assisted TKA resulted in an overall lower 90-day episode-of-care cost when compared to manual TKA. The 90-day EOC cost savings of rTKA were driven by reduced facility costs, LOS and readmissions, and an economically beneficial discharge destination