To combat the growing cost of care for patients undergoing arthroplasty procedures, the Centers for Medicare & Medicaid Services (CMS) has instituted bundled reimbursement models that include both elective and nonelective arthroplasty procedures.1 These models incorporate a single target payment for the 90-day episode of care, which is adjusted for the presence of a fracture as well as medical severity diagnosis-related groups, based on the presence of a major complication or comorbidity.2,3 Patients who have a fracture or a major complication or comorbidity receive a higher target price for the episode of care. However, if the cost of care exceeds the target price, the result is a shared financial burden for the provider and financial loss.2 Importantly, in CMS bundled reimbursement initiatives, this structure applies to patients who have hip fracture and undergo total hip arthroplasty (THA).
Previous authors noted that these reimbursement models unfairly penalize hospitals that regularly treat patients with hip fracture.4 This patient population has a greater number of emergency department visits, a higher number of readmissions, and a wider standard deviation of episode charges compared with patients who undergo elective arthroplasty.5 The current reimbursement model does not appear to account for these discrepancies.6 Nonetheless, for many institutions, this model represents the current platform for billing and reimbursement.
Several authors have shown that reduced time to surgery for patients with hip fracture improves mortality and reduces length of stay (LOS).7 Therefore, in an effort to improve patient outcomes and reduce the cost of care, early operative intervention is advocated. Additionally, high-volume centers and surgeons reduce the cost of care for patients undergoing elective arthroplasty.8 However, the same outcomes have not been shown for patients with hip fracture, and the literature is conflicting.9–12 It is unclear whether high-volume arthroplasty surgeons should be on call to treat hip fractures that require THA within a bundled reimbursement system. Therefore, the authors conducted this study to determine whether fellowship-trained arthroplasty surgeons outperformed non-arthroplasty surgeons when treating hip fractures. The authors hypothesized that arthroplasty surgeons would achieve a shorter LOS, fewer discharges to inpatient facilities, and fewer 90-day readmissions compared with nonarthroplasty surgeons. These outcome metrics commonly serve as a surrogate for cost of care within bundled reimbursement models.
Materials and Methods
After approval was obtained from each center's institutional review board, the authors performed a multicenter retrospective review of patients with hip fracture who underwent THA between June 2013 and February 2018 at 2 tertiary referral academic centers participating in CMS bundled reimbursement models. All patients included in this series were Medicare patients who were treated for hip fracture. The study included a total of 4 hospitals across 2 academic centers. Academic trainees are involved in patient care at all 4 hospitals. At 3 of the 4 hospitals, patients with hip fracture are admitted to a medical service and undergo a preoperative evaluation, and surgery occurs within 24 to 48 hours by protocol. At 1 hospital, patients undergo medical screening, and if healthy, they are admitted directly to orthopedics rather than to a medical service. Postoperatively, patients are made weight bearing as tolerated, and they work with physical therapy (beginning on postoperative day 0 or 1) until discharge is appropriate. Final discharge disposition is determined with the help of physical therapists.
Patients were identified from the electronic database at each center and were stratified according to whether the operating surgeon was trained in arthroplasty. Patients who underwent other operative procedures, including hemiarthroplasty or placement of sliding hip screws, intramedullary nails, or percutaneous screws, were excluded. Additionally, patients with polytrauma, defined as those with other fractures, were excluded. The final cohort included 291 patients with hip fracture undergoing THA. Mean age was 81.1 years (range, 54–103 years), and 208 (71.5%) of the patients were women. Assuming disposition to a skilled nursing facility for 85% of patients with hip fracture, to detect a 15% difference between groups, this study would require approximately 270 total patients, for an alpha of 0.05 and power of 0.80.
Patient demographic data, including age, sex, and body mass index (BMI), were collected, in addition to American Society of Anesthesiologists (ASA) score (categorized as 1 or 2 vs 3 or 4). Perioperative outcome metrics included operative time, postoperative hospital LOS, disposition to home vs inpatient facility, and 90-day readmissions.
Continuous data were analyzed with a Mann–Whitney U test and are presented as median (lower quartile, upper quartile). Categorical data were analyzed with a chi-square test and are presented as count (percentage). Multivariable logistic regression analysis, controlling for age, sex, BMI, and ASA score, was performed for each outcome variable of interest: 90-day readmissions, disposition to a skilled nursing facility or rehabilitation, and LOS of 72 hours or more. These data are presented as odds ratios with 95% confidence intervals. Statistical analysis was performed with Mac Wizard Pro (E. Miller, Chicago, Illinois) and SPSS Statistics version 24 (IBM, Armonk, New York). P<.05 indicated statistical significance unless otherwise noted.
Of the 291 patients included in the study, 120 patients (41.2%) underwent surgery performed by a fellowship-trained adult reconstruction surgeon, and the rest underwent THA performed by other subspecialists as part of trauma coverage. No significant difference was found in patient age, sex, or ASA score between these 2 cohorts; however, patients treated by arthroplasty surgeons had a higher median BMI (24.4 vs 23.3 kg/m2, P=.010; Table 1).
For patients with hip fracture who underwent THA performed by arthroplasty surgeons, no significant difference was found in operative time or any of the outcome variables: 90-day readmissions, disposition, and LOS (Table 2). Multivariable logistic regression for each outcome variable was performed while controlling for patient age, sex, BMI, and ASA score. Arthroplasty surgeons were not more likely to have a reduction in 90-day readmissions (odds ratio, 1.539; 95% confidence interval, 0.704–3.365; P=.280), disposition to home (odds ratio, 1.232; 95% confidence interval, 0.556–2.733; P=.607), or LOS less than 72 hours (odds ratio, 3.988; 95% confidence interval, 0.355–44.829; P=.262).
Despite an expanding population of elderly patients, fortunately, the incidence of hip fracture has been decreasing.13 Still, the cost of treating patients with hip fracture is rising and is expected to reach $17 billion by 2025.14 Because of the successful outcomes with THA for osteoarthritis and fracture, this procedure has become an increasingly common option. In light of the high cost of care, CMS has chosen to include hip fractures as part of bundled initiatives in an effort to increase the value of care. As part of an effort to delineate value-based care in this patient population, the authors sought to determine whether fellowship-trained adult reconstruction surgeons had improved short-term outcomes compared with non-arthroplasty surgeons when treating patients with hip fracture. This study showed that arthroplasty surgeons did not achieve significant improvements in LOS, home discharge dispositions, or 90-day readmissions.
This study has important implications for several reasons. It is the first study to compare the acute outcomes of patients with hip fracture treated by surgeons in different specialties. Because they participate in bundled reimbursement initiatives, the authors' institutions have considered having an arthroplasty surgeon available rather than surgeons of other specialties, including orthopedic trauma specialists, when joint replacement is indicated for hip fracture. This practice could be expensive and logistically challenging for a hospital system. However, this study showed that none of the acute outcome measures evaluated were significantly different when patients were stratified based on surgeon specialization.
Wiegers et al9 performed a meta-analysis of 24 studies that included more than 2 million patients with hip fracture, focusing on the effect of hospital and surgeon volume on postoperative outcomes. They did not find a significant difference in postoperative mortality or complications based on surgeon volume. However, they identified 3 studies that explored LOS, all of which favored high-volume surgeons.10,15,16 In an analysis of patients undergoing hemiarthroplasty, Clement et al11 found that low-volume surgeons had more complications and dislocations. This finding is similar to that of Browne et al,10 who found that high-volume surgeons and hospitals influenced LOS, but only high-volume hospitals (not surgeons) significantly influenced cost. Importantly, these large registry studies had limited data availability. Additionally, these authors defined “high-volume” surgeons differently, as having 2 or more cases per year11 vs 16 or more cases per year.10 This difference in the definition may contribute to the current finding that the results were similar for arthroplasty specialists and other specialists. The other studies included surgeons with a fairly low number of cases per year, and both institutions in the current study have regularly rotating call schedules for trauma. Further, all hospitals have resident and fellow coverage as well as perioperative protocols to attempt to standardize patient care, regardless of surgeon training. The earlier studies also had inherent bias because high-volume centers are more likely to have high-volume surgeons. In these instances, as in the current study, more efficient perioperative (especially postoperative) practices and therapy may contribute to the reduced LOS seen with high-volume surgeons compared with low-volume surgeons at centers that do not regularly treat fragility fractures. By comparing surgeons within 2 tertiary referral institutions (4 hospitals), the current authors controlled for perioperative differences between high-and low-volume centers.
Lott et al17 reviewed the acute outcomes for patients with hip fracture before and after implementation of a bundled reimbursement initiative. They noted a significant reduction in LOS and discharge to facilities; however, they did not see a significant reduction in hospital readmissions or cost per episode of care. Because of the wide variation in episode costs among patients with hip fracture, several authors have argued that these patients should not be included in a bundled reimbursement initiative.5,6 Inability to truly alter episode of care costs for these patients may be affected by the need to proceed with surgical intervention within 24 to 48 hours of presentation, limiting preoperative optimization, which commonly occurs over several months or years before elective arthroplasty.
This study had several important limitations, as with any retrospective review. First, the number of patients with hip fracture who were treated under bundled reimbursement initiatives was limited. However, in light of the multicenter approach, the authors were able to detect a difference of approximately 15% in facility discharges, which was not observed in this cohort. Although a much larger cohort could detect more subtle differences in both disposition and readmissions, the authors believe that this will likely require a meta-analysis that combines data from multiple institutions. Additionally, the authors used LOS, disposition, and readmissions as surrogates for value-based care, as is commonly done in the literature. In light of the slightly reduced LOS and reduction in facility discharges for patients treated by arthroplasty surgeons, it is likely that there was a difference in the 90-day cost of care that was not evaluated in the current study and will require future research. It is also possible that arthroplasty surgeons were performing more technically difficult cases compared with non-arthroplasty surgeons, which was a potential confounding factor. Currently, the CMS does not consider case complexity within the Comprehensive Care for Joint Replacement bundle, and all patients are treated equally. Therefore, the authors considered it most appropriate to analyze them in this way. Another limitation was that because the procedures were performed at 2 academic centers, residents and fellow trainees routinely provided perioperative care for patients under hospital hip fracture protocols. Therefore, the results may not be transferable to community hospitals, where orthopedic surgeons are the primary admitting providers. Finally, the authors evaluated 90-day readmissions; however, in-hospital complications that otherwise could have resulted in a readmission were not considered. Although it is important, the authors did not define which complications were significant enough to potentially have triggered a readmission, and instead they used LOS as a surrogate for potentially significant postoperative complications during admission.
Surgeon training and specialty did not influence the early outcomes of patients with hip fracture who were treated with THA at 2 high-volume academic centers with established perioperative care pathways. Rather than focusing on the availably of arthroplasty surgeons, hospitals should concentrate on expeditious surgical intervention and appropriate perioperative care for this vulnerable patient population. Specialists in other orthopedic disciplines can achieve comparable early outcomes. Ultimately, further multicenter studies are needed to provide additional evidence of outcomes for patients with hip fracture after THA based on surgical subspecialty.
- Centers for Medicare & Medicaid Services. CMS.gov. https://www.cms.gov. Accessed April 14, 2020.
- Ryan SP, Goltz DE, Howell CB, et al. Predicting costs exceeding bundled payment targets for total joint arthroplasty. J Arthroplasty. 2019;34(3):412–417. doi:10.1016/j.arth.2018.11.012 [CrossRef] PMID:30518476
- McLawhorn AS, Buller LT. Bundled payments in total joint replacement: keeping our care affordable and high in quality. Curr Rev Musculoskelet Med. 2017;10(3):370–377. doi:10.1007/s12178-017-9423-6 [CrossRef] PMID:28741101
- Yoon RS, Mahure SA, Hutzler LH, Iorio R, Bosco JA. Hip arthroplasty for fracture vs elective care: one bundle does not fit all. J Arthroplasty. 2017;32(8):2353–2358. doi:10.1016/j.arth.2017.02.061 [CrossRef] PMID:28366309
- Schroer WC, Diesfeld PJ, LeMarr AR, Morton DJ, Reedy ME. Hip fracture does not belong in the elective arthroplasty bundle: presentation, outcomes, and service utilization differ in fracture arthroplasty care. J Arthroplasty. 2018;33(7S):S56–S60. doi:10.1016/j.arth.2018.02.091 [CrossRef] PMID:29622493
- Grace TR, Patterson JT, Tangtiphaiboontana J, Krogue JD, Vail TP, Ward DT. Hip fractures and the bundle: a cost analysis of patients undergoing hip arthroplasty for femoral neck fracture vs degenerative joint disease. J Arthroplasty. 2018;33(6):1681–1685. doi:10.1016/j.arth.2018.01.071 [CrossRef] PMID:29506928
- Bohm E, Loucks L, Wittmeier K, Lix LM, Oppenheimer L. Reduced time to surgery improves mortality and length of stay following hip fracture: results from an intervention study in a Canadian health authority. Can J Surg. 2015;58(4):257–263. doi:10.1503/cjs.017714 [CrossRef] PMID:26204364
- Courtney PM, Frisch NB, Bohl DD, Della Valle CJ. Improving value in total hip and knee arthroplasty: the role of high volume hospitals. J Arthroplasty. 2018;33(1):1–5. doi:10.1016/j.arth.2017.07.040 [CrossRef] PMID:28844631
- Wiegers EJA, Sewalt CA, Venema E, et al. The volume-outcome relationship for hip fractures: a systematic review and meta-analysis of 2,023,469 patients. Acta Orthop. 2019;90(1):26–32. doi:10.1080/17453674.2018.1545383 [CrossRef] PMID:30712501
- Browne JA, Pietrobon R, Olson SA. Hip fracture outcomes: does surgeon or hospital volume really matter?J Trauma. 2009;66(3):809–814. doi:10.1097/TA.0b013e31816166bb [CrossRef] PMID:19276758
- Clement RC, Strassle PD, Ostrum RF. Should all orthopaedists perform hemiarthroplasty for femoral neck fractures? A volume-outcome analysis. J Orthop Trauma. 2018;32(7):354–360. doi:10.1097/BOT.0000000000001176 [CrossRef] PMID:29664883
- Malik AT, Panni UY, Masri BA, Noordin S. The impact of surgeon volume and hospital volume on postoperative mortality and morbidity after hip fractures: a systematic review. Int J Surg. 2018;54(pt B):316–327. doi:10.1016/j.ijsu.2017.10.072 [CrossRef] PMID:29102691
- Lewiecki EM, Wright NC, Curtis JR, et al. Correction to: hip fracture trends in the United States, 2002 to 2015. Osteoporos Int. 2018;29(11):2583. doi:10.1007/s00198-018-4685-4 [CrossRef] PMID:30151621
- Burge R, Dawson-Hughes B, Solomon DH, Wong JB, King A, Tosteson A. Incidence and economic burden of osteoporosis-related fractures in the United States, 2005–2025. J Bone Miner Res. 2007;22(3):465–475. doi:10.1359/jbmr.061113 [CrossRef] PMID:17144789
- Lavernia CJ. Hemiarthroplasty in hip fracture care: effects of surgical volume on short-term outcome. J Arthroplasty. 1998;13(7):774–778. doi:10.1016/S0883-5403(98)90029-8 [CrossRef] PMID:9802663
- Okike K, Chan PH, Paxton EW. Effect of surgeon and hospital volume on morbidity and mortality after hip fracture. J Bone Joint Surg Am. 2017;99(18):1547–1553. doi:10.2106/JBJS.16.01133 [CrossRef] PMID:28926384
- Lott A, Haglin JM, Belayneh R, Konda S, Egol KA. Bundled payment initiative for hip fracture arthroplasty patients: one institution's experience. J Orthop Trauma. 2019;33(3):e89–e92. doi:10.1097/BOT.0000000000001378 [CrossRef] PMID:30562253
|Characteristic||Arthroplasty Surgeon (n=120)||Non-Arthroplasty Surgeon (n=171)||P|
|Age, median (lower quartile, upper quartile), y||82 (74, 89)||82 (76, 88)||.560|
|Female sex, No.||84 (70.0%)||124 (72.5%)||.640|
| White||108 (90.0%)||151 (91.0%)|
| Black||1 (0.8%)||3 (1.8%)|
| Asian||2 (1.7%)||5 (3.0%)|
| Other||7 (5.9%)||7 (4.2%)|
|Body mass index, median (lower quartile, upper quartile), kg/m2||24.4 (22.5, 27.3)||23.3 (20.5, 26.2)||.010a|
|American Society of Anesthesiologists score 1 or 2, No.||36 (30.0%)||68 (39.8%)||.087|
|Outcome||Arthroplasty Surgeon (n=120)||Non-Arthroplasty Surgeon (n=171)||P|
|Operating room time, median (lower quartile, upper quartile), min (n=251)||94 (79, 120)||87 (78, 108)||.210|
|Length of stay, median (lower quartile, upper quartile), h||120 (95, 168)||120 (96, 192)||.149|
| Home||29 (24.2%)||29 (17.0%)|
| Inpatient facility||91 (75.8%)||144 (83.0%)|
|90-day readmission, No.||16 (13.3%)||15 (8.8%)||.214|