Abstract
Background.
Black women are more likely to die of breast cancer than White women. This study evaluated the contribution of time to primary surgical management and surgical facility characteristics to racial disparities in breast cancer mortality among both Black and White women.
Methods.
The study identified 2224 Black and 3787 White women with a diagnosis with stages I to III breast cancer (2010–2014). Outcomes included time to surgical treatment (> 30 days from diagnosis) and breast cancer mortality. Odds ratios (ORs) and 95% confidence intervals (CIs) associating surgical facility characteristics with surgical delay were computed, and Cox proportional hazards regression was used to compute hazard ratios (HRs) and 95% CIs associating delay and facility characteristics with breast cancer mortality.
Results.
Black women were two times more likely to have a surgical delay (OR, 2.15; 95% CI, 1.92–2.41) than White women. Racial disparity in surgical delay was least pronounced among women treated at a non-profit facility (OR, 1.95; 95% CI, 1.70–2.25). The estimated mortality rate for Black women was two times that for White women (HR, 2.00; 95% CI, 1.83–2.46). Racial disparities in breast cancer mortality were least pronounced among women who experienced no surgical delay (HR, 1.81; 95% CI, 1.28–2.56), received surgery at a government facility (HR, 1.31; 95% CI, 0.76–2.27), or underwent treatment at a Commission on Cancer-accredited facility (HR, 1.82; 95% CI, 1.38–2.40).
Conclusions.
Black women were more likely to experience a surgical delay and breast cancer death. Persistent racial disparities in breast cancer mortality were observed across facility characteristics except for government facilities.
In the United States, despite similar screening prevalence and incidence rates of breast cancer,1Black women experience persistent disparities in breast cancer mortality compared with White women.1–4 These racial disparities in mortality are most pronounced for breast cancers with effective treatment regimens, such as early-stage or estrogen receptor-positive disease.5 The observed disparities are complex and multifactorial, but the potential mechanisms underlying these inequities may include differential access to and quality of care. Black women are more likely to experience delays in receiving their initial breast cancer diagnosis6–8 and more likely to experience delays in receipt of surgery after diagnosis.7,9–11 Delays in time to surgical treatment have been associated with poor postoperative outcomes, including reduced survival.9,12 This association is more pronounced among Black women.9
Delay in upfront surgery may vary by characteristics of the treating facility, but has been explored limitedly as a potential explanation of racial disparities in breast cancer mortality. One previous study reported that community hospitals had shorter surgical delays, whereas National Cancer Institute (NCI)-designed Comprehensive Cancer Centers and Veterans Administration hospitals had longer wait times.11
In addition to surgical delay, other facility characteristics related to quality of care have been associated with breast cancer outcomes, including surgical volume, a possible proxy for institutional capacity and more comprehensive care, availability of specialty and health services, and breast cancer-related accreditations.13,14 Black women also are less likely to receive care at higher-quality hospitals,14,15 which may contribute to differences in mortality. However, it remains unclear what role facility characteristics play in time to surgical treatment and whether delays in upfront surgery contribute to overall outcomes or outcome disparities.
This study was the first to evaluate associations between facility characteristics, time to surgical treatment, and breast cancer mortality overall and the racial disparities of the associations.
METHODS
Study Population
Using the Georgia Cancer Registry (GCR), we identified women with breast cancer diagnosed between 2010 and 2014 while residing in metropolitan Atlanta. Breast cancer patients were included if they had a diagnosis of stages I to III primary breast tumor and were classified as non-Hispanic Black (NHB) or non-Hispanic White (NHW). Race was abstracted from the medical abstracts pertaining to the breast cancer diagnosis by Certified Tumor Registrars based on U.S. Census Bureau definitions, and Hispanic ethnicity was determined by the North American Association of Central Cancer Registries Hispanic Identification Algorithm.16 Patients who did not receive surgery whose surgery date was equal to the date of diagnosis, those who underwent neoadjuvant therapy, and those whose surgical facility could not be identified were excluded from the study.
Exposure Assessment: Facility Characteristics
Characteristics of the facility where each patient received breast surgery were obtained from the NCI Hospital File for the same year the patient received surgery.17 Information in the NCI Hospital Files is derived from several sources, including the Healthcare Cost Report and the Provider of Service survey conducted by the Center for Medicare and Medicaid Services.18 The GCR captures the facility performing the patient surgery. The GCR personnel linked facility variables to the patient-level dataset, keeping facility names and locations confidential.
The facility characteristics of interest were annual patient volume (high vs low), facility type (non-profit, proprietary, government), medical school affiliation, accreditation by the Commission on Cancer (COC), and participation in the National Surgical Adjuvant Breast and Bowel Project (NSABP) as of 2005. Accreditations that applied to only one hospital (e.g., NCI comprehensive center status) were not included in the analysis. Annual patient volume was defined based on the breast cancer surgical counts for the state of Georgia for the study period and facilities were categorized as high- versus low-volume facilities based on a cut point at the 10th percentile of the distribution of breast cancer surgeries. The top 10% of facilities performing breast cancer surgeries accounted for approximately 50% of all the patients.
Outcome Assessment
The primary outcomes of interest were time to upfront surgery and breast cancer mortality. Surgical delay was defined as surgery received more than 30 days after diagnosis versus 30 days or less after diagnosis. This cutoff was determined based on literature review and clinical input.19 However, because no standard exists currently, we also defined surgical delay as more than 60 days versus 60 days or less after diagnosis in a sensitivity analysis for comparability with prior studies. The date of initial surgery and the date of diagnosis were available from the GCR. Breast cancer mortality (ICD10-C50) was determined from death certificate data. The GCR links annually with both state vital records and the National Death Index. The follow-up period was defined as time in months from the date of surgery until death, last date of contact in registry, 31 December 2019, or end of the study follow-up period.
Covariates
We collected information on age at diagnosis (continuous), stage (I–III), type of breast cancer surgery (breast-conserving surgery, mastectomy, mastectomy with reconstruction or removal of the contralateral breast), derived breast cancer subtype based on hormone receptor (HR) and human epidermal growth factor receptor-2 (HER2) expression (HR+/HER2−, HR+/HER2+, HR−/HER2+, or HR−/HER2−), marital status (single, married–including domestic partner, divorced/separated/other), insurance status (private, Medicare, Medicaid, military/other, or uninsured), and a Census-derived area-based measure of socioeconomic status [SES] (0 % to <5 %, 5 % to <10 %, 10 % to <20 %, 20 % to 100 % below poverty). The SES was based on Census tract-level poverty data published annually from the American Community Survey.20,21
Statistical Methods
Descriptive statistics were calculated as mean values with standard deviation or frequency and percentage for covariates of interest across categories of surgical delay. We additionally report descriptive statistics across categories of race and ethnicity. We used multivariable-adjusted logistic regression models to compute the odds ratios (ORs) and 95% confidence intervals (CIs) associating facility characteristics with surgical delay and multivariable-adjusted Cox proportional hazards regression to compute the hazard ratios (HRs) and 95% CIs associating surgical delay and facility characteristics with breast cancer mortality.
We assessed the impact of racial disparities on surgical delay and breast cancer mortality and whether these associations were modified by facility characteristics. Interaction describes differences in the effect of one exposure across the strata of another exposure, which depends on the scale. The presence of interaction between race and facility characteristics and its effect on surgical delay was estimated with the common referent approach to calculate the relative excess risk due to interaction (RERI), evaluating the departure of the effect on the additive scale.22,23
We computed the 95% CI for the RERI using the delta method.24 The presence of multiplicative interactions, indicating whether the combined effect of race and facility characteristics or surgical delay is greater than the product of the individual effects, was assessed by comparing stratum-specific effect estimates.22,25 To evaluate the contribution of delay and facility characteristics to racial disparities in breast cancer outcomes, we considered the distribution of delay and facility characteristics among NHB and NHW women, as well as the presence of additive/multiplicative interaction.26
Potential confounders included in the models were based on a priori knowledge and graphic-based methods (DAG).27 For the association between facility characteristics and surgical delay, the potential confounders were age at diagnosis and type of breast cancer surgery. For the association between facility characteristics and breast cancer mortality, the confounders were age at diagnosis, disease stage, tumor subtype, SES, and insurance status. In a sensitivity analysis, we excluded women whose breast cancer diagnosis was the first of two or more primaries (n = 623, 10 %). No hypothesis testing was performed.28,29 All analyses were conducted using SAS v9.4 (SAS, Cary, NC, USA).
RESULTS
Among the 6011 breast cancer patients in the study, 3516 (58%) experienced a delay of more than 30 days to primary surgical management (Table 1). NHB women were more likely to experience a delay than NHW women (69% vs 52%, respectively). Women who received mastectomy or mastectomy with reconstruction/removal of the contralateral breast as part of their surgical management were more likely to have a delay in surgery (64% and 74% respectively) than the women who received breast-conserving surgery (51%). Surgical delay also was more common among the patients who received surgery at non-profit hospitals (63%), facilities with a medical school affiliation (64%), and facilities without COC accreditation (61%). Surgical delay was least likely among breast cancer patients treated at a government facility (41%).
TABLE 1.
Patient demographic, tumor, and facility characteristics by surgical delay among 6011 patients with breast cancer diagnosed in the metropolitan Atlanta area (2010–2014)
| ≤ 30 Days from diagnosis 2495 (42 %) | > 30 Days from diagnosis 3516 (58 %) | |
|---|---|---|
|
| ||
| Patient characteristics | Mean ± SD | Mean ± SD |
| Age at diagnosis (years) | 60 ± 13 | 59 ± 12 |
| Follow-up (years) | 7.0 ± 2.2 | 6.7 ± 2.1 |
| n (%) | n (%) | |
| Race | ||
| White | 1808 (48) | 1979 (52) |
| Black | 687 (31) | 1537 (69) |
| Breast cancer-specific death | 136 (5.5) | 226 (6.4) |
| Insurance type | ||
| Private | 1498 (41) | 2193 (59) |
| Uninsured | 30 (33) | 60 (67) |
| Medicaid | 130 (38) | 212 (55) |
| Medicare | 788 (45) | 964 (55) |
| Military | 21 (34) | 41 (66) |
| Unknown | 28 (40) | 42 (60) |
| Socioeconomic status | ||
| 0 to < 5% | 603 (45) | 741 (55) |
| 5 to < 10% | 674 (45) | 833 (55) |
| 10 to < 20% | 734 (43) | 989 (57) |
| 20%+ | 484 (36) | 953 (66) |
| Cancer stage | ||
| I | 1609 (44) | 2083 (56) |
| II | 726 (38) | 1168 (62) |
| III | 160 (38) | 265 (62) |
| ER status | ||
| ER+ | 2140 (41) | 3045 (59) |
| ER− | 335 (42) | 460 (58) |
| Unknown | 20 | 11 |
| Tumor molecular subtype | ||
| Luminal A | 1782 (41) | 2566 (59) |
| Luminal B | 234 (39) | 361 (61) |
| HER2 | 70 (40) | 105 (60) |
| TNBC | 223 (42) | 302 (58) |
| Unknown | 186 | 182 |
| Surgery type | ||
| Conserving | 1714 (49) | 1752 (51) |
| Mastectomy | 376 (36) | 681 (64) |
| Mastectomy with reconstruction or removal of the contralateral breast | 339 (26) | 964 (74) |
| Facility characteristics | ||
| Annual patient discharge | ||
| High | 1994 (42) | 2796 (58) |
| Low | 501 (41) | 720 (59) |
| Medical school affiliation | ||
| Yes | 580 (37) | 1009 (64) |
| No | 1911 (43) | 2498 (57) |
| Unknown | 4 | 9 |
| Facility type | ||
| Non-profit | 1639 (37) | 2750 (63) |
| For-profit | 400 (47) | 448 (53) |
| Government | 452 (59) | 309 (41) |
| Unknown | 4 | 9 |
| Accreditations or affiliations | ||
| COC-accredited | ||
| Yes | 1587 (43) | 2093 (57) |
| No | 908 (39) | 1423 (61) |
| NSABP affiliation | ||
| Yes | 1894 (41) | 2750 (59) |
| No | 601 (44) | 766 (56) |
SD, standard deviation; ER, estrogen receptor; HER2, human epidermal growth factor receptor-2; TNBC, triple-negative breast cancer; COC, Commission on Cancer; NSABP, National Surgical Adjuvant Breast and Bowel Project
Table 2 presents the distribution of facility characteristics by race. In the total study population, 3787 (63%) of the patients were NHW and 2224 (37%) were NHB. The NHB women were less likely to receive surgery at a high-volume facility (69% vs 85%), a non-profit facility (68% vs 76%), a COC-accredited facility (50% vs 68%), or an NSABP-affiliated facility (70% vs 82%) than the NHW women. However, the NHB women were more likely to receive surgery at a facility with a medical school affiliation (37% vs 21%) than the NHW women.
TABLE 2.
Patient demographic, time-to-surgery, and facility characteristics by race among 6011 patients with breast cancer diagnosed in the metropolitan Atlanta area (2010–2014)
| NHW | NHB | |
|---|---|---|
|
| ||
| 3787 (63 %) | 2224 (37 %) | |
| Patient characteristics | Mean ± SD | Mean ± SD |
| Age at diagnosis (years) | 60 ± 13 | 57 ± 12 |
| Follow-up (years) | 6.9 ± 2.1 | 6.8 ± 2.2 |
| n (%) | n (%) | |
| Breast cancer-specific death | 173 (4.6) | 189 (3.1) |
| Surgical delay (days) | ||
| Delay (>30) | 1979 (52) | 1537 (69) |
| No delay (≤30) | 1808 (48) | 687 (31) |
| Surgical delay | ||
| Delay (>60) | 448 (12) | 553 (25) |
| No delay (≤60) | 3339 (88) | 1671 (75) |
| Surgery type | ||
| Conserving | 2173 (59) | 1293 (60) |
| Mastectomy | 614 (17) | 443 (21) |
| Mastectomy with reconstruction or removal of the contralateral breast | 885 (24) | 418 (19) |
| Facility characteristics | ||
| Annual patient discharge | ||
| High | 3266 (86) | 1524 (69) |
| Low | 521 (14) | 700 (31) |
| Medical school affiliation | ||
| Yes | 779 (21) | 810 (37) |
| No | 3008 (79) | 1401 (63) |
| Facility type | ||
| Non-profit | 2882 (76) | 1507 (68) |
| For-profit | 454 (12) | 394 (18) |
| Government | 451 (12) | 310 (14) |
| Unknown | 13 (0.6) | |
| Accreditations or affiliations | ||
| COC-accredited | ||
| Yes | 2572 (68) | 1108 (50) |
| No | 1215 (32) | 1116 (50) |
| NSABP affiliation | ||
| Yes | 3091 (82) | 1553 (70) |
| No | 696 (18) | 671 (30) |
NHW, non-Hispanic White; NHB, non-Hispanic Black; SD, standard deviation; COC, Commission on Cancer; NSABP, National Surgical Adjuvant Breast and Bowel Project
Facility Characteristics and Surgical Delay
Table 3 presents the multivariable-adjusted associations between facility characteristics and surgical delay overall as well as racial disparities in surgical delay by facility characteristics. Compared with the breast cancer patients treated at non-profit facilities, surgical delay was less likely among those treated at government facilities (OR, 0.40; 95% CI, 0.34–0.47) and those treated at for-profit centers (OR, 0.71; 95% CI, 0.61–0.82). The patients who received surgical management at a non-COC-accredited facility were 1.24 times more likely to experience a delay in surgery than the breast cancer patients who received surgery at a COC-accredited facility (95% CI, 1.11–1.38).
TABLE 3.
Multivariable-adjusted odds ratios associating facility characteristics with surgical treatment delays overall, and racial disparities in delay by facility characteristics among 6011 women with breast cancer diagnosed in the metropolitan Atlanta area (2010–2014)
| Delay (n) |
Overall OR (95 % CI)1 |
Delay n (%) |
Common referent OR (95 % CI)1 |
Stratified effects OR (95 % CI)a |
||||
|---|---|---|---|---|---|---|---|---|
|
|
|
|||||||
| Overall | NHW | NHB | NHW | NHB | RERI | |||
|
| ||||||||
| Racial disparity | 1979 | 1537 | – | – | – | 2.15 (1.92–2.41) | ||
| Annual patient discharge | ||||||||
| Low | 2796 | 1.06 (0.93 to 1.21) | 236 | 484 | 0.73 (0.60 to 0.89) | 2.11 (1.76 to 2.53) | 0.32 (−0.11 to 0.75) | 2.89 (2.25 to 3.70) |
| High | 720 | Referent | 1743 | 1053 | Referent | 2.04 (1.79 to 2.33) | Referent | 2.04 (1.79 to 2.33) |
| Medical school affiliation | ||||||||
| No | 2498 | 0.72 (0.64 to 0.81) | 1541 | 957 | 0.79 (0.67 to 0.93) | 1.70 (1.41 to 2.40) | −0.06 (−0.45 to 0.34) | 2.15 (1.87 to 2.48) |
| Yes | 1009 | Referent | 438 | 571 | Referent | 1.94 (1.57 to 2.40) | Referent | 1.94 (1.57 to 2.40) |
| Facility type | ||||||||
| Government | 309 | 0.40 (0.34 to 0.47) | 120 | 189 | 0.25 (0.20 to 0.32) | 1.20 (0.94 to 1.54) | 0.23 (−0.19 to 0.66) | 4.72 (3.41 to 6.52) |
| For-profit | 448 | 0.71 (0.61 to 0.82) | 190 | 258 | 0.54 (0.44 to 0.67) | 1.49 (1.19 to 1.87) | 2.13 (1.87 to 2.40) | 2.72 (2.04 to 3.64) |
| Non-profit | 2750 | Referent | 1669 | 1081 | Referent | 1.95 (1.70 to 2.25) | Referent | 1.95 (1.70 to 2.25) |
| Accreditations or affiliations | ||||||||
| COC-accredited | ||||||||
| No | 1423 | 1.24 (1.11 to 1.38) | 643 | 780 | 1.10 (0.95 to 1.27) | 2.28 (1.95 to 2.66) | 0.08 (−0.34 to 0.50) | 2.07 (1.74 to 2.47) |
| Yes | 2093 | Referent | 1336 | 757 | Referent | 2.16 (1.85 to 2.52) | Referent | 2.16 (1.85 to 2.52) |
| NSABP affiliation | ||||||||
| No | 2750 | 0.93 (0.82 to 1.06) | 1675 | 1075 | 0.69 (0.58 to 0.83) | 2.03 (1.69 to 2.43) | 0.33 (−0.8 to 0.74) | 2.92 (2.32 to 3.68) |
| Yes | 766 | Referent | 304 | 462 | Referent | 2.01 (1.76 to 2.30) | Referent | 2.01 (1.76 to 2.30) |
Adjusted for age at diagnosis and type of surgery
OR, odds ratio; CI, confidence interval; RERI, relative excess risk due to interaction; NHW, non-Hispanic White; NHB, non-Hispanic Black; COC, commission on cancer; NSABP, National Surgical Adjuvant Breast and Bowel Project
Overall, the NHB women were 2.15 times more likely to experience a delay in surgical management (95% CI, 1.92–2.41) than the NHW women. This disparity in surgical delay persisted across facility characteristics, although the magnitude of the disparity varied. Among the patients who received surgery at a government facility, the NHB women were 4.72 times more likely to experience a surgical delay (95% CI, 3.41–6.52) than the NHW women. The NHB women treated at a for-profit facility were 2.72 times more likely to experience a surgical delay than their NHW counterparts (95% CI, 2.04–3.64). We observed evidence of additive interaction for the NHB women treated at for-profit centers (RERI, 2.13; 95% CI, 1.87–2.40). The racial disparity in delay was least pronounced, but still present, among the women treated at a non-profit facility (OR, 1.95; 95% CI, 1.70–2.25).
Facility Characteristics and Breast Cancer Mortality
Table 4 illustrates the multivariable associations between surgical delay, facility characteristics, and breast cancer mortality overall, and the racial disparities in breast cancer mortality by delay and facility characteristics. We observed a slight increase in the hazard of breast cancer mortality among the breast cancer patients who experienced more than a 30-day delay in surgical management (HR, 1.11; 95% CI, 0.88–1.38) compared with those who did not. The breast cancer patients who received surgery more than 60 days after diagnosis had breast cancer mortality rate 1.28 times greater (95% CI: 0.99, 1.66) than those who received surgery 60 days or less after diagnosis. The patients who received surgery at a low-volume facility versus a high-volume facility had a hazard of breast cancer mortality rate 1.14 times higher (HR, 1.14; 95% CI, 0.89–1.47). Surgery received at a for-profit facility was associated with higher mortality than surgery received at a non-profit facility (HR, 1.25; 95% CI, 0.94–1.66).
TABLE 4.
Multivariable-adjusted hazard ratios (HRs) and 95 % confidence intervals (CIs) associating delay, facility characteristics with breast cancer mortality as well as overall and racial disparities by facility characteristics among 6011 women with breast cancer diagnosed in the metropolitan Atlanta area (2010–2014)
| Deaths (n) |
Overall HR (95 % CI a |
Deaths (n) |
Common referent HR (95 % CI a |
RERI (95 % CI) | Stratified effects HR (95 % CI)b |
|||
|---|---|---|---|---|---|---|---|---|
|
|
|
|||||||
| Overall | NHW | NHB | NHW | NHB | ||||
|
| ||||||||
| Overall disparity | 173 | 189 | – | – | 2.00 (1.63 to 2.46) | |||
| Dx to surgery | ||||||||
| Delay (>30 days) | 226 | 1.11 (0.88 to 1.38) | 90 | 136 | 0.98 (0.73 to 1.34) | 1.46 (1.07 to 1.99) | 0.19 (—0.33 to 0.71) | 2.05 (1.57 to 2.68) |
| No delay (≤30 days) | 136 | Referent | 83 | 53 | Referent | 1.27 (0.87 to 1.85) | Referent | 1.81 (1.28 to 2.56) |
| Dx to surgery | ||||||||
| Delay (>60 days) | 82 | 1.28 (0.99 to 1.66) | 28 | 54 | 1.31 (0.87 to 1.99) | 1.58 (1.10 to 2.26) | −0.19 (−0.93 to 0.55) | 1.62 (1.02 to 2.55) |
| No delay (≤60 days) | 280 | Referent | 145 | 135 | Referent | 1.46 (1.11 to 1.91) | Referent | 1.98 (1.56 to 2.51) |
| Annual patient volume | ||||||||
| Low | 94 | 1.14 (0.89 to 1.47) | 29 | 65 | 1.18 (0.78 to 1.8) | 1.50 (1.07 to 2.12) | −0.09 (−0.76 to 0.58) | 1.87 (1.20 to 2.90) |
| High | 268 | Referent | 144 | 124 | Referent | 1.43 (1.08 to 1.89) | Referent | 1.95 (1.53 to 2.48) |
| Medical school affiliation | ||||||||
| Yes | 108 | Referent | 31 | 99 | Referent | 1.80 (1.12 to 2.89) | Referent | 2.61 (1.72 to 3.95) |
| No | 253 | 1.10 (0.86 to 1.40) | 142 | 111 | 1.37 (0.91 to 2.08) | 1.85 (1.19 to 2.86) | −0.34 (−1.13 to 0.46) | 1.81 (1.41 to 2.32) |
| Facility type | ||||||||
| Government | 51 | 1.06 (0.78 to 1.45) | 27 | 24 | 1.29 (0.84 to 1.98) | 1.27 (0.80 to 2.10) | −0.74 (−1.76 to 0.27) | 1.31 (0.76 to 2.27) |
| For-profit | 67 | 1.25 (0.94 to 1.66) | 26 | 41 | 1.15 (0.74 to 1.80) | 1.88 (1.27 to 2.78) | 0.04 (−0.72 to 0.80) | 2.04 (1.24 to 3.34) |
| Non-profit | 243 | Referent | 120 | 123 | Referent | 1.46 (1.09 to 1.96) | Referent | 2.11 (1.64 to 2.72) |
| Accreditations or affiliations | ||||||||
| COC-accredited | ||||||||
| Yes | 207 | Referent | 120 | 87 | Referent | 1.27 (0.94 to 1.73) | Referent | 1.82 (1.38 to 2.40) |
| No | 155 | 0.98 (0.79 to 1.22) | 53 | 102 | 0.82 (0.58 to 1.14) | 1.36 (1.00 to 1.86) | 0.27 (−0.21 to 0.75) | 2.29 (1.64 to 3.20) |
| NSABP affiliation | ||||||||
| Yes | 267 | Referent | 135 | 132 | Referent | 1.41 (1.06 to 1.88) | Referent | 2.08 (1.63 to 2.64) |
| No | 95 | 1.11 (0.87 to 1.42) | 38 | 57 | 1.11 (0.76 to 1.62) | 1.52 (1.07 to 2.15) | 0.29 (−0.76 to 1.34) | 1.72 (1.14 to 2.60) |
Dx, diagnosis; HR, hazards ratio; CI, confidence interval; RERI, relative excess risk due to interaction; NHW, non-Hispanic White; NHB, non-Hispanic Black; COC, Commission on Cancer; NSABP, National Surgical Adjuvant Breast and Bowel Project
Adjusted for age at diagnosis, insurance, subtype, stage, and SES
Adjusted for age at diagnosis
Overall, the NHB breast cancer patients had an estimated mortality rate two times higher than the NHW women (HR, 2.00; 95% CI, 1.63–2.46). Even after restriction of the analysis to women who experienced a surgical delay (> 30 days), the NHB women a mortality rate two times higher than the NHW women (HR, 2.05; 95% CI, 1.57–2.68). The racial disparity in breast cancer mortality was less pronounced among the women who did not experience a surgical delay (HR, 1.81; 95% CI, 1.28–2.56), those who received surgery at a government facility (HR, 1.31; 95% CI, 0.76–2.27), and those treated at a COC-accredited facility (HR, 1.82; 95% CI, 1.38–2.40). Racial disparities in breast cancer mortality were most pronounced among the women who received surgery at a facility with a medical affiliation (HR, 2.61; 95% CI, 1.72–3.95), at a non-profit facility (HR, 2.11; 95% CI, 1.64–2.72), or at a non-COC-accredited facility (HR, 2.29; 95% CI, 1.64–3.20).
In our sensitivity analysis that excluded women whose breast cancer diagnosis was the first of multiple primaries, the results were similar, although less precise (data not shown).
DISCUSSION
To our knowledge, this is the first study to examine the associations between facility characteristics, surgical delay, and breast cancer mortality, as well as the potential impact of facility characteristics and surgical delay on racial disparities in breast cancer mortality. We found that breast cancer patients who received surgery at government facilities, non-profit facilities, and facilities with COC accreditation were less likely to experience a surgical delay. The NHB women were more likely to experience a surgical delay than the NHW women, and racial disparities in surgical delay were largely consistent across facility characteristics. The patients who experienced a surgical delay longer than 60 days had a higher hazard of breast cancer mortality. However, irrespective of surgical delay, the NHB women had a hazard of breast cancer mortality nearly two times greater than the NHW women. Similarly, racial disparities in breast cancer mortality persisted across facility characteristics, although they were least pronounced among the patients who received surgery at government facilities.
Variation in the timeliness of surgical care has been noted in previous publications. A study using data from the National Cancer Database reported that patients with Medicaid or no insurance and patients attending academic/research facilities or high-volume facilities were more likely to experience a delay in surgery.30 Moreover, the authors noted that throughout the study period (2003–2011), the average wait time for surgical care increased by approximately 1 week.30
In the current study, we found slight surgical delays among the patients treated in high-volume facilities. However, we also found that the patients receiving surgery at government and for-profit facilities were less likely to experience surgical delays than the patients treated at non-profit facilities. This may have been due to standardized care protocols at government facilities. For-profit facilities may have additional diagnostic workups, such as tumor boards, that may lead to delays relative to government facilities.
Racial disparities in receipt of surgery also have been documented.31 In a clinic-based study among insured Black and White women, the Black women had an average 47-day delay between diagnosis and surgery compared with the White women, who had an average 33-day delay between diagnosis and surgery. A study that defined surgical delay as 60 days or longer after diagnosis reported that Black women were three times more likely to experience a surgical delay.6 Our results were consistent with these previous studies, but these studies did not examine racial disparities in surgical delay by facility characteristics.
Surgical delays among women with early-stage disease may lead to axillary upstaging and are associated with breast cancer mortality.32,33 A study using Surveillance, Epidemiology, and End Results (SEER)-Medicare data and the National Cancer Database reported that women who received surgery more than 60 days after diagnosis have an estimated mortality rate 1.26 times higher than women who received surgery 60 days or less after diagnosis.34
In our analyses, we defined surgical delay as more than 30 days after diagnosis based on clinical input, but in our sensitivity analysis defining surgical delay as a delay longer than 60 days, we observed comparable results, highlighting the importance of timeliness of surgery. In our study, Black women were more likely to experience a surgical delay, which may contribute to racial disparities in breast cancer mortality and highlights that timeliness of surgical care may be especially important among vulnerable populations. We did not explore the interaction between stage, race, and delay, but future studies may benefit from an investigation into the role of treatment delays in upstaging in the context of racial disparities in breast cancer outcomes.
To date, literature on surgical facility characteristics in breast cancer mortality and racial disparities in breast cancer mortality is limited. Increasing hospital volume—a possible proxy for institutional capacity and more comprehensive care—also is reported to be associated with a lower breast cancer mortality rate.35 Greenup et al.35 used the National Cancer Database to examine the association between hospital volume and breast cancer mortality, comparing high- and low-volume facilities (HR, 0.89; 95% CI, 0.73–0.83). We observed similar, albeit less precise, results. In our study, the NHB women were more likely to receive surgery at a low-volume facility (31% vs 14%), suggesting that quality of care may be a factor contributing to racial disparities in breast cancer mortality, but would not eliminate the disparities because disparities persisted within strata of facility volume. This also was suggested by the racial disparities in COC facilities, with reduced disparities observed among patients who received surgery at COC-accredited facilities (HR, 1.82 vs 2.29), but Black women were less likely to receive surgery at a COC-accredited facility (50% vs 68%).
This study, although the first to explore race differences in surgical delay and breast cancer mortality by facility characteristics, had limitations to consider. First, even though we had information on surgical facility characteristics, we did not have information on surgical teams or demographic information of physicians at the surgical facilities, which may influence care.36 Previous studies have suggested, especially in maternal health, that concordance between physician and patient race may influence care and outcomes, which may be important to explore.36,37 However, breast cancer treatment is complex and involves multiple specialties. Therefore, understanding the care facility characteristics may be more relevant.
Second, although we examined facility characteristics of the surgical facility for women with a diagnosis of breast cancer in Atlanta, we did not have information on the facility where they received other adjuvant therapies or whether they received other standard-of-care treatment, which also may influence mortality. However, by restricting our study population to women residing in the metropolitan Atlanta area, it is likely that most of the women received adjuvant therapies at the same care facility where they received their surgery.
Third, we did not have information on comorbid conditions, which have been suggested previously as important for surgical delays because patients with more underlying health conditions may require more complex procedures.38,39 Finally, where a woman chooses to receive care may be influenced by distance, insurance, SES, and education, which reduce the exchangeability of patients across facility types.15 We controlled for neighborhood SES and insurance in our analyses to mitigate confounding by these known factors. However, unknown and unmeasured reasons exist to explain why women may choose certain care facilities to receive surgery.
In conclusion, we observed pronounced racial disparities in surgical delay and breast cancer mortality. Surgical delay, hospital volume, and facility accreditations may partially contribute to racial disparities in mortality. Women treated at government care facilities had the lowest racial disparities in breast cancer mortality, although we observed racial disparities in surgical delay at government facilities. Future research may benefit from more comprehensive studies investigating quality of care, delays and completion of care, and breast cancer mortality to further our understanding of racial disparities in breast cancer mortality.
ACKNOWLEDGMENTS
This this work was supported in part by the Cancer Prevention and Control Research program; the Winship Research Informatics shared resources, a core supported by the Winship Cancer Institute of Emory University; and the Komen Foundation (CCR19608510) awarded to Lauren E McCullough. Lindsay J. Collin was supported by TL1TR002540 from the National Center for Advancing Translational Sciences of the National Institutes of Health. Jasmine M. Miller-Kleinhenz was supported NIH/NIGMS (2K12GM000680). The collection of cancer incidence data used in this study was supported by contract HHSN261201800003I, Task Order HHSN26100001 from the NCI and cooperative agreement 5NU58DP003875-04 from the U.S. Centers for Disease Control and Prevention.
Footnotes
DISCLOSURE There are no conflicts of interest.
REFERENCES
- 1.DeSantis CE, Fedewa SA, Goding Sauer A, Kramer JL, Smith RA, Jemal A. Breast cancer statistics, 2015: convergence of incidence rates between black and white women. CA Cancer J Clin. 2016;66:31–42. 10.3322/caac.21320. [DOI] [PubMed] [Google Scholar]
- 2.Aizer AA, Wilhite TJ, Chen M-H, et al. Lack of reduction in racial disparities in cancer-specific mortality over a 20-year period. Cancer. 2014;120:1532–9. 10.1002/cncr.28617. [DOI] [PubMed] [Google Scholar]
- 3.DeSantis CE, Siegel RL, Sauer AG, et al. Cancer statistics for African Americans, 2016: progress and opportunities in reducing racial disparities. CA Cancer J Clin. 2016;66:290–308. 10.3322/caac.21340. [DOI] [PubMed] [Google Scholar]
- 4.DeSantis CE, Ma J, Sauer AG, Newman LA, Jemal A. Breast cancer statistics, 2017, racial disparity in mortality by state. CA Cancer J Clin. 2017;67:439–48. 10.3322/caac.21412. [DOI] [PubMed] [Google Scholar]
- 5.Collin LJ, Jiang R, Ward KC, et al. Racial disparities in breast cancer outcomes in the metropolitan Atlanta area: new insights and approaches for health equity. JNCI Cancer Spectr. 2019. 10.1093/jncics/pkz053. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.George P, Chandwani S, Gabel M, et al. Diagnosis and surgical delays in African American and white women with early-stage breast cancer. J Womens Health Larchmt. 2015;24:209–17. 10.1089/jwh.2014.4773. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Gorin SS, Heck JE, Cheng B, Smith SJ. Delays in breast cancer diagnosis and treatment by racial/ethnic group. Arch Intern Med. 2006;166:2244–52. 10.1001/archinte.166.20.2244. [DOI] [PubMed] [Google Scholar]
- 8.Miller-Kleinhenz JM, Collin LJ, Seidel R, et al. Racial disparities in diagnostic delay among women with breast cancer. J Am Coll Radiol <AQ4> Published online 16 July 2021:S1546-1440(21)00549-4. Doi: 10.1016/j.jacr.2021.06.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Smith EC, Ziogas A, Anton-Culver H. Delay in surgical treatment and survival after breast cancer diagnosis in young women by race/ethnicity. JAMA Surg. 2013;148:516–23. 10.1001/jamasurg.2013.1680. [DOI] [PubMed] [Google Scholar]
- 10.Fedewa SA, Edge SB, Stewart AK, Halpern MT, Marlow NM, Ward EM. Race and ethnicity are associated with delays in breast cancer treatment (2003–2006). J Health Care Poor Underserved. 2011;22:128–41. [DOI] [PubMed] [Google Scholar]
- 11.Bilimoria KY, Ko CY, Tomlinson JS, et al. Wait times for cancer surgery in the United States: trends and predictors of delays. Ann Surg. 2011;253:779–85. 10.1097/SLA.0b013e318211cc0f. [DOI] [PubMed] [Google Scholar]
- 12.McLaughlin JM, Anderson RT, Ferketich AK, Seiber EE, Balkrishnan R, Paskett ED. Effect on survival of longer intervals between confirmed diagnosis and treatment initiation among low-income women with breast cancer. J Clin Oncol. 2012;30:4493–500. 10.1200/JCO.2012.39.7695. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Guller U, Safford S, Pietrobon R, Heberer M, Oertli D, Jain NB. High hospital volume is associated with better outcomes for breast cancer surgery: analysis of 233,247 patients. World J Surg. 2005;29: 994–9; Discussion 999–1000. Doi: 10.1007/s00268-005-7831-z. [DOI] [PubMed] [Google Scholar]
- 14.Keating NL, Kouri E, He Y, Weeks JC, Winer EP. Racial differences in definitive breast cancer therapy in older women: Are they explained by the hospitals where patients undergo surgery? Med Care. 2009;47:765–73. 10.1097/MLR.0b013e31819e1fe7. [DOI] [PubMed] [Google Scholar]
- 15.Wheeler SB, Reeder-Hayes KE, Carey LA. Disparities in breast cancer treatment and outcomes: biological, social, and health system determinants and opportunities for research. Oncologist. 2013;18:986–93. 10.1634/theoncologist.2013-0243. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Howe HL. NAACCR Guideline for Enhancing Hispanic-Latino Identification: Revised NAACCR Hispanic/Latino Identification Algorithm [NHIA V2].; 2005. <PUB1>
- 17.Warren JL, Klabunde CN, Schrag D, Bach PB, Riley GF. Overview of the SEER-Medicare data: content, research applications, and generalizability to the United States elderly population. Med Care. 2002;40: IV3–18. [DOI] [PubMed] [Google Scholar]
- 18.Enewold L, Parsons H, Zhao L, et al. Updated overview of the SEER-Medicare data: enhanced content and applications. JNCI Monogr. 2020;2020:3–13. 10.1093/jncimonographs/lgz029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Bleicher RJ. Timing and delays in breast cancer evaluation and treatment. Ann Surg Oncol. 2018;25:2829–38. 10.1245/s10434-018-6615-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Knoble NB, Alderfer MA, Hossain MJ. Socioeconomic status (SES) and childhood acute myeloid leukemia (AML) mortality risk: analysis of SEER data. Cancer Epidemiol. 2016;44:101–8. 10.1016/j.canep.2016.07.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kish JK, Yu M, Percy-Laurry A, Altekruse SF. Racial and ethnic disparities in cancer survival by neighborhood socioeconomic status in Surveillance, Epidemiology, and End Results (SEER) registries. J Natl Cancer Inst Monogr. 2014;2014:236–43. 10.1093/jncimonographs/lgu020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Knol MJ, VanderWeele TJ. Recommendations for presenting analyses of effect modification and interaction. Int J Epidemiol. 2012;41:514–20. 10.1093/ije/dyr218. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.VanderWeele TJ, Knol MJ. A tutorial on interaction. Epidemiol Methods. 2014;3:33–72. 10.1515/em-2013-0005. [DOI] [Google Scholar]
- 24.Hosmer DW, Lemeshow S. Confidence interval estimation of interaction. Epidemiology. 1992;3:452–6. [DOI] [PubMed] [Google Scholar]
- 25.Starr JR, McKnight B. Assessing interaction in case-control studies: type I errors when using both additive and multiplicative scales. Epidemiology. 2004;15:422–7. [DOI] [PubMed] [Google Scholar]
- 26.Ward JB, Gartner DR, Keyes KM, Fliss MD, McClure ES, Robinson WR. How do we assess a racial disparity in health? Distribution, interaction, and interpretation in epidemiological studies. Ann Epidemiol. 2019;29:1–7. 10.1016/j.annepidem.2018.09.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Howards PP, Schisterman EF, Poole C, Kaufman JS, Weinberg CR. Toward a clearer definition of confounding revisited with directed acyclic graphs. Am J Epidemiol. 2012;176:506–11. 10.1093/aje/kws127. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Lash TL. The harm done to reproducibility by the culture of null hypothesis significance testing. Am J Epidemiol. 2017;186:627–35. 10.1093/aje/kwx261. [DOI] [PubMed] [Google Scholar]
- 29.Wasserstein RL, Lazar NA. The ASA Statement on p values: context, process, and purpose. Am Statistician. 2016;70:129–33. 10.1080/00031305.2016.1154108. [DOI] [Google Scholar]
- 30.Liederbach E, Sisco M, Wang C, et al. Wait times for breast surgical operations, 2003–2011: a report from the National Cancer Database. Ann Surg Oncol. 2015;22:899–907. 10.1245/s10434-014-4086-7. [DOI] [PubMed] [Google Scholar]
- 31.Sheppard VB, Oppong BA, Hampton R, et al. Disparities in breast cancer surgery delay: the lingering effect of race. Ann Surg Oncol. 2015;22:2902–11. 10.1245/s10434-015-4397-3. [DOI] [PubMed] [Google Scholar]
- 32.Khader A, Chang S-C, Santamaria-Barria J, et al. Delay in surgery is associated with axillary upstaging of clinically node negative breast cancer patients. J Surg Oncol. 2021;123:854–65. 10.1002/jso.26332. [DOI] [PubMed] [Google Scholar]
- 33.Wagner JL, Warneke CL, Mittendorf EA, et al. Delays in primary surgical treatment are not associated with significant tumor size progression in breast cancer patients. Ann Surg. 2011;254:119–24. 10.1097/SLA.0b013e318217e97f. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Bleicher RJ, Ruth K, Sigurdson ER, et al. Time to surgery and breast cancer survival in the United States. JAMA Oncol. 2016;2:330–9. 10.1001/jamaoncol.2015.4508. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Greenup RA, Obeng-Gyasi S, Thomas S, et al. The effect of hospital volume on breast cancer mortality. Ann Surg. 2018;267:375–81. 10.1097/SLA.0000000000002095. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Greenwood BN, Hardeman RR, Huang L, Sojourner A. Physician–patient racial concordance and disparities in birthing mortality for newborns. PNAS. 2020;117:21194–200. 10.1073/pnas.1913405117. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Meghani SH, Brooks JM, Gipson-Jones T, Waite R, Whitfield-Harris L, Deatrick JA. Patient–provider race-concordance: does it matter in improving minority patients’ health outcomes? Ethnicity Health. 2009;14:107–30. 10.1080/13557850802227031. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Zhou J, Enewold L, Zahm SH, et al. Breast conserving surgery versus mastectomy: the influence of comorbidities on choice of surgical operation in the Department of Defense health care system. Am J Surg. 2013;206:393–9. 10.1016/j.amjsurg.2013.01.034. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Simunovic M, Thériault M-E, Paszat L, et al. Using administrative databases to measure waiting times for patients undergoing major cancer surgery in Ontario, 1993–2000. Can J Surg. 2005;48:137–42. [PMC free article] [PubMed] [Google Scholar]