Abstract
Background
Implementation of payment reform for breast reconstruction following mastectomy demands a comprehensive understanding of costs related to the complex process of reconstruction. Bundled payments for services to women with breast cancer may profoundly impact reimbursement and access to breast reconstruction. Our objectives were to determine the contribution of cancer therapies, comorbidities, revisions, and complications to costs following immediate reconstruction and the optimal duration of episodes to incentivize cost containment for bundled payment models.
Methods
The cohort was comprised of women who underwent immediate breast reconstruction between years 2009 to 2016 from the MarketScan Commercial Claims and Encounters database. Continuous enrollment for three months prior and 24 months following reconstruction was required. Total costs were calculated within predefined episodes (30 days, 90 days, one year, and two years). Multivariable models assessed predictors of costs.
Results
Among 15,377 women in the analytic cohort, 11,592 (75%) underwent tissue expander (TE), 1,279 (8%) direct to implant (DTI), and 2,506 (16%) autologous reconstruction. Adjuvant therapies increased costs at one year [TE +$39,978 (p=<0.001), DTI +$34,365 (p=<0.001), and autologous +$29,226 (p=<0.001)]. At one year, most patients had undergone TE exchange (76%) and revisions (81%), and a majority of complications had occurred (87%). Comorbidities, revisions, and complications increased costs for all episode scenarios.
Conclusions
Bundling should consider separate bundles for medical and surgical care with adjustment for procedure type, cancer therapies, and comorbidities to limit the adverse impact on access to reconstruction. Our findings suggest that a one-year time horizon may optimally capture reconstruction events and complications.
INTRODUCTION
Implementation of innovative payment reforms to limit unsustainable health spending is occurring throughout the United States.1,2 Although the Centers for Medicare and Medicaid (CMS) is testing several alternative payment reform models, episode-based bundled payments have emerged as a viable option to limit unnecessary spending without affecting the overall quality of care. Bundled payments represent an arrangement wherein hospitals and providers are paid a predetermined lump sum in exchange for all services furnished to a patient during a time period or over the course of a defined clinical condition.3,4 Bundling aims to indirectly increase financial accountability of providers and hospitals for the costs and quality of care delivered by linking financial reimbursement to clinical outcomes.
The American College of Surgeons Coding and Reimbursement committee established criteria that surgical procedures should satisfy to be considered for bundled payment models.5 Among the suggested criteria for bundling, procedures should be both relatively common throughout the United States and costly with measurable outcomes and costs. Breast reconstruction following mastectomy is one of the most common reconstructive operations in the United States with an estimated 106,295 women who underwent these procedures in 2017.6 Despite innovations in the surgical approaches over the past few decades, prior studies suggest rates of major complications between 20% and 35%.7–11 Given the relatively high complication rates seen in prospective studies, the need to improve quality and value in breast reconstruction has become a priority in oncologic surgery. Although breast reconstruction has not been specifically identified for bundling by CMS, it satisfies many of the stated criteria to be considered in the future as a specific episode of care or as one component of care to women with breast cancer.12 There are now several examples of bundling in the private sector that have resembled those in practice by CMS, which further increases the possibility of bundling for breast reconstruction.13,14 Furthermore, it is estimated that 30% of hospital systems and 20% of self-insured employers are engaged in some form of bundled payment contracting.15
Approaches to breast reconstruction following mastectomy have evolved significantly over the past few decades. Plastic surgeons are now key members of a multidisciplinary team providing care to women with breast cancer. Reconstruction is often performed in conjunction with mastectomy and tends to involve a single- or multi-stage approach based upon the individual clinical and anatomic challenges of each patient.16 Furthermore, the clinical outcomes of reconstruction are strongly affected by adjuvant treatments such as radiation therapy.11 Given the complexities inherent to breast reconstruction in the current paradigm of care, our objectives were to characterize the contribution of cancer therapies, patient comorbidities, revisions, and complications to costs following immediate reconstruction and the optimal duration of payment episodes for consideration in future bundled payment models.
METHODS
Data Source and Analytic Cohort
The MarketScan Commercial Claims and Encounters Database includes insurance claims from more than 43 million individuals in the United States. This database is the largest convenience sample available among proprietary datasets and believed to be representative of individuals in the United States with employer-provided insurance. In addition to longitudinal tracking of enrollees over time, the database captures complete episodes of care through claims from inpatient stays, outpatient procedures, office visits, prescription information, and laboratory tests.
Women with a diagnosis of breast cancer or a strong family history of breast cancer who underwent mastectomy with immediate breast reconstruction between years 2009 and 2016 were selected from the MarketScan database. To be eligible for inclusion, women must have been at least 18 years of age and with continuous enrollment within three months prior to and 24 months following the mastectomy. Women with distant metastases within three months of the mastectomy and enrollees with missing cost information or two-year bundled payment costs >99th percentile were excluded.
Episodes of Care Duration
The episode of care durations were designed to capture common pathways for breast reconstruction performed in either single or multiple stages. Because breast reconstruction most often occurs in a staged process over a period time based upon surgeon and patient preference in addition to clinical context, we compared total costs over periods of 30 days, 90 days, one year, and two years. We also separately assessed costs in the preoperative period to capture use of services such as preoperative angiography that are often used in preparation for autologous reconstruction.
Costs of Breast Reconstruction
Total costs were calculated within the predefined episodes with the total payments variable in the dataset, which represents the total payment for a service from all payers prior to the application of discounts like copayments, deductibles, or coordination of benefits. Because these values are reported as payments from one or more payers to a provider, costs in this study are indicative of the amount of money that was reimbursed. We adjusted all costs to the 2016 United States market value using the Consumer Price Index. The costs for all claims during the specific episode durations were added together and then divided by the number of the patients to estimate an average cost of treatment for the surgical treatment among the cohort.
Independent and Dependent Variables
The MarketScan dataset includes clinical and socioeconomic variables that were utilized in univariate and multivariate analyses. Breast reconstruction was classified as tissue expander, direct-to-implant, or autologous reconstruction. Patients who underwent latissimus dorsi reconstruction were included in the autologous reconstruction group. Patient age, median household income in area of residence, and type of insurance plan were compared between women who underwent common approaches to reconstruction. Comorbidity burden was estimated with a modified Elixhauser score.17 The timing and costs related to chemotherapy and radiation therapy were available in the dataset. Major complications were defined as complications that resulted in an unplanned operation, unplanned admission, or reconstruction failure. Balancing procedures were defined as procedures on the unreconstructed contralateral breast to obtain symmetry with the reconstructed breast for patients undergoing unilateral mastectomies.
Statistical Analysis
Baseline clinical and demographic characteristics were compared using chi-square or Fisher’s Exact tests for categorical variables and ANOVA, standard t-tests, or Kruskal-Wallis tests for continuous variables. Multivariable models were built using demographic and clinically relevant variables to assess the impact of major complications and radiation therapy on total costs at different bundled payment episode durations after adjusting for relevant clinical and socioeconomic variables. All statistical analyses were performed in SAS 9.4 (SAS Institute, Cary, NC). All tests were two-sided and a P-value of less than .05 was considered statistically significant. This study was judged to be exempt from human participant review by the institutional review board of the University of Michigan, Ann Arbor.
RESULTS
Among 15,377 women in the analytic cohort, 11,592 (76%) underwent tissue expander reconstruction, 1,279 (8%) underwent direct-to-implant reconstruction, and 2,506 (16%) underwent autologous reconstruction (Table 1). There were 4,066 (26.4%) women who underwent bilateral mastectomies and 969 (6.3%) women with a strong family history of breast cancer or a genetic predisposition for breast cancer. There were 774 (5.0%) women received radiation therapy and 4,459 (28.9%) women who received any chemotherapy during the study period. At one year, most patients had undergone exchange of tissue expanders for permanent implants (76%) and other revision procedures (81%). Acellular dermal matrix (ADM) was utilized in implant-based breast reconstruction for 3,972 (44.6%) women. Although more than 75% of women had Elixhauser comorbidity scores ≤9, there were 3,771 (24.5%) of women with comorbidity scores >9 (Table 1).
Table 1.
Baseline sociodemographic and clinical characteristics of the analytic cohort by procedure type
| Patient Characteristics | Tissue Expander Reconstruction | Direct-to-Implant Reconstruction | Autologous Reconstruction | P-value* |
|---|---|---|---|---|
| Total | 11,592 (75%) | 1,279 (8%) | 2,506 (16%) | |
| Age (years) | ||||
| ≤40 | 1,505 (13%) | 166 (13%) | 242 (10%) | <0.001 |
| 41–50 | 4,191 (36%) | 446 (35%) | 856 (34%) | |
| 51–60 | 4,184 (36%) | 441 (34%) | 1,052 (42%) | |
| >60 | 1,712 (15%) | 226 (18%) | 356 (14%) | |
| Median household income | ||||
| <40,000 | 259 (2%) | 19 (1%) | 45 (2%) | 0.004 |
| 40,000–50,000 | 2,648 (23%) | 268 (21%) | 529 (21%) | |
| 50,000–60,000 | 4,664 (40%) | 514 (40%) | 1,039 (41%) | |
| 60,000–70,000 | 1,976 (17%) | 220 (17%) | 482 (19%) | |
| >70,000 | 665 (6%) | 97 (8%) | 147 (6%) | |
| Missing | 1,380 (12%) | 161 (13%) | 264 (11%) | |
| Type of insurance plan | ||||
| PPO | 7,186 (62%) | 786 (61%) | 1,502 (60%) | <0.001 |
| Comprehensive | 519 (4%) | 60 (5%) | 85 (3%) | |
| HMO | 1,389 (12%) | 139 (11%) | 354 (14%) | |
| POS | 909 (8%) | 145 (11%) | 233 (9%) | |
| Other | 1,260 (11%) | 110 (9%) | 255 (10%) | |
| Missing | 329 (3%) | 39 (3%) | 77 (3%) | |
| Comorbidity score | ||||
| ≤3 | 1,174 (10%) | 162 (13%) | 244 (10%) | <0.001 |
| 4–5 | 7,627 (66%) | 868 (68%) | 1,531 (61%) | |
| 6–9 | 1,688 (15%) | 148 (12%) | 464 (19%) | |
| >9 | 1,103 (10%) | 101 (8%) | 267 (11%) | |
| Region | ||||
| Northeast | 2,441 (21%) | 330 (26%) | 601 (24%) | <0.001 |
| North central | 2,621 (23%) | 253 (20%) | 432 (17%) | |
| South | 4,173 (36%) | 409 (32%) | 1,061 (42%) | |
| West | 2,142 (18%) | 259 (20%) | 368 (15%) | |
| Unknown | 215 (2%) | 28 (2%) | 44 (2%) | |
| Any chemotherapy | ||||
| No | 8,135 (70%) | 988 (77%) | 1,795 (72%) | <0.001 |
| Yes | 3,457 (30%) | 291 (23%) | 711 (28%) | |
| Adjuvant chemotherapy | ||||
| No | 8,592 (74%) | 1,033 (81%) | 1,942 (77%) | <0.001 |
| Yes | 3,000 (26%) | 246 (19%) | 564 (23%) | |
| Neoadjuvant chemotherapy | ||||
| No | 10,773 (93%) | 1,200 (94%) | 2,255 (90%) | <0.001 |
| Yes | 819 (7%) | 79 (6%) | 251 (10%) | |
| Any radiation therapy | ||||
| No | 10,961 (95%) | 1,233 (96%) | 2,409 (96%) | <0.001 |
| Yes | 631 (5%) | 46 (4%) | 97 (4%) | |
| Acellular dermal matrix | ||||
| No | 8,161 (70%) | 738 (58%) | 2,506 (100%) | <0.001 |
| Yes | 3,431 (30%) | 541 (42%) | 0 (0%) | |
| Bilateral mastectomy | ||||
| No | 8,501 (73%) | 909 (71%) | 1,901 (76%) | 0.004 |
| Yes | 3,091 (27%) | 370 (29%) | 605 (24%) |
Chi-square and Fisher’s exact tests used for categorical variables
Values listed as n (%)
The timing and frequency of major complications differed by procedure type. For instance, the rate of major complications within thirty days was 8%, 12%, and 15% for tissue expander, direct-to-implant, and autologous reconstruction, respectively (p<0.001) (Table 2). At one year, the rate of major complications was 25%, 24%, and 28% for tissue expander, direct-to-implant, and autologous reconstruction, respectively (p=0.007). Approximately 86.8% of women who experienced at least one major complication were captured in the one year bundled payment period.
Table 2.
Postoperative outcomes and healthcare resources use by type of reconstruction
| Patient outcomes | Tissue Expander Reconstruction | Direct-to-Implant Reconstruction | Autologous Reconstruction | P-value* |
|---|---|---|---|---|
| Radiation therapy | ||||
| 30 days following index procedure | 49 (0%) | 13 (1%) | 20 (1%) | 0.003 |
| 90 days following index procedure | 286 (2%) | 28 (2%) | 49 (2%) | 0.284 |
| 1 year following index procedure | 616 (5%) | 45 (4%) | 91 (4%) | <0.001 |
| 2 years following index procedure | 631 (5%) | 46 (4%) | 97 (4%) | <0.001 |
| Chemotherapy | ||||
| 30 days following index procedure | 1,274 (11%) | 128 (10%) | 328 (13%) | 0.004 |
| 90 days following index procedure | 3,099 (27%) | 267 (21%) | 635 (25%) | <0.001 |
| 1 year following index procedure | 3,354 (29%) | 286 (22%) | 684 (27%) | <0.001 |
| 2 years following index procedure | 3,457 (30%) | 291 (23%) | 711 (28%) | <0.001 |
| Major complications | ||||
| 30 days following index procedure | 904 (8%) | 152 (12%) | 373 (15%) | <0.001 |
| 90 days following index procedure | 1,518 (13%) | 206 (16%) | 491 (20%) | <0.001 |
| 1 year following index procedure | 2,892 (25%) | 301 (24%) | 692 (28%) | 0.007 |
| 2 years following index procedure | 3,349 (29%) | 340 (27%) | 785 (31%) | 0.006 |
| Planned or unplanned revisions | ||||
| 30 days following index procedure | 932 (8%) | 203 (16%) | 298 (12%) | <0.001 |
| 90 days following index procedure | 2,464 (21%) | 308 (24%) | 482 (19%) | 0.002 |
| 1 year following index procedure | 10,046 (87%) | 687 (54%) | 1,714 (68%) | <0.001 |
| 2 years following index procedure | 10,632 (92%) | 764 (60%) | 1,880 (75%) | <0.001 |
| Exchange following TE | ||||
| 30 days following index procedure | 30 (0%) | 0 (0%) | 0 (0%) | - |
| 90 days following index procedure | 1,000 (9%) | 0 (0%) | 0 (0%) | - |
| 1 year following index procedure | 8,813 (76%) | 0 (0%) | 0 (0%) | - |
| 2 years following index procedure | 9,500 (82%) | 0 (0%) | 0 (0%) | - |
| Balancing procedures | ||||
| 30 days following index procedure | 11 (0%) | 2 (0%) | 2 (0%) | 0.067 |
| 90 days following index procedure | 177 (2%) | 10 (1%) | 33 (1%) | 0.09 |
| 1 year following index procedure | 1,836 (16%) | 88 (7%) | 339 (14%) | <0.001 |
| 2 years following index procedure | 2,081 (18%) | 110 (9%) | 405 (16%) | <0.001 |
TE, Tissue Expander. Values listed as n (%)
Chi-square and Fisher’s exact tests used for categorical variables.
The median costs for the initial surgery (including mastectomy) were approximately $24,623, $24,626, and $38,019 for tissue expander, direct-to-implant, and autologous reconstruction, respectively (Figure 1). Postoperative costs increased significantly between the 90-day bundled payment period and the one-year period, due in part to the occurrence of breast reconstruction revisions, exchange of tissue expanders for permanent implants, and procedures on the contralateral breast for symmetry.
Figure 1.
Preoperative, operative, and postoperative costs for each procedure type and predefined bundled payment period
The costs associated with major complications differed by procedure type and increased throughout all bundled payment periods (Table 3). For instance, median costs for any complications at thirty days were approximately $3,586, $5,148, and $7,197 for tissue expander, direct-to-implant, and autologous reconstruction, respectively (p<0.001). At one year, median costs for major complications increased to approximately $5,751, $6,080, and $10,644 for tissue expander, direct-to-implant, and autologous reconstruction, respectively (p=0.003). Postoperative costs increased across all procedure groups and all bundled payment periods with an increasing burden of patient comorbidities (See Table, Supplemental Digital Content 1, which shows the influence of patient comorbidities on preoperative, operative, and postoperative costs for women who underwent breast reconstruction following mastectomy, http://links.lww.com/PRS/E246).
Table 3.
Unadjusted costs of complications and other reconstruction procedures within bundled payment periods
| 30-Day Bundled Payment | 90-Day Bundled Payment | 1-year Bundled Payment | 2-year Bundled Payment | |
|---|---|---|---|---|
| Complications * , median (IQR) | ||||
| TE | 3,586 (9,612) | 4,337 (10,742) | 5,751 (13,261) | 6,366 (14,012) |
| DTI | 5,148 (13,298) | 5,863 (14,964) | 6,080 (16,925) | 6,642 (17,217) |
| Autologous | 7,197 (20,034) | 9,156 (21,141) | 10,644 (22,941) | 10,959 (23,087) |
| P-value | <0.001 | <0.001 | 0.003 | 0.004 |
| Planned or unplanned revisions, mean (SD) | ||||
| TE | 235 (1,438) | 946 (3,947) | 7,131 (10,457) | 9,113 (13,086) |
| DTI | 441 (2,756) | 825 (3,602) | 4,233 (10,048) | 5,902 (15,244) |
| Autologous | 296 (1,994) | 699 (3,670) | 5,173 (9,228) | 6,443 (10,687) |
| P-value | <0.001 | <0.001 | <0.001 | <0.001 |
| Exchange following TE, mean (SD) | ||||
| TE | 8 (298) | 345 (1,797) | 2,923 (4,360) | 3,288 (4,722) |
| DTI | 0 (0) | 0 (0) | 0 (0) | 0 (0) |
| Autologous | 0 (0) | 0 (0) | 0 (0) | 0 (0) |
| P-value | 0.009 | <0.001 | <0.001 | <0.001 |
| Balancing procedures, mean (SD) | ||||
| TE | 1 (68) | 34 (400) | 400 (1,470) | 477 (1,665) |
| DTI | 0 (9) | 7 (100) | 203 (1,263) | 291 (2,213) |
| Autologous | 1 (50) | 51 (778) | 374 (1,565) | 466 (1,795) |
| P-value | 0.962 | 0.047 | <0.001 | <0.001 |
| Number of planned/unplanned revisions, mean (SD) | ||||
| TE | 0.1 (0.3) | 0.3 (0.6) | 1.6 (1.1) | 2.1 (1.5) |
| DTI | 0.2 (0.4) | 0.3 (0.6) | 0.9 (1.2) | 1.2 (1.5) |
| Autologous | 0.2 (0.7) | 0.3 (0.8) | 1.3 (1.3) | 1.6 (1.6) |
| P-value | <0.001 | 0.007 | <0.001 | <0.001 |
| Total costs † , median (IQR) | ||||
| TE | 0 (0) | 0 (74) | 7,820 (13,082) | 9,851 (15,279) |
| DTI | 0 (0) | 0 (488) | 923 (7,512) | 1,951 (10,352) |
| Autologous | 0 (0) | 0 (214) | 3,509 (10,521) | 4,809 (12,275) |
| P-value | <0.001 | 0.002 | <0.001 | <0.001 |
TE, Tissue Expander; DTI, Direct-to-implant
P-Values calculated with ANOVA test for all variables, except Kruskal-Wallis tests for complications and total costs in skewed data.
Among patients with at least one complication
Excluding costs related to the index procedure
In multivariable models, the occurrence of at least one complication was associated with an increase of approximately $7,051, $9,172, and $7,595 to overall costs during the one year bundled payment period for tissue expander, direct-to-implant, and autologous reconstruction, respectively (Table 4). Adjusted models also demonstrated that radiation therapy contributed approximately $36,300, $27,785, and $22,395 to total costs at one year for tissue expander, direct-to-implant, and autologous reconstruction, respectively (Table 5). Similarly, in adjusted models chemotherapy contributed approximately $40,413, $35,588, and $30,333 to total costs at one year for tissue expander, direct-to-implant, and autologous reconstruction, respectively (Table 5).
Table 4.
Adjusted costs attributed to postoperative complications by procedure type and bundled payment periods
| No complications | Any complication | Adjusted difference | P-value | |
|---|---|---|---|---|
| Tissue expander reconstruction | ||||
| 30 days | 30,542 (18,309) | 33,959 (20,241) | 3,550 (2,277, 4,823) | <0.001 |
| 90 days | 36,360 (22,262) | 40,362 (23,977) | 4,101 (2,868, 5,334) | <0.001 |
| 1 year | 51,283 (37,227) | 58,584 (41,289) | 7,051 (5,549, 8,553) | <0.001 |
| 2 years | 54,433 (40,756) | 66,004 (47,304) | 11,009 (9,410, 12,608) | <0.001 |
| Direct-to-implant reconstruction | ||||
| 30 days | 32,425 (23,587) | 35,200 (23,522) | 2,894 (−942, 6,731) | 0.139 |
| 90 days | 36,727 (25,946) | 39,792 (25,777) | 3,422 (−355, 7,201) | 0.076 |
| 1 year | 45,323 (35,072) | 54,433 (41,552) | 9,172 (4,554, 13,789) | <0.001 |
| 2 years | 48,105 (38,434) | 59,770 (46,732) | 11,711 (6,828, 16,594) | <0.001 |
| Autologous reconstruction | ||||
| 30 days | 46,643 (31,120) | 51,586 (30,780) | 5,037 (1,496, 8,578) | 0.005 |
| 90 days | 51,347 (32,954) | 55,305 (30,642) | 4,430 (1,170, 7,691) | 0.008 |
| 1 year | 63,215 (42,831) | 69,949 (43,175) | 7,595 (3,821, 11,368) | <0.001 |
| 2 years | 65,809 (44,813) | 75,680 (48,387) | 11,245 (7,363, 15,126) | <0.001 |
Values listed as mean (standard deviation), with exception of adjusted difference, which is listed as the adjusted mean difference (95% confidence interval). Adjusted difference calculated from generalized linear regression model with a log link and gamma distribution using whether or not a complication occurred during each period as the primary predictor, controlling for patient age, median household income in the area resided, type of insurance and score of comorbidity.
Table 5.
Comparison of postoperative costs for women who received adjuvant therapies by each procedure type for each bundled payment period
| No radiation or chemotherapy | Both radiation and chemotherapy | Adjusted difference | P-value | |
|---|---|---|---|---|
| Tissue expander reconstruction | ||||
| One year | 41,006 (25,179) | 80,813 (47,953) | 39,978 (38,389, 41,568) | <0.001 |
| Two years | 43,832 (27,315) | 88,422 (54,094) | 44,467 (42,743, 46,190) | <0.001 |
| Direct-to-implant reconstruction | ||||
| One year | 39,398 (28,176) | 73,569 (48,039) | 34,365 (28,766, 39,964) | <0.001 |
| Two years | 42,004 (31,317) | 80,343 (53,155) | 38,354 (32,258, 44,451) | <0.001 |
| Autologous reconstruction | ||||
| One year | 56,288 (36,206) | 86,912 (50,272) | 29,226 (25,082, 33,370) | <0.001 |
| Two years | 58,572 (37,196) | 93,131 (55,277) | 33,184 (28,865, 37,502) | <0.001 |
| No radiation therapy | Radiation therapy | Adjusted difference | P-value | |
| Tissue expander reconstruction | ||||
| One year | 51,094 (36,951) | 88,933 (45,607) | 36,300 (31,980, 40,620) | <0.001 |
| Two years | 55,366 (41,138) | 99,647 (53,209) | 42,538 (37,700, 47,375) | <0.001 |
| Direct-to-implant reconstruction | ||||
| One year | 46,584 (36,191) | 71,674 (46,846) | 27,785 (10,431, 39,139) | <0.001 |
| Two years | 50,263 (40,399) | 76,499 (51,341) | 26,183 (10,748, 41,617) | <0.001 |
| Autologous reconstruction | ||||
| One year | 64,350 (42,513) | 84,296 (51,507) | 22,395 (11,170, 33,619) | <0.001 |
| Two years | 67,914 (45,401) | 93,420 (57,557) | 28,039 (16,019, 40,059) | <0.001 |
| No chemotherapy | Chemotherapy | Adjusted difference | P-value | |
| Tissue expander reconstruction | ||||
| One year | 41,443 (25,467) | 81,747 (48,496) | 40,413 (38,768, 42,057) | <0.001 |
| Two years | 44,318 (27,640) | 89,447 (54,715) | 44,917 (43,135, 46,700) | <0.001 |
| Direct-to-implant reconstruction | ||||
| One year | 39,532 (28,332) | 75,017 (48,229) | 35,588 (29,773, 41,403) | <0.001 |
| Two years | 42,159 (31,472) | 81,924 (53,394) | 39,670 (33,342, 45,998) | <0.001 |
| Autologous reconstruction | ||||
| One year | 56,413 (36,511) | 88,146 (50,013) | 30,333 (26,051, 34,614) | <0.001 |
| Two years | 58,767 (37,581) | 94,487 (55,140) | 34,328 (29,862, 38,794) | <0.001 |
Values listed as mean (standard deviation), with exception of adjusted difference, which is listed as the adjusted mean difference (95% confidence interval). Adjusted difference calculated from generalized linear regression model with a log link and gamma distribution using whether or not received radiation and/or chemotherapy during each period as the primary predictor, controlling for patient age, median household income in the area resided, type of insurance and score of comorbidity.
DISCUSSION
Implementation of payment reform for breast reconstruction following mastectomy demands a comprehensive understanding of costs related to the process of breast reconstruction. Broad variations in the procedural and complication-related costs for reconstruction in our study highlight an opportunity for bundled payment plans to incentivize quality improvement and value in breast reconstruction. However, the contribution of adjuvant therapies and patient comorbidities to postoperative costs, the number of distinct pathways available to women considering reconstruction, and the well-known association of radiation therapy with increased complications underscores the challenges of disentangling costs for breast reconstruction from the current practice of multidisciplinary breast care. Without appropriate consideration for the nuances of breast reconstruction, bundled payments may reduce reimbursement to plastic surgeons, exacerbate existing disparities in access to reconstruction, and limit the number of options available to women considering these procedures following mastectomy.
Although specific details of how bundling would apply to breast reconstruction remain unclear, our findings shed some light on the appropriateness of a one-year time horizon. At one year postoperatively, the majority of patients have undergone tissue expander exchange, revisions, and a majority of postoperative complications have occurred. This observation is further supported in a study by Fischer et al showing that the number of secondary breast procedures and total charges converge at one year following immediate reconstruction.18 Future research is needed to determine how adjuvant therapies affect timing of procedures to complete reconstruction. In choosing a time period for bundling, accommodations would likely have to be made for patients undergoing adjuvant therapy to account for the delays in further reconstructive procedures in these patients.
Our study also highlights the contribution of patient comorbidities to increased costs in the postoperative period. These findings suggest that bundled payment models should consider adjusting for comorbidities in determining payments for the provision of these services. Without adjustment for comorbidities, bundling may inadvertently create disincentives to surgeons from performing reconstruction on patients at increased risk for complications. Similarly, radiation therapy and chemotherapy contribute substantially to costs within the first-year following initiation of reconstruction and thus, attempts should be made to separate these services from bundles for reconstruction completely.
In previous examples of bundled payment programs, voluntary participation by hospitals have demonstrated either unchanged or reduced spending without altering the quality of care delivered.19,20 More recently, a large multi-center randomized controlled trial of the Medicare Comprehensive Care for Joint Replacement bundling program demonstrated a modest reduction of spending for hip or knee replacement procedures at participating centers.21 The observed reduction in spending was primarily due to decreased used of post-acute care services. Similar trends in reduced post-acute care service utilization from bundling have been noted by other studies of bundling and alternative payment models.19,22–25 Few women are discharged requiring post-acute care services following breast reconstruction. Therefore, the mechanism by which bundled payments may reduce costs for breast reconstruction without compromising on care quality is not clear. It is possible, however, that bundling may incentive cost reduction related to use of materials and resources such as acellular dermal matrices in implant-based reconstruction, preoperative imaging, and postoperative intensive care unit stays following autologous reconstruction. Improvement in clinical outcomes from preoperative angiography for autologous reconstruction and use of acellular dermal matrices in implant-based reconstruction continue to be debated.26–29
Although it is clear that efforts to improve value are necessary, payment reform should not discourage continued innovations in breast reconstruction. These reconstructive procedures are now often performed with autologous tissue or a variety of implant-based approaches, depending on patient preferences and the clinical and anatomic challenges individual to each patient.16 Innovations in surgical technique, such as autologous reconstruction using abdominal, thigh, or gluteal tissues, have increased the options available to women who may not have otherwise considered reconstruction. Furthermore, plastic surgeons have refined implant and tissue expander technologies and autologous fat grafting techniques to achieve natural-appearing breast reconstruction outcomes. These different approaches to reconstruction vary in complexity and accordingly vary in the time and effort necessary for delivery of surgical care. With careful consideration, payment strategies can be implemented to ensure that some of the more complicated options for autologous reconstructions can still be provided to women who stand to benefit from them in the long-term.30
Although bundling has been spearheaded by CMS, commercial insurers and private sector health systems have adopted similar and other emerging payment reforms as well. Pilot studies assessing the impact of bundling or other alternative payment reforms on cost containment for breast reconstruction and the potential unintended consequences for quality and access to these services must be performed prior to widespread implementation. Condition-specific bundling (i.e. for breast cancer, rather than breast reconstruction) and payment reform may impact the reimbursement received by plastic surgeons most significantly since the reconstruction typically occurs after treatment has concluded. Plastic surgeons would likely be the final group to receive compensation for services among other multidisciplinary providers in a condition-specific bundling scenario. For providers in small group or individual practices, negotiating a portion of reimbursement from the bundled payment may pose significant administrative and economic challenges. As a result, these providers may be less likely to offer these procedures for women with insurances that participate in bundled payments. Thus, it should not be surprising that plastic surgeons have already begun to raise concerns that bundling may decrease opportunities for continued innovations in the field and exacerbate existing disparities in access to reconstruction.31–34
Although this study benefitted from a large sample size and the database included comprehensive cost information for the process of breast reconstruction, it has limitations inherent to observational data. Our payment models are based upon ICD-9 coding methodologies, which may lack clinical granularity needed for robust risk-adjustment in multivariable models. Furthermore, the sample is primarily comprised of claims from commercial insurers to represent the majority of women who undergo breast reconstruction who are insured; however, this cohort may not be generalizable to publically insured populations. It is important to note that although autologous reconstruction was associated with increased costs in this study, there are likely important cost-savings advantages to this approach relative to implants over longer time periods. Finally, it remains possible that missing or incorrect variables may explain some of the observed associations.
The variations in procedural and complication-related costs for breast reconstruction in our study highlight a potential opportunity for bundled payment plans to incentivize quality improvement and value in breast reconstruction. However, the contribution of adjuvant therapies and patient comorbidities to postoperative costs, the number of distinct pathways available to women considering reconstruction, and the well-known association of radiation therapy with increased complications underscores the challenges of disentangling costs for breast reconstruction from the current practice of multidisciplinary breast care. Given the complexity of reconstruction pathways, our findings suggest that a one-year episode time horizon may optimally capture reconstruction events and complications. If bundling were to be applied to reconstruction, consideration should be given to separate bundles for medical and surgical care with adjustment for procedure type, patient comorbidities, and cancer therapies to limit the adverse impact of bundled payments on access to reconstruction.
Supplementary Material
See Table, which shows the influence of patient comorbidities on preoperative, operative, and postoperative costs for women who underwent breast reconstruction following mastectomy, http://links.lww.com/PRS/E246.
Funding/Support:
Support for this study was provided by the Midcareer Investigator Award in Patient-Oriented Research (K24-AR053120-06) to KCC. This work is also supported by the Chang Gung Memorial Hospital (projects CORPG3G0111, CORPG3G0161, and CORPG3H0071) to KCC. Chang Gung Memorial Hospital provided analysis assistance and support from the Maintenance project of the Center for Artificial Intelligence Research in Medicine (grant CIRPG3H0021).
Footnotes
Financial Disclosure Statement: The authors have no financial disclosures.
Presented at: American Association of Plastic Surgeons Annual Meeting in April 2019 (Baltimore, MD)
REFERENCES
- 1.Porter ME, Lee TH. From Volume to Value in Health Care: The Work Begins. JAMA. 2016;316(10):1047–1048. [DOI] [PubMed] [Google Scholar]
- 2.Burwell SM. Setting value-based payment goals--HHS efforts to improve U.S. health care. N Engl J Med. 2015;372(10):897–899. [DOI] [PubMed] [Google Scholar]
- 3.Press MJ, Rajkumar R, Conway PH. Medicare’s New Bundled Payments: Design, Strategy, and Evolution. JAMA. 2016;315(2):131–132. [DOI] [PubMed] [Google Scholar]
- 4.Alphs Jackson H, Walsh B, Abecassis M. A Surgeon’s Guide to Bundled Payment Models for Episodes of Care. JAMA Surg. 2016;151(1):3–4. [DOI] [PubMed] [Google Scholar]
- 5.American College of Surgeons. ACS release primer on bundled payment. Bulletin of the American College of Surgeons. 2013; http://bulletin.facs.org/2013/06/acs-releases-primer-on-bundled-payment/. Accessed Feb 1, 2019. [Google Scholar]
- 6.American Society of Plastic Surgeons. Plastic Surgeons Statistics Report. 2017; https://www.plasticsurgery.org/documents/News/Statistics/2017/plastic-surgery-statistics-full-report-2017.pdf. Accessed Feb 1, 2019.
- 7.Berlin NL, Tandon VJ, Qi J, Hamill JB, Kim HM, Momoh AO, Pusic AL, Wilkins EG. Hospital Variations in Clinical Complications and Patient Reported Outcomes at Two Years Following Immediate Breast Reconstruction. Ann Surg. 2018;[published online February 27, 2018]. [DOI] [PubMed] [Google Scholar]
- 8.Bennett KG, Qi J, Kim HM, Hamill JB, Pusic AL, Wilkins EG. Comparison of 2-Year Complication Rates Among Common Techniques for Postmastectomy Breast Reconstruction. JAMA Surg. 2018;153(10):901–908. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Wilkins EG, Hamill JB, Kim HM, et al. Complications in Postmastectomy Breast Reconstruction: One-year Outcomes of the Mastectomy Reconstruction Outcomes Consortium (MROC) Study. Ann Surg. 2018;267(1):164–170. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Fischer JP, Wes AM, Nelson JA, et al. Propensity-matched, longitudinal outcomes analysis of complications and cost: comparing abdominal free flaps and implant-based breast reconstruction. J Am Coll Surg. 2014;219(2):303–312. [DOI] [PubMed] [Google Scholar]
- 11.Chetta MD, Aliu O, Zhong L, et al. Reconstruction of the Irradiated Breast: A National Claims-Based Assessment of Postoperative Morbidity. Plast Reconstr Surg. 2017;139(4):783–792. . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Sheckter CC, Razdan SN, Disa JJ, Mehrara BJ, Matros E. Conceptual Considerations for Payment Bundling in Breast Reconstruction. Plast Reconstr Surg. 2018;141(2):294–300. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Rastogi A, Mohr BA, Williams JO, Soobader MJ, de Brantes F. Prometheus payment model: application to hip and knee replacement surgery. Clin Orthop Relat Res. 2009;467(10):2587–2597. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Spinks T, Guzman A, Beadle BM, et al. Development and Feasibility of Bundled Payments for the Multidisciplinary Treatment of Head and Neck Cancer: A Pilot Program. J Oncol Pract. 2018;14(2):e103–e112. [DOI] [PubMed] [Google Scholar]
- 15.PWC. Strategy & Annual Bundles Survey. 2015–2016; https://www.strategyand.pwc.com/media/file/Annual-Bundles-Survey-2015-2016-Results.pdf. Accessed Feb 1, 2019.
- 16.Panchal H, Matros E. Current Trends in Postmastectomy Breast Reconstruction. Plast Reconstr Surg. 2017;140(5S Advances in Breast Reconstruction):7S–13S. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.van Walraven C, Austin PC, Jennings A, Quan H, Forster AJ. A modification of the Elixhauser comorbidity measures into a point system for hospital death using administrative data. Med Care. 2009;47(6):626–633. [DOI] [PubMed] [Google Scholar]
- 18.Fischer JP, Fox JP, Nelson JA, Kovach SJ, Serletti JM. A Longitudinal Assessment of Outcomes and Healthcare Resource Utilization After Immediate Breast Reconstruction-Comparing Implant- and Autologous-based Breast Reconstruction. Ann Surg. 2015;262(4):692–699. [DOI] [PubMed] [Google Scholar]
- 19.Dummit LA, Kahvecioglu D, Marrufo G, et al. Association Between Hospital Participation in a Medicare Bundled Payment Initiative and Payments and Quality Outcomes for Lower Extremity Joint Replacement Episodes. JAMA. 2016;316(12):1267–1278. [DOI] [PubMed] [Google Scholar]
- 20.Chen LM, Ryan AM, Shih T, Thumma JR, Dimick JB. Medicare’s Acute Care Episode Demonstration: Effects of Bundled Payments on Costs and Quality of Surgical Care. Health Serv Res. 2018;53(2):632–648. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Barnett ML, Wilcock A, McWilliams JM, et al. Two-Year Evaluation of Mandatory Bundled Payments for Joint Replacement. N Engl J Med. 2019;380(3):252–262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Finkelstein A, Ji Y, Mahoney N, Skinner J. Mandatory Medicare Bundled Payment Program for Lower Extremity Joint Replacement and Discharge to Institutional Postacute Care: Interim Analysis of the First Year of a 5-Year Randomized Trial. JAMA. 2018;320(9):892–900. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Navathe AS, Troxel AB, Liao JM, et al. Cost of Joint Replacement Using Bundled Payment Models. JAMA Intern Med. 2017;177(2):214–222. [DOI] [PubMed] [Google Scholar]
- 24.Tsai TC, Joynt KE, Wild RC, Orav EJ, Jha AK. Medicare’s Bundled Payment initiative: most hospitals are focused on a few high-volume conditions. Health Aff (Millwood). 2015;34(3):371–380. [DOI] [PubMed] [Google Scholar]
- 25.Jubelt LE, Goldfeld KS, Chung WY, Blecker SB, Horwitz LI. Changes in Discharge Location and Readmission Rates Under Medicare Bundled Payment. JAMA Intern Med. 2016;176(1):115–117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Sorkin M, Qi J, Kim HM, et al. Acellular Dermal Matrix in Immediate Expander/Implant Breast Reconstruction: A Multicenter Assessment of Risks and Benefits. Plast Reconstr Surg. 2017;140(6):1091–1100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Potter S, Browning D, Savovic J, Holcombe C, Blazeby JM. Systematic review and critical appraisal of the impact of acellular dermal matrix use on the outcomes of implant-based breast reconstruction. Br J Surg. 2015;102(9):1010–1025. [DOI] [PubMed] [Google Scholar]
- 28.Hallberg H, Rafnsdottir S, Selvaggi G, et al. Benefits and risks with acellular dermal matrix (ADM) and mesh support in immediate breast reconstruction: a systematic review and meta-analysis. J Plast Surg Hand Surg. 2018;52(3):130–147. [DOI] [PubMed] [Google Scholar]
- 29.Berlin NL, Wilkins EG. Commentary on: Outcomes of Acellular Dermal Matrix for Immediate Tissue Expander Reconstruction with Radiotherapy: A Retrospective Cohort Study. Aesthet Surg J. 2019;39(3):289–291. [DOI] [PubMed] [Google Scholar]
- 30.Pusic AL, Matros E, Fine N, et al. Patient-Reported Outcomes 1 Year After Immediate Breast Reconstruction: Results of the Mastectomy Reconstruction Outcomes Consortium Study. J Clin Oncol. 2017;35(22):2499–2506. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Berlin NL, Wilkins EG, Alderman AK. Addressing Continued Disparities in Access to Breast Reconstruction on the 20th Anniversary of the Women’s Health and Cancer Rights Act. JAMA Surg. 2018;153(7):603–604. [DOI] [PubMed] [Google Scholar]
- 32.Alderman AK, Atisha D, Streu R, et al. Patterns and correlates of postmastectomy breast reconstruction by U.S. Plastic surgeons: results from a national survey. Plast Reconstr Surg. 2011;127(5):1796–1803. [DOI] [PubMed] [Google Scholar]
- 33.Alderman AK, Hawley ST, Janz NK, et al. Racial and ethnic disparities in the use of postmastectomy breast reconstruction: results from a population- based study. J Clin Oncol. 2009;27(32):5325–5330. . [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Alderman AK, McMahon L Jr., Wilkins EG. The national utilization of immediate and early delayed breast reconstruction and the effect of sociodemographic factors. Plast Reconstr Surg. 2003;111(2):695–703. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
See Table, which shows the influence of patient comorbidities on preoperative, operative, and postoperative costs for women who underwent breast reconstruction following mastectomy, http://links.lww.com/PRS/E246.