Abstract
Introduction:
Racial and socioeconomic disparities in health access and outcomes for many conditions are well known. However, for time-sensitive high-acuity diseases such as traumatic injuries, disparities in access and outcomes should be significantly diminished. Our primary objective was to characterize racial disparities across majority, mixed-race, and minority hospitals for African American (AA) versus white males with high-grade splenic injuries.
Methods:
Data from the National Trauma Data Bank was utilized from 2007 to 2015. A total of 24 855 AA or white males with high-grade splenic injuries were included. Multilevel mixed effects regression analysis was used to evaluate disparities in outcomes and resource allocation.
Results:
Mortality was significantly higher for AA males at mixed-race (odds ratio [OR] 1.6; 95% CI 1.3-2.1; P < .001) and minority (OR 2.1; 95% CI 1.5-3.0; P < .001) hospitals, but not at majority hospitals. At minority hospitals, AA males were significantly less likely to be admitted to the intensive care unit (OR 0.7; 95% CI 0.49-0.97; P = .04) and experienced a significantly longer time to surgery (IRR 1.5; P = .02). Minority hospitals were significantly more likely to have failures from angiographic embolization requiring operative intervention (OR 2.2; P = .009). At both types of nonmajority hospitals, AA males with penetrating injuries were more likely to be managed with angiography (mixed-race hospitals: OR 1.7; P = .046 vs minority hospitals: OR 1.6; P = .08).
Discussion:
While multiple studies have shown that minority hospitals have increased mortality compared to majority hospitals, this study found this disparity only existed for AAs.
Introduction
Being male in the United States is associated with social and economic advantages. Globally, however, men have higher mortality rates than women for 14 of the top 15 leading causes of death.1 Men of color, particularly black men, account for much of the gender difference in mortality: the difference in life expectancy for black (vs white) men far exceeds the difference in life expectancy for men (vs women).2 US-born black and Hispanic men have higher rates of fatal chronic conditions and a shorter average life expectancy than their white and female counterparts.3 The age-adjusted mortality rate for black men is 19% higher than for white men.4
African American (AA) and Hispanic men face many racialized and gendered social norms, cultural expectations, and stressors that can negatively shape their health behaviors and explain their higher rates of morbidity and mortality. Furthermore, AA men are among the most disadvantaged racial/gender groups in the United States across several domains (eg, unemployment, incarceration, discrimination, and homicide); these social determinants of health contribute to disproportionately high rates of unhealthy behaviors, functional limitations, and premature mortality among men of color.5–7
Racial and socioeconomic disparities in health access and outcomes for many conditions, especially chronic diseases, are well known.8 As compared with white males, AA patients are 2 to 3 times more likely to undergo bilateral orchiectomies and lower limb amputations due to suboptimal management of chronic diseases.3 However, for time-sensitive high-acuity diseases such as traumatic injuries, which are not necessarily correlated with race or socioeconomic status, disparities in access and outcomes should be significantly diminished. Yet are they?
Previous studies have demonstrated, even after controlling for payer status, that AA patients are more likely to die after major trauma.9,10 Haider et al11 have shown that black-white disparities in trauma mortality may result from the fact that hospitals treating higher proportions of minority patients have worse outcomes compared with hospitals treating predominantly white patients.
In this study, our primary objective was to characterize racial disparities across majority, mixed-race, and minority hospitals for AA (vs white) males with high-grade splenic injuries. Our secondary objective was to identify differences in resource allocation by race and by hospital racial type.
Methods
Data
The data source for this study was the American College of Surgeons (ACS) National Trauma Data Bank (NTDB), the largest trauma registry in the United States.12 It includes patient-level and hospital data on traumatic injuries and clinical outcomes for more than 800 trauma centers. We limited our analysis to years after 2006 because of data quality improvements implemented in 2007.
This study was approved by the University of Minnesota institutional review board (STUDY00001489).
Participants
We obtained patient-level data from the NTDB from January 1, 2007 through December 31, 2015. The inclusion criteria were as follows:
Age ≥ 16 years
Abbreviated Injury Score (AIS) grade 3 or higher splenic injury
Excluded from our study were all females; males whose racial status was other than “black non-Hispanic” or “white non-Hispanic”; hospitals that submitted fewer than 50 patients to the NTDB; males with no signs of life at the initial evaluation in the emergency department (ED), that is, systolic blood pressure (SBP) = 0, pulse = 0, Glasgow Coma Scale (GCS) score = 3; and due to low sample size, males (n = 3) treated at an ACS-verified level 4 trauma center.
Measures
We defined splenic injuries using the following three categories: (1) admission AIS codes (544299.2, 544210.2, 544212.2, 544214.3, 544220.2, 544222.2, 544224.3, 544226.4, 544228.5, 544240.3); (2) ICD-9 diagnosis codes (865, 865.XX); and (3) ICD-10 diagnosis codes (S36.031A, S36.032A, S36.021A, S36.030A, S36.00XA, S36.020A, S36.029A, S36.039A, S36.09XA).
Approximately 5% or less of patient variables were missing; however, a small number of such variables were missing at a higher rate (alcohol level, 10% missing; minutes spent in transit via emergency medical services [EMS], 26% missing). We defined “minutes spent in transit via EMS” as the total elapsed time from dispatch of the EMS transporting unit to hospital arrival. To account for missing data, we used the Stata multiple imputation (mi) suite of commands, with 10 imputations for each missing value, applying all demographic, covariate, and outcome variables.
We defined angiography using the following two categories: (1) ICD-9 procedure codes (88.4, 88.40, 88.42, 88.45, 88.46-88.49, 39.71, 30.73-39.99) and (2) ICD-10 procedure codes (04L33DZ, 04L33ZZ, 04L03DZ, 04R04JZ, 04U04JZ, 04V03DZ, 04V34ZZ, 04L43DZ, 04L43ZZ, 04L44CZ, 04L44DZ, 04L44ZZ).
We defined operations using the following two categories: (1) ICD-9 procedure codes (41.42-41.45, 41.95, 41.99, 54.10-54.12, 54.19) and (2) ICD-10 procedure codes (075P0ZZ, 07974ZX, 079P0ZZ, 079P4ZZ, 079T0ZX, 079T3ZZ, 07BB0ZX, 07B.C.0ZX, 7BH3ZX, 07BP0ZX, 07BP0ZZ, 07BP3ZZ, 07C10ZZ, 07C14ZZ, 07C20ZZ, 07C50ZZ, 07CM0ZZ, 07CP0ZZ, 07JP0ZZ, 07Q20ZZ, 07TP0ZZ, 07TP4ZZ).
We categorized hospital racial type by the percent of AA patients in all trauma admissions in 2007 through 2015: minority (>50%), mixed race (25% through 50%), and majority (<25%).
Statistical Analysis
To analyze demographic differences by hospital type and race, we used the χ2 test (for binary variables) and the Kruskal-Wallis test (for minutes spent in transit via EMS). For AA versus white males within and across hospitals, we compared the adjusted odds of all-cause in-hospital mortality, early mortality (defined as death within 48 hours after hospital arrival), intensive care unit (ICU) admission, and treatment (angiography or surgery); to do so, we used a multilevel mixed-effects regression model, accounting for hospital-level random effects. Our models included the following fixed effects: age, insurance status, presence of bowel injuries, blunt or penetrating mechanism of injury, Injury Severity Score (ISS), grade of splenic injuries, GCS score, ED heart rate, ED SBP, ED respiratory distress (defined as respiratory rate >29 or <10), transfer status, alcohol level, minutes spent in transit via EMS, year of admission, hospital ACS-verified level, and significant comorbidities on univariate analysis (bleeding disorders, congenital anomalies, congestive heart failure, chronic renal failure, diabetes mellitus, disseminated cancer, advanced directives limiting care, history of myocardial infarction [MI], hypertension requiring medication, obesity, cirrhosis, major psychiatric illness, and drag use disorders).
Our primary outcome was all-cause in-hospital mortality. Secondary outcomes included early mortality, ICU admission, major complications, treatment strategy, and time to treatment. We defined major complications as systemic sepsis, pulmonary embolism, pneumonia, acute renal failure, acute respiratory distress syndrome, or a cardiovascular complication (arrest, MI, or cerebrovascular accident). Those major complications have previously been verified to have the highest attributable mortality among trauma patients.13,14 To evaluate all-cause in-hospital mortality, early mortality, ICU admission, major complications, and treatment strategy, we used multilevel mixed-effects logistic regression models. To evaluate time to treatment, we used mixed-effects negative binomial regression models.
For all statistical analyses, we used SAS (version 9.4, SAS Institute, Cary, NC, United States) and Stata MP, version 15 (StataCorp, College Station, TX, United States). A P-value < .05 was considered statistically significant.
Results
In the 2007 to 2015 NTDB database, records were available for 6768 156 patients (Figure 1). After we applied our inclusion and exclusion criteria, 24 855 patients with high-grade (≥ grade 3) splenic injuries were eligible for our study.
Figure 1.
Study diagram detailing selection of patients in 2007 to 2015 National Trauma Data Bank. AIS, Abbreviated Injury Scale; ACS, American College of Surgeons; AA, African American; CA, Caucasian American; NTDB, National Trauma Data Bank.
The baseline characteristics and comorbidities of patients, by hospital racial type, are shown in Table 1; their injuries and morbidities, in Table 2. AA (vs white) males were significantly more likely to suffer penetrating injuries, across all hospital racial types; to suffer bowel injuries; to have self-pay or Medicaid insurance status; and to be younger.
Table 1.
Baseline Demographics and Comorbid Characteristics of 24 855 Patients with Grade 3 or Higher Spleen Injury.
| Hospital minority status by race, no. (%) |
|||||||
|---|---|---|---|---|---|---|---|
| Majority |
Mixed |
Minority |
|||||
| CA patients | AA patients | CA patients | AA patients | CA patients | AA patients | ||
|
|
|||||||
| Variable | (n = 17 843) | (n = 1221) | (n = 3421) | (n = 1119) | (n = 549) | (n = 702) | P value |
| Age | |||||||
| 16-25 | 5595 (31.4) | 451 (36.9) | 1030 (30.1) | 373 (33.3) | 145 (26.4) | 235 (33.5) | <.001 |
| 26-45 | 5721 (32.0) | 456 (37.4) | 1101 (32.2) | 447 (40.0) | 190 (34.6) | 292 (41.5) | |
| 46-65 | 5027 (28.2) | 276 (22.6) | 1019 (29.8) | 264 (23.6) | 168 (30.6) | 159 (22.7) | |
| 66-75 | 901 (5.0) | 29 (2.4) | 162 (4.7) | 30 (2.7) | 30 (5.5) | 11 (1.6) | |
| >75 | 598 (3.4) | 9 (0.7) | 109 (3.2) | 5 (0.4) | 16 (2.9) | 5 (0.7) | |
| Insurance type | |||||||
| Medicaid | 2107 (12.6) | 304 (26.3) | 427 (13.2) | 260 (24.5) | 85 (16.7) | 191 (29.8) | <.001 |
| Medicare | 1417 (8.5) | 56 (4.8) | 311 (9.6) | 63 (5.9) | 39 (7.7) | 24 (3.8) | |
| Private | 10 610 (63.6) | 460 (39.8) | 1784 (55.1) | 371 (35.0) | 268 (52.8) | 232 (36.2) | |
| Self-pay | 2555 (15.3) | 336 (29.1) | 716 (22.1) | 367 (34.6) | 116 (22.8) | 193 (30.2) | |
| Year of admission | |||||||
| 2007 | 497 (2.8) | 36 (3.0) | 78 (2.3) | 23 (2.1) | 40 (7.3) | 66 (9.4) | <.001 |
| 2008 | 1138 (6.4) | 82 (6.7) | 231 (6.8) | 68 (6.1) | 44 (8.0) | 63 (9.0) | |
| 2009 | 1377 (7.7) | 92 (7.5) | 191 (5.6) | 75 (6.7) | 45 (8.2) | 63 (9.0) | |
| 2010 | 1767 (9.9) | 122 (10.0) | 276 (8.1) | 95 (8.5) | 51 (9.3) | 66 (9.4) | |
| 2011 | 2066 (11.6) | 129 (10.6) | 390 (11.4) | 133 (11.9) | 38 (6.9) | 70 (10.0) | |
| 2012 | 2503 (14.0) | 151 (12.4) | 543 (15.9) | 157 (14.0) | 56 (10.2) | 70 (10.0) | |
| 2013 | 2515 (14.1) | 168 (13.8) | 519 (15.2) | 168 (15.0) | 78 (14.2) | 87 (12.4) | |
| 2014 | 2845 (15.9) | 216 (17.7) | 551 (16.0) | 187 (16.7) | 85 (15.5) | 94 (13.4) | |
| 2015 | 3135 (17.6) | 225 (18.5) | 642 (18.7) | 213 (19.0) | 112 (20.4) | 123 (17.4) | |
| Other comorbidity | 4014 (23.7) | 227 (20.3) | 728 (23.0) | 213 (20.9) | 148 (28.3) | 162 (25.3) | .002 |
| Bleeding disorder | 570 (3.4) | 23 (2.1) | 107 (3.4) | 18 (1.8) | 12 (2.3) | 4 (0.6) | <.001 |
| Congestive heart failure | 203 (1.2) | 13 (1.2) | 43 (1.4) | 6 (0.6) | 4 (0.8) | 3 (0.5) | .2 |
| Chronic renal failure | 71 (0.4) | 9 (0.8) | 15 (0.5) | 10 (1.0) | 3 (0.6) | 2 (0.3) | .08 |
| Diabetes mellitus | 1139 (6.7) | 55 (4.9) | 210 (6.6) | 67 (6.6) | 27 (5.2) | 31 (4.8) | .06 |
| Disseminated cancer | 82 (0.5) | 3 (0.3) | 10 (0.3) | 1 (0.1) | 1 (0.2) | 0 (0) | .1 |
| Advanced directive | 80 (0.5) | 4 (0.4) | 11 (0.4) | 0 (0) | 2 (0.4) | 0 (0) | .1 |
| History of MI | 155 (0.9) | 4 (0.4) | 29 (0.9) | 3 (0.3) | 3 (0.6) | 5 (0.8) | .1 |
| Hypertension | 2888 (17.1) | 168 (15.0) | 565 (17.9) | 188 (18.5) | 64 (12.3) | 63 (9.8) | <.001 |
| Obesity | 872 (5.2) | 46 (4.1) | 174 (5.5) | 52 (5.1) | 13 (2.5) | 15 (2.3) | <.001 |
| Cirrhosis | 165 (1.0) | 7 (0.6) | 35 (1.1) | 9 (0.9) | 14 (2.7) | 2 (0.3) | .001 |
| Major psychiatric illness | 712 (4.2) | 41 (3.7) | 166 (5.3) | 35 (3.4) | 35 (6.7) | 20 (3.1) | .002 |
| Drug use disorder | 1071 (6.3) | 91 (8.2) | 237 (7.5) | 111 (10.9) | 45 (8.6) | 38 (5.9) | <.001 |
Abbreviations: CA, Caucasian American: AA, African American; MI, myocardial infarction.
Table 2.
Baseline Injury Severity Characteristics of 24 855 Patients with Grade 3 or Higher Spleen Injury.
| Hospital minority status by race, no. (%) |
|||||||
|---|---|---|---|---|---|---|---|
| Majority |
Mixed |
Minority |
|||||
| CA patients | AA patients | CA patients | AA patients | CA patients | AA patients | ||
|
|
|||||||
| Variable | (n = 17 843) | (n = 1221) | (n = 3421) | (n = 1119) | (n = 549) | (n = 702) | P value |
| ISS, n (%) | |||||||
| 9-15 | 3407 (19.6) | 165 (13.8) | 502 (15.8) | 127 (12.0) | 96 (18.7) | 80 (11.8) | <.001 |
| 16-24 | 5253 (30.3) | 308 (25.8) | 945 (29.8) | 291 (27.5) | 148 (28.7) | 180 (26.6) | |
| 25-35 | 5142 (29.6) | 426 (35.7) | 1009 (31.8) | 410 (38.7) | 150 (29.1) | 262 (38.8) | |
| >35 | 3566 (20.5) | 295 (24.7) | 721 (22.7) | 232 (21.8) | 121 (23.5) | 154 (22.8) | |
| Respiratory distress | |||||||
| Yes | 2859 (16.0) | 248 (20.3) | 546 (16.0) | 193 (17.3) | 119 (21.7) | 148 (21.1) | <.001 |
| No | 14984 (84.0) | 973 (79.7) | 2875 (84.0) | 926 (82.7) | 430 (78.3) | 554 (78.9) | |
| Blood pressure | |||||||
| ≤60 mmHg | 741 (4.2) | 86 (7.0) | 185 (5.4) | 76 (6.8) | 23 (4.2) | 61 (8.7) | <.001 |
| 61-90 mmHg | 1915 (10.7) | 123 (10.1) | 347 (10.1) | 119 (10.6) | 57 (10.4) | 82 (11.7) | |
| >91 mmHg | 15 187 (85.1) | 1012 (82.9) | 2889 (84.5) | 924 (82.6) | 469 (85.4) | 559 (79.6) | |
| Penetrating injury | 651 (3.8) | 410 (35.0) | 141 (4.2) | 392 (36.2) | 44 (8.4) | 332 (48.5) | <.001 |
| Bowel injury | 1769 (9.9) | 321 (26.3) | 442 (12.9) | 304 (27.2) | 73 (13.3) | 263 (37.5) | <.001 |
| Pulse | |||||||
| ≤50 bpm | 169 (1.0) | 10 (0.9) | 20 (0.6) | 4 (0.4) | 5 (0.9) | 4 (0.6) | <.001 |
| 51-119 bpm | 14 492 (83.0) | 953 (81.2) | 2676 (80.8) | 884 (81.8) | 424 (79.6) | 535 (80.3) | |
| ≥120 bpm | 2797 (16.0) | 210 (17.9) | 617 (18.6) | 193 (17.9) | 104 (19.5) | 127 (19.1) | |
| GCS score | |||||||
| 3-8 | 2846 (16.5) | 219 (18.4) | 613 (18.5) | 181 (16.7) | 104 (19.4) | 106 (15.5) | <.001 |
| 9-13 | 769 (4.5) | 92 (7.8) | 172 (5.2) | 67 (6.2) | 26 (4.8) | 38 (5.6) | |
| 14-15 | 13 636 (79.0) | 876 (73.8) | 2520 (76.3) | 838 (77.1) | 407 (75.8) | 540 (78.9) | |
| Splenic grade (AIS) | |||||||
| 3 | 1266 (7.1) | 75 (6.1) | 192 (5.6) | 66 (5.9) | 35 (6.4) | 44 (6.3) | <.001 |
| 4 | 12 348 (69.2) | 851 (69.7) | 2376 (69.5) | 758 (67.7) | 381 (69.4) | 438 (62.4) | |
| 5 | 4229 (23.7) | 295 (24.2) | 853 (24.9) | 295 (26.4) | 133 (24.2) | 220 (31.3) | |
| Transfer in | 5797 (32.5) | 224 (18.4) | 1055 (30.8) | 242 (21.6) | 114 (20.8) | 59 (8.4) | <.001 |
| EMS minutes, median Alcohol | 45.0 | 29.0 | 48.0 | 31.0 | 29.0 | 24.0 | <.001 |
| Yes—above legal | 2048 (12.9) | 175 (16.0) | 502 (15.8) | 176 (16.8) | 60 (12.2) | 93 (15.8) | <.001 |
| Yes—below legal | 983 (6.2) | 100 (9.1) | 195 (6.1) | 82 (7.8) | 42 (8.5) | 76 (12.9) | |
| No—confirmed | 6992 (43.9) | 426 (38.8) | 1413 (44.3) | 433 (41.3) | 254 (51.4) | 252 (42.9) | |
| Not tested | 5894 (37.0) | 396 (36.1) | 1077 (33.8) | 357 (34.1) | 138 (27.9) | 167 (28.4) | |
Abbreviations: CA, Caucasian American; AA, African American; ISS, Injury Severity Score; GCS, Glasgow Coma Scale; AIS, Abbreviated Injury Scale; EMS, emergency medical services; bpm, beats per minute.
All-Cause In-Hospital Mortality
AA males at mixed-race hospitals (vs white males majority hospitals) had significantly higher mortality (odds ratio [OR] 1.6; 95% CI 1.3-2.1; P < .001). AA males at minority hospitals (vs white males majority hospitals) had significantly higher mortality (OR 2.1; 95% CI 1.5-3.0; P < .001). Within minority hospitals AA (vs white) males had significantly higher mortality (OR 1.55; 95% CI 1.0-2.4; P = .050). But AA (vs white) males at majority hospitals did not have significantly higher mortality (OR 1.2; 95% CI 0.96-1.6; P = .1). For all of those comparisons, see Figure 2.
Figure 2.
All-cause in-hospital mortality across races and hospital types. Reference category: whites at a majority hospital. C-statistic = 0.9. AA, African American; CA, Caucasian American; OR, odds ratio.
Secondary Outcomes
Evaluating early mortality (a surrogate for a hemorrhagic cause of death), we found that AA males had a stepwise increase in early mortality at all hospital racial types. AA males (vs white males at majority hospitals) had 34% higher odds of early mortality (OR 1.3; 95% CI 1.008-1.8; P = .04) at majority hospitals, 61% higher odds at mixed-race hospitals (OR 1.6; 95% CI 1.2-2.2; P = .003), and 86% higher odds at minority hospitals (OR 1.9; 95% CI 1.3-2.7; P = .002). But we found no significant differences in early mortality for white males at majority (vs minority or mixed-race) hospitals (Figure 3A).
Figure 3.
Secondary outcomes across races and hospital types: (A) odds of 48-hour all-cause in-hospital mortality; (B) odds of development of major complication; (C) odds of ICU admission. Reference category: whites at a majority hospital. OR, odds ratio.
AA (vs white) males at mixed-race hospitals—but not at minority hospitals—had a significantly higher rate of major complications (Figure 3B).
AA males in all hospital racial types were less likely to be admitted to the ICU compared with white males at majority hospitals, but the difference was not statistically significant (Figure 3C). However, within minority hospitals, AA (vs white) males were significantly less likely to be admitted to the ICU (OR 0.69; 95% CI 0.49-0.97; P = .035).
For patients managed nonoperatively, we found that white (vs AA) males were significantly more likely to undergo angiography at minority hospitals (OR 1.7; 95% CI 1.08-2.7; P = .02; Figure 4A).
Figure 4.
Treatment evaluation across races and hospital types: (A) odds of receiving angiography for patients managed nonoperatively; (B) odds of angiographic failure requiring operative intervention; (C) odds of receiving angiography as initial/sole therapy in penetrating injuries; (D) incidence rate ratio of time to surgery. Reference category: whites at a majority hospital. AA, African American; CA, Caucasian American; OR, odds ratio; IRR, incidence rate ratio.
Minority hospitals were significantly more likely to have failures from angiographic embolization requiring operative intervention compared with majority and mixed-race hospitals (CA males: OR 2.2, P = .009; and AA males: OR 1.8, P = .053; Figure 4B).
Additionally, AA males in mixed-race and minority hospitals with penetrating injuries were more likely to be managed by angiography compared with white males treated at majority hospitals (mixed: OR 1.7, P = .046; minority: OR 1.6, P = .08; Figure 4C).
There were no differences in time to angiography across races and hospital groups. However, minority hospitals had significantly longer times to operative intervention compared with white males treated at majority hospitals (CA males: risk ratio 1.5, P = .03; AA males: risk ratio 1.5, P = .02; Figure 4D).
Discussion
In the United States, racial disparities in outcomes and in healthcare processes have been documented for patients with a wide spectrum of illnesses.8 In our study, we demonstrated racial disparities in hospital mortality for AA (vs white) males with splenic injuries at majority, mixed-race, and minority hospitals. We found that, as compared with white males at majority hospitals, AA males had higher odds of mortality at mixed-race (60% higher) and minority hospitals (110% higher). Furthermore, within minority hospitals AA males had significantly higher mortality than white males.
In terms of early mortality (again, defined as death within 48 hours after hospital arrival), we found that, as compared with white males at majority hospitals, AA males had 34% higher odds of early mortality at majority hospitals, 61% higher odds at mixed-race hospitals, and 86% higher odds at minority hospitals. But we found no significant differences in early mortality for white males across hospital types.
Among the common injury mechanisms seen in the United States, gunshot wound is associated with the highest mortality rate. AA suffer a disproportionate burden of mortality from this mechanism.15,16 We used splenic injury which typically occurs following blunt injury from motor vehicle collisions or falls in this study so as to overcome the disproportionate burden of penetrating trauma in black men. Within majority hospitals, we found no statistically significant difference in mortality between CA and AA men. The differences in outcomes based on racial categorizations of hospitals may be attributable to hospital characteristics that are reflective of the socioeconomic status of the catchment population it serves. In other words, minority, mixed, and majority hospital designations serve as a proxy for quality of care.
Disparities in the quality of hospital care between white and AA males can be attributed to differences within or across hospitals, or perhaps both. Such disparities might be a symptom of racial segregation that unofficially but persistently occurs in the healthcare delivery system or by residence. It is possible that white and AA males are typically served by different hospital racial types (as defined in our study) and that minority hospitals provide lower-quality care than majority hospitals. Residential segregation likely plays a key role in segregating the healthcare delivery system, but other factors contribute. Historically, until the 1960s, hospitals were racially segregated in the south and in most northern cities.17,18 During the 1960s, sanctioned forms of hospital segregation were essentially eliminated, but in reality, segregation remains, partly due to social and economic pressures that are unique to health care. Other factors that can divide hospitals racially include racial differences in physician referrals, in transportation systems, in hospital ED capacity, and in patients’ preferences. Studies have shown that patients with high-grade solid organ injuries have improved outcomes when treated at tertiary trauma centers.14,19 Unfortunately we were unable to evaluate transfer practices between majority, mixed-race, and minority hospitals. Hospital segregation may, in turn, affect health outcomes and service utilization differently than does residential segregation, for example, through racial differences in providers’ medical practice patterns and in patients’ access to high-quality providers and specialized services.
This study further confirms the findings of Dimick et al20 revealing that AA patients are more likely to undergo surgery at low-quality hospitals, particularly in segregated regions. They found a strong relationship between residential segregation and the use of low-quality hospitals. This is supported by our findings that minority hospitals were significantly more likely to have failures from angiographic embolization requiring operative intervention compared with majority and mixed hospitals. In addition, we found in this study that AA males in mixed and minority hospitals with penetrating injuries were more likely to be managed by angiography compared with white males. This finding is concerning as the standard of care for patients with penetrating abdominal trauma is operative intervention and not angiography.
This study has several limitations. First, like all administrative databases, the exact method and accuracy of racial designation are unknown. We cannot tell whether racial categories are self-assigned or administratively designated. Second, though the racial categorization of hospitals used in this study has been previously utilized in other studies, we make assumptions as to the capabilities of the hospitals based on the arbitrary categorization.
Conclusion
While multiple studies have shown that minority hospitals have increased mortality compared to majority hospitals, this study found this disparity only existed for AAs. We were able to identify significant racial disparities in resource allocations. Future studies should characterize modifiable factors associated with disparities in resource allocations to guide interventions aimed at reducing disparities in health care.
Acknowledgments
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Mark R. Hemmila receives support from Blue Cross Blue Shield of Michigan and Blue Care Network (a nonprofit mutual company) for conduct of The Michigan Trauma Quality Improvement Program with a Collaborative Quality Initiatives grant. This research was funded by a grant awarded to CJT from the University of Minnesota’s Center for Healthy African American Men through Partnerships (CHAAMPS—NIH U54MD008620).For the remaining authors, no conflicts were declared.
Footnotes
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Mark R. Hemmila receives support from Blue Cross Blue Shield of Michigan and Blue Care Network (a nonprofit mutual company) for conduct of The Michigan Trauma Quality Improvement Program with a Collaborative Quality Initiatives grant.
References
- 1.Braboy Jackson P, Williams DR, Mullings AJ, Mullings AJ. The intersection of race, gender and SES: Health paradoxes. In: Schulz AJ, Mullings AJ, ed. Gender, Race, Class & Health: Intersectional Approaches. Jossey-Bass; 2006:131–162. [Google Scholar]
- 2.Courtenay W A global perspective on the field of men’s health: an editorial. Int J Mens Health. 2002; 1(1): 1-. doi: 10.3149/jmh.0101.1 [DOI] [Google Scholar]
- 3.Brown TH, Hargrove TW, Griffith DM. Racial/Ethnic disparities in men’s health: examining psychosocial mechanisms. Fam Community Health. 2015;38(4):307–318. doi: 10.1097/FCH.0000000000000080 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Gornick ME, Eggers PW, Reilly TW, et al. Effects of race and income on mortality and use of services among Medicare beneficiaries. N Engl J Med. 1996;335(11):791–799. doi: 10.1056/NEJM199609123351106 [DOI] [PubMed] [Google Scholar]
- 5.Williams DR. The health of men: structured inequalities and opportunities. Am J Public Health. 2003,93(5).724–73l. doi: 10.2105/AJPH.93.5.724 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Griffith DM, Ellis KR, Allen JO. An intersectional approach to social determinants of stress for African American men: men’s and women’s perspectives. Am J Mens Health. 2013;7(4 Suppl):19S–30. doi: 10.1177/1557988313480227 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Xanthos C, Treadwell HM, Holden KB. Social determinants of health among African—American men. J Mens health. 2010;7(1): 11–19. doi: 10.1016/j.jomh.2009.12.002 [DOI] [Google Scholar]
- 8.Smedley BD, Stith AY, Nelson AR. Unequal treatment: Confronting racial and ethnic disparities in health care. National Academies Press; 2003. [PubMed] [Google Scholar]
- 9.Haider AH, Ong’uti S, Efron DT, et al. Association between hospitals caring for a disproportionately high percentage of minority trauma patients and increased mortality: a nation-wide analysis of 434 hospitals. Arch Sing. 2012; 147( 1): 63–70. doi: 10.1001/archsurg.2011.254 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Maybury RS, Bolorunduro OB, Villegas C, et al. Pedestrians struck by motor vehicles further worsen race- and insurance-based disparities in trauma outcomes: the case for inner-city pedestrian injury prevention programs. Surgery. 2010; 148(2):202–208. doi: 10.1016/j.surg.2010.05.010 [DOI] [PubMed] [Google Scholar]
- 11.Haider AH, Chang DC, Efron DT, et al. Race and insurance status as risk factors for trauma mortality. Arch Sing. 2008;143(10):945–949. doi: 10.1001/archsurg.143.10.945 [DOI] [PubMed] [Google Scholar]
- 12.Haider AH, Saleem T, Leow JJ, et al. Influence of the National trauma data bank on the study of trauma outcomes: is it time to set research best practices to further enhance its impact? J Am Coll Sing. 2012;214(5):756–768. doi: 10.1016/j.jamcollsurg.2011.12.013 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Ingraham AM, Xiong W, Hemmila MR, et al. The attributable mortality and length of stay of trauma-related complications: a matched cohort study. Ann Surg. 2010;252(2): 358–362. doi: 10.1097/SLA.0b013e3181e623bf [DOI] [PubMed] [Google Scholar]
- 14.Tignanelli CJ, Joseph B, Jakubus JL, et al. Variability in management of blunt liver trauma and contribution of level of American College of surgeons Committee on trauma verification status on mortality. J Trauma Acute Care Surg. 2018;84(2):273–279. doi: 10.1097/TA.0000000000001743 [DOI] [PubMed] [Google Scholar]
- 15.WISQARS Injury Mortality Reports. 1999–2004. Accessed on 12/29/2019. http://webappa.cdc.gov/sasweb/ncipc/mortrate10_sy.html
- 16.Tignanelli CJ, Vander Kolk WE, Mikhail JN, Delano MJ, Hemmila MR. Noncompliance with American College of surgeons Committee on trauma recommended criteria for full trauma team activation is associated with undertriage deaths. J Trauma Acute Care Surg. 2018;84(2):287–294. doi: 10.1097/TA.0000000000001745 [DOI] [PubMed] [Google Scholar]
- 17.Halperin EC. Desegregation of hospitals and medical societies in North Carolina. N Engl J Med. 1988;318(1): 58–63. doi: 10.1056/NEJM198801073180127 [DOI] [PubMed] [Google Scholar]
- 18.Smith DB. Addressing racial inequities in health care: civil rights monitoring and report cards. J Health Polit Policy Law. 1998;23(1):75–105. doi: 10.1215/03616878-23-1-75 [DOI] [PubMed] [Google Scholar]
- 19.Adzemovic T, Murray T, Jenkins P, et al. Should they stay or should they go? who benefits from interfacility transfer to a higher-level trauma center following initial presentation at a lower-level trauma center. J Trauma Acute Care Surg. 2019;86(6):952–960. doi: 10.1097/TA.0000000000002248 [DOI] [PubMed] [Google Scholar]
- 20.Dimick J, Ruhter J, Sarrazin MV, Birkmeyer JD. Black patients more likely than whites to undergo surgery at low-quality hospitals in segregated regions. Health Aff. 2013;32(6): 1046–1053. doi: 10.1377/hlthaff.2011.1365 [DOI] [PMC free article] [PubMed] [Google Scholar]