Keywords
aortic arch surgery - social vulnerability index - health disparities - ethnicity
and surgical outcomes - access to care
Introduction
There have been several studies demonstrating the influence that ethnicity and social
vulnerability have on disease severity, morbidity, and mortality.[1]
[2]
[3]
[4] Such data highlight the importance of equitable access to care and underscores the
need for continued research to help mitigate disparities. While the individual impact
of different demographic and socioeconomic factors has been studied, few have investigated
the combined effects of the intersecting variables. Analyzing the cumulative impact
of these factors may provide insight into how social vulnerability affects different
ethnicities and provide a more targeted approach to improving access to care.
The CDC's social vulnerability index (SVI) quantifies neighborhood-level disparity,
with a higher SVI indicating more social vulnerability. This measure has been used
to study the impact of social vulnerability on surgical outcomes in trauma, coronary
artery disease, and abdominal aortic aneurysms; however, it has not been applied to
aortic arch surgery.[5]
[6]
[7]
[8] Given the heterogeneity across racial groups and socioeconomic classes, it is of
paramount importance to evaluate their impact in aortic surgery.
The purpose of this study is to investigate differences in presentation, outcomes,
and follow-up across different racial and socioeconomic groups in aortic arch surgery.
Furthermore, performing simultaneous analysis of race and the individual and cumulative
effects of SVI will better define presentation and subsequent outcomes. The results
of this study will identify and help develop tailored approaches to expand equitable
care for aortic surgery patients.
Materials and Methods
A retrospective review of a single institution prospectively maintained aortic database
was performed for all patients who underwent aortic arch surgery from 2011 to 2022.
All adults (age 18 years or older) who underwent open aortic arch surgery, including
hemiarch, total arch, or other aortic arch replacement were included. In total, 837
patients were identified. All data were collected through manual review of these medical
records in compliance with policies of the institution. The study design was reviewed
and approved for exception by the Colorado Multiple Institutional Review Board (COMIRB
#17-0198, approval date: February 6, 2017).
Data compiled within this registry include data submitted to the Society of Thoracic
Surgeons (STS) as well as other unique variables, including demographic characteristics
as well as preoperative, intraoperative, and outcome variables. In total, 92 variables
were included in the analyses. Preoperative variables, including age, sex, body mass
index, history of medical comorbidities such as hyperlipidemia, hypertension, type
2 diabetes mellitus, chronic kidney disease, pulmonary disease, coronary artery disease,
and smoking history were included. Ethnicity was obtained from the medical record
as self-reported by patients. Additionally, baseline variables including systolic
and diastolic blood pressures, international normalized ratio (INR), creatinine (Cr),
and hemoglobin A1C were obtained. For elective patients, INR, Cr, and blood pressures
were obtained from preoperative clinic visits or laboratory appointments. For nonelective
patients, these values were taken from either primary care or other provider visits
within 3 months of surgery. If these variables were not obtainable within 3 months
prior to presentation, these values were obtained from vitals or laboratories at the
time of initial presentation if there was no evidence of malperfusion, shock, or hypertensive
crisis. In the few nonelective cases where patients did not have any prior clinical
assessment within 3 months of operation and presented in any of the above conditions,
these values were excluded from analysis. Baseline hemoglobin A1c was obtained either
from preoperative assessments or from initial postoperative laboratories (within first
24 hours of surgery).
For operative and postoperative variables, these were obtained through manual review
of operative and medical records. Regarding postoperative in-hospital morbidity and
mortality, postoperative stroke was defined as new-onset neurological deficit lasting
>24 hours with imaging evidence of infarction or hemorrhage. Operative mortality was
based off the STS definition and included all deaths occurring within the index hospital
stay or death within 30 days of the procedure if discharged. New renal replacement
therapy requirement included acute kidney injury requiring hemodialysis.
Additionally, patient follow-up with a cardiovascular provider postdischarge was assessed,
along with a composite of cardiovascular- or procedure-related emergency department
(ED) visits as well as unplanned readmissions within one year of surgery. In addition
to reviewing patient-provided information, CareEverywhere was also searched for these
variables to ensure thorough data collection. Cardiovascular-related presentations
were defined as any presentation related to concern for arrhythmia, myocardial infarction,
heart failure, new-onset renal failure, hypertensive crisis, or thromboembolic event.
Procedure-related presentations included postoperative infections, pain associated
with the operation or baseline aortic pathology, hemorrhagic events related to postprocedure
anticoagulation, or unplanned readmissions for the management of aortic pathology.
Social Vulnerability Index Assessment
This study implemented custom built Python code (version 3.12.2, Python Software Foundation)
to develop an automated system for SVI calculation that replicates the manual process
available via the interactive CDC SVI map. The system utilized the 2020 SVI census
tract dataset in conjunction with geocoded address data from the patient population
to evaluate social vulnerability.[9] Address data were matched to SVI data using Federal Information Processing Standards
(FIPS) codes, which were derived from geospatial longitude and latitude coordinates.
Custom data processing techniques were employed to align geocoded patient address
data with the validated CDC SVI dataset. These techniques included the use of regular
expressions for sophisticated text data manipulation and numerical transformation
methods to ensure accurate data alignment. Regular expressions are a powerful tool
used for pattern matching and manipulation within text, enabling sophisticated search
and replace operations in strings of data. Geocoding was conducted using the U.S.
Census Bureau geocoder tool, which relies on the Master Address File/Topologically
Integrated Geographic Encoding and Referencing (MAF/TIGER) database to convert addresses
or location coordinates into geocoded information. Designed to facilitate efficient
and precise public geocoding, this tool provides results based on an address match
score. This geocoding tool yields interpolated coordinates by approximating the physical
location of an address within the TIGER database address ranges. FIPS codes were generated
in accordance with the National Institute of Standards and Technology (NIST) to precisely
match geocoded addresses with their relevant SVI metrics. The analysis identified
the SVI percentile ranking for each address, highlighting those ranked in the 75th
percentile or higher. This categorization process was designed to identify areas of
heightened vulnerability, thereby facilitating the analysis of potential correlations
between social determinants and health outcomes.
Analysis
Subgroup analyses were undertaken to elucidate the associations of ethnicity, SVI,
and their combination with various clinical and demographic outcomes. Continuous variables
were subjected to inferential statistics including analysis of variance for ethnicity
and combined ethnicity SVI groups, and the Mann–Whitney U test (also known as the
Wilcoxon rank-sum test) for differentiating SVI subgroups. Binary variables underwent
hypothesis testing with Fisher's exact tests across all aforementioned groups, establishing
a significance level at a p-value of ≤0.05.
Comparative assessments were performed on a range of demographic and clinical variables,
including operation urgency, history of aortic dissection, tobacco smoking, and incidence
of total arch replacements. This involved contrasting Caucasian against African American
subgroups, as well as high SVI (≥75%) against low SVI (<75%) categories. Furthermore,
a cross-analysis considering both ethnicity and SVI compared Caucasian patients with
low SVI to their African American counterparts with high SVI values. Contingency tables
were constructed to calculate odds ratios and p-values via the Fisher's exact test. Odds ratios were accompanied by 95% confidence
intervals computed using the standard error of the log of the odds ratio. To present
these associations visually, a forest plot was generated, highlighting the effect
sizes and corresponding confidence intervals for each comparison.
Results
Regarding racial and ethnic distribution of the cohort, 76.6% (641/837) identified
as Caucasian, 9.7% (81/837) identified as Black, 8.7% (73/837) identified as Hispanic,
2.4% (20/837) identified as Asian, and 2.6% (22/837) identified as Other. Demographic
city data compared with included patients were 16.6 versus 9.7% for the African American
cohort, 6.6 versus 2.4% for the Asian cohort, 43.5 versus 76% for the Caucasian cohort,
12 versus 8.7% for the Hispanic cohort, and 22 versus 2.6% for the Other cohort. The
patient cohort was not consistent with city demographics, with under representation
of non-Caucasian minorities in the patient cohort.
As seen in [Table 1], Regarding ethnicity only, African American and Hispanic individuals presented at
a younger age (p = 0.001). African American patients also presented with higher baseline systolic
(p = 0.002) and diastolic blood pressures (p < 0.001) and were significantly more likely to present urgently or emergently (p = 0.002) with aortic dissection pathology (p = 0.003). Regarding intraoperative characteristics, African Americans and Asians
required longer cardiopulmonary bypass (CPB; p = 0.029), aortic cross-clamp (p = 0.012), and circulatory arrest times (p = 0.005). Postoperatively, African Americans were more likely to have procedure-related
ED presentations within 1 year of surgery (p < 0.001) as seen in [Fig. 1]. However, no statistically significant differences were seen for length of stay,
in-hospital morbidity, in-hospital mortality, readmission, or follow-up rates with
cardiovascular providers when investigating ethnicity alone.
As seen in [Table 2], Regarding the SVI analysis alone, SVI was categorized as <75% or SVI ≥ 75%. Preoperatively,
high SVI patients presented at a younger age (p = 0.007), had a history of smoking (p = 0.007), and presented urgently or emergently (p = 0.001) as compared with lower SVI patients. There were no differences in BMI, baseline
laboratory data, aortic presentation, past cardiothoracic or aortic surgical history,
or other comorbidities. Intraoperatively, patients with high SVI were more likely
to require total arch replacement (p = 0.048) and have significantly longer CPB times (p = 0.008), aortic cross clamp times (p = 0.048), and circulatory arrest times (p = 0.001). They also had a lower intraoperative nadir bladder temperature (p < 0.001). There was no significant difference in the amount of intraoperative blood
product transfusion. Postoperatively, high SVI patients required more mechanical circulatory
support (p = 0.025) and utilized the ED more frequently in the first year after surgery (p = 0.003). There were no significant differences in length of stay, intensive care
unit morbidity, follow-up rates, or mortality for patients in the higher SVI group.
As seen in [Fig. 2], several of the differences between ethnicities became statistically significant
when the cumulative impact of SVI and ethnicity was considered. Preoperatively, regardless
of SVI, African American and Hispanic patients presented at a younger age (p = 0.001), with high SVI patients in general more likely to present at a younger age
(p = 0.007). African American and high SVI Asian patients presented with higher baseline
systolic and diastolic blood pressures (p = 0.002). African American and high SVI patients, regardless of race, were significantly
more likely to present urgently or emergently (p < 0.001) with aortic dissection pathology (p = 0.006) as seen in [Fig. 3].
Fig. 1 Comparative analysis of mean arterial pressure (MAP) and procedure-related emergency
department (ED) presentations within 1 year, categorized by ethnicity.
Fig. 2 Cumulative impact of SVI and ethnicity on preoperative, intraoperative, and postoperative
characteristics. SVI, socioeconomic vulnerability index.
Intraoperatively, African American individuals had longer circulatory arrest times
than other ethnicities only at high SVI (p = 0.002). Regardless of ethnicity, patients with high SVI were more likely to require
total arch replacement (p = 0.048) and have significantly longer CPB times (p = 0.008), aortic cross clamp times (p = 0.048), and circulatory arrest times (p = 0.001). There were no differences observed between the groups in the use of procedures
involving stent grafts, such as frozen elephant trunks, nor were there disparities
associated with ethnicity or SVI in the application of techniques utilizing advanced
devices.
Postoperatively, significant differences were seen in the number of procedure-related
ED presentations within 1 year (p < 0.001), with notably high usage among African Americans regardless of SVI and low
usage among high SVI Asian patients. No significant differences were seen in rates
of readmission or follow-up with a cardiovascular provider.
In the multivariate logistic regression analysis, neither race nor SVI ≥ 75% were
independently associated with postoperative mortality or readmission following aortic
arch surgery. Urgent or emergent operative status emerged as a significant predictor
of increased postoperative mortality (odds ratio [OR]: 3.16, 95% confidence interval
[CI]: 1.44–6.93, p = 0.004), emphasizing the critical role of surgical acuity in patient outcomes ([Table 3]). Other intraoperative factors, including CPB time (OR: 1.02, 95% CI: 1.01–1.03,
p < 0.001) and nadir bladder temperature (OR: 1.22, 95% CI: 1.06–1.41, p = 0.006), were significantly associated with mortality risk, whereas cross-clamp
time showed a mild protective effect (OR: 0.99, 95% CI: 0.98–1.00, p = 0.005). Variables such as BMI and comorbidities demonstrated limited or borderline
significance, underscoring the multifactorial nature of outcomes in this patient population.
These findings suggest that intraoperative management strategies and the urgency of
surgical presentation are pivotal in mitigating mortality risk.
Table 1
Analysis of preoperative, intraoperative, and postoperative variables by race and
ethnicity
|
Variable
|
Overall
|
African American
|
Asian
|
Caucasian
|
Hispanic
|
Other
|
p-Value
|
|
Sample size
|
837
|
81 (9.7%)
|
20 (2.4%)
|
641 (76.6%)
|
73 (8.7%)
|
22 (2.6%)
|
N/A
|
|
Preoperative characteristics
|
|
Age
|
61 [50, 69]
|
57 [45, 65]
|
61 [49, 73]
|
62 [52, 70]
|
57 [46, 67]
|
60 [46, 69]
|
0.002
|
|
Sex (male)
|
599 (71.6%)
|
65 (80.2%)
|
14 (70.0%)
|
455 (71.0%)
|
52 (71.2%)
|
13 (59.1%)
|
0.307
|
|
BMI
|
27.6 [24.3, 31.9]
|
28.0 [24.2, 32.3]
|
23.0 [20.6, 25.7]
|
27.6 [24.5, 31.9]
|
27.5 [24.3, 30.7]
|
26.7 [22.4, 33.3]
|
0.035
|
|
Baseline systolic BP
|
130 [117, 144]
|
140 [125, 156]
|
129 [121, 144]
|
129 [116, 142]
|
130 [118, 144]
|
130 [102, 147]
|
0.002
|
|
Baseline diastolic BP
|
75 [67, 86]
|
80 [70, 95]
|
80 [75, 85]
|
75 [67, 85]
|
74 [70, 86]
|
70 [58, 82]
|
<0.001
|
|
HTN
|
575 (68.7%)
|
67 (82.7%)
|
16 (80.0%)
|
429 (66.9%)
|
46 (63.0%)
|
17 (77.3%)
|
0.023
|
|
Smoking
|
203 (24.3%)
|
23 (28.4%)
|
6 (30.0%)
|
154 (24.0%)
|
17 (23.3%)
|
3 (13.6%)
|
0.641
|
|
Renal disease
|
88 (10.5%)
|
16 (19.8%)
|
3 (15.0%)
|
60 (9.4%)
|
4 (5.5%)
|
5 (22.7%)
|
0.007
|
|
Aortic presentation
|
|
|
|
|
|
|
|
|
Aneurysm
|
619 (74.0%)
|
52 (64.2%)
|
16 (80.0%)
|
482 (75.2%)
|
54 (74.0%)
|
15 (68.2%)
|
0.260
|
|
Dissection
|
289 (34.5%)
|
44 (54.3%)
|
7 (35.0%)
|
207 (32.3%)
|
23 (31.5%)
|
8 (36.4%)
|
0.003
|
|
Operative urgency
|
|
|
|
|
|
|
|
|
Elective
|
525 (62.7%)
|
34 (42.0%)
|
11 (55.0%)
|
418 (65.2%)
|
47 (64.4%)
|
15 (68.2%)
|
0.002
|
|
Urgent emergent
|
312 (37.3%)
|
47 (58.0%)
|
9 (45.0%)
|
223 (34.8%)
|
26 (35.6%)
|
7 (31.8%)
|
0.002
|
|
Intraoperative characteristics
|
|
Procedure type
|
|
|
|
|
|
|
|
|
Hemiarch
|
602 (71.9%)
|
52 (64.2%)
|
12 (60.0%)
|
473 (73.8%)
|
51 (69.9%)
|
14 (63.6%)
|
0.214
|
|
Total arch
|
235 (28.1%)
|
29 (35.8%)
|
8 (40.0%)
|
168 (26.2%)
|
22 (30.1%)
|
8 (36.4%)
|
0.214
|
|
Operative variables
|
|
|
|
|
|
|
|
|
CPB time
|
156 [123, 211]
|
170 [129, 234]
|
158 [124, 189]
|
152 [121, 208]
|
162 [133, 210]
|
195 [146, 239]
|
0.029
|
|
Cross-clamp time
|
98 [71, 135]
|
111 [79, 162]
|
91 [63, 127]
|
98 [71, 133]
|
97 [66, 135]
|
110 [89, 171]
|
0.016
|
|
HCA time
|
13 [8, 23]
|
17 [10, 26]
|
15 [11, 18]
|
12 [8, 22]
|
16 [8, 27]
|
18 [13, 32]
|
0.005
|
|
Nadir bladder temperature
|
26.9 [25.0, 28.0]
|
26.1 [23.5, 27.7]
|
26.9 [25.5, 27.9]
|
27.1 [25.3, 28.0]
|
26.6 [24.0, 27.8]
|
26.0 [24.4, 28.0]
|
0.115
|
|
Nadir hemoglobin
|
8.5 [7.5, 10.0]
|
8.4 [7.5, 9.2]
|
7.8 [7.4, 8.5]
|
8.6 [7.5, 10.1]
|
8.4 [7.7, 10.4]
|
7.7 [7.3, 8.1]
|
0.015
|
|
Postoperative characteristics
|
|
Length of stay
|
8.0 [6.0, 13.0]
|
9.0 [7.0, 14.0]
|
9.0 [8.0, 13.3]
|
8.0 [6.0, 13.0]
|
7.0 [6.0, 10.0]
|
9.0 [7.0, 12.0]
|
0.490
|
|
ICU length of stay
|
3.0 [2.0, 6.0]
|
3.0 [2.0, 6.0]
|
4.0 [3.0, 5.5]
|
3.0 [2.0, 6.0]
|
3.0 [1.3, 4.0]
|
4.0 [3.0, 6.5]
|
0.581
|
|
Postoperative mortality
|
63 (7.5%)
|
5 (6.2%)
|
2 (10.0%)
|
46 (7.2%)
|
7 (9.6%)
|
3 (13.6%)
|
0.712
|
|
Postoperative readmission
|
229 (27.4%)
|
32 (39.5%)
|
4 (20.0%)
|
166 (25.9%)
|
22 (30.1%)
|
5 (22.7%)
|
0.113
|
|
ED presentations in 1 y
|
0.0 [0.0, 1.0]
|
1.0 [0.0, 1.0]
|
0.0 [0.0, 0.0]
|
0.0 [0.0, 1.0]
|
0.0 [0.0, 1.0]
|
0.0 [0.0, 0.8]
|
<0.001
|
|
Follow-up CV
|
728 (87.0%)
|
73 (90.1%)
|
18 (90.0%)
|
555 (86.6%)
|
63 (86.3%)
|
19 (86.4%)
|
0.648
|
|
Postoperative late death
|
47 (5.6%)
|
3 (3.7%)
|
1 (5.0%)
|
39 (6.1%)
|
4 (5.5%)
|
0 (0.0%)
|
0.708
|
Abbreviations: BP, blood pressure; CPB, cardiopulmonary bypass; CV, cardiovascular;
HCA, hypothermic circulatory arrest; HTN, hypertension; ICU, intensive care unit;
N/A, not applicable.
Note: Values are n (%) or median (25–75% interquartile range).
Table 2
Analysis of preoperative, intraoperative, and postoperative variables by socioeconomic
vulnerability index
|
Variable
|
Overall
|
SVI < 75%
|
SVI ≥ 75%
|
p-Value
|
|
Sample size
|
837
|
682 (81.5%)
|
155 (18.5%)
|
N/A
|
|
Preoperative characteristics
|
|
Age
|
61 [50, 69]
|
62 [51, 70]
|
59 [47, 67]
|
0.007
|
|
BMI
|
27.6 [24.3, 31.9]
|
27.5 [24.3, 31.6]
|
28.0 [24.0, 32.4]
|
0.247
|
|
Baseline systolic BP
|
130 [117, 144]
|
130 [117, 143]
|
129 [114, 147]
|
0.670
|
|
Baseline diastolic BP
|
75 [67, 86]
|
75 [67, 86]
|
77 [67, 87]
|
0.425
|
|
Smoking
|
203 (24.3%)
|
152 (22.3%)
|
51 (32.9%)
|
0.007
|
|
Operative urgency
|
|
|
|
|
|
Elective
|
525 (62.7%)
|
447 (65.5%)
|
78 (50.3%)
|
<0.001
|
|
Urgent emergent
|
312 (37.3%)
|
235 (34.5%)
|
77 (49.7%)
|
<0.001
|
|
Intraoperative characteristics
|
|
Procedure type
|
|
|
|
|
|
Hemiarch
|
602 (71.9%)
|
501 (73.5%)
|
101 (65.2%)
|
0.048
|
|
Total arch
|
235 (28.1%)
|
181 (26.5%)
|
54 (34.8%)
|
0.048
|
|
Adjunctive root procedure
|
307 (36.7%)
|
251 (36.8%)
|
56 (36.1%)
|
0.948
|
|
No adjunctive structural procedure
|
191 (22.8%)
|
144 (21.1%)
|
47 (30.3%)
|
0.018
|
|
Operative variables
|
|
|
|
|
|
CPB time
|
156 [123, 211]
|
152 [121, 208]
|
178 [130, 232]
|
0.008
|
|
Cross-clamp time
|
98 [71, 135]
|
97 [71, 134]
|
103 [72, 143]
|
0.048
|
|
HCA time
|
13 [8, 23]
|
12 [8, 22]
|
16 [10, 26]
|
<0.001
|
|
Nadir bladder temperature
|
26.9 [25.0, 28.0]
|
27.1 [25.3, 28.0]
|
26.0 [23.8, 27.8]
|
<0.001
|
|
Nadir hemoglobin
|
8.5 [7.5, 10.0]
|
8.5 [7.6, 10.0]
|
8.40 [7.3, 9.8]
|
0.532
|
|
Postoperative characteristics
|
|
Length of stay
|
8.0 [6.00, 13.0]
|
8.0 [6.0, 13.0]
|
9.0 [7.0, 14.0]
|
0.053
|
|
ICU length of stay
|
3.0 [2.0, 6.0]
|
3.0 [2.0, 6.0]
|
4.0 [2.0, 6.0]
|
0.115
|
|
New hemodynamic support
|
47 (5.6%)
|
32 (4.7%)
|
15 (9.7%)
|
0.025
|
|
Postoperative mortality
|
63 (7.5%)
|
48 (7.0%)
|
15 (9.7%)
|
0.339
|
|
Postoperative readmission
|
229 (27.4%)
|
185 (27.1%)
|
44 (28.4%)
|
0.575
|
|
ED presentations in 1 y
|
0.0 [0.0, 1.0]
|
0.0 [0.0, 1.0]
|
0.0 [0.0, 1.0]
|
0.003
|
|
Follow-up CV
|
728 (87.0%)
|
599 (87.8%)
|
129 (83.2%)
|
0.192
|
|
Postoperative late death
|
47 (5.6%)
|
37 (5.4%)
|
10 (6.5%)
|
0.758
|
Abbreviations: BP, blood pressure; CPB, cardiopulmonary bypass; CV, cardiovascular;
HCA, hypothermic circulatory arrest; HTN, hypertension; ICU, intensive care unit.
Note: Values are n (%) or median (25–75% interquartile range).
Table 3
Multivariate logistic regression analysis evaluating predictors of postoperative mortality
following aortic arch surgery
|
Variable
|
OR (95% CI)
|
p-Value
|
|
Patient characteristics
|
|
|
|
Age
|
1.02 (1.00, 1.05)
|
0.057
|
|
BMI
|
0.94 (0.88, 0.99)
|
0.029
|
|
Baseline systolic BP
|
1.01 (0.99, 1.02)
|
0.517
|
|
Baseline diastolic BP
|
0.99 (0.97, 1.01)
|
0.445
|
|
SVI ≥ 75%
|
1.12 (0.49, 2.57)
|
0.794
|
|
Race
|
|
|
|
African American
|
3.46 (0.45, 26.43)
|
0.232
|
|
Asian
|
3.04 (0.39, 23.64)
|
0.288
|
|
Caucasian
|
2.15 (0.57, 8.20)
|
0.261
|
|
Hispanic
|
3.12 (0.65, 14.89)
|
0.154
|
|
Comorbidities
|
|
|
|
HTN
|
1.19 (0.56, 2.57)
|
0.649
|
|
Smoking
|
1.19 (0.59, 2.38)
|
0.628
|
|
Renal disease
|
0.99 (0.38, 2.58)
|
0.977
|
|
Aortic presentation
|
|
|
|
Aortic aneurysm
|
0.77 (0.39, 1.53)
|
0.452
|
|
Aortic dissection
|
1.25 (0.58, 2.73)
|
0.571
|
|
Operative urgency
|
|
|
|
Urgent/emergent
|
3.16 (1.44, 6.93)
|
0.004
|
|
Operative variables
|
|
|
|
CPB time
|
1.02 (1.01, 1.03)
|
<0.001
|
|
Cross-clamp time
|
0.99 (0.98, 1.00)
|
0.005
|
|
HCA time
|
1.02 (0.99, 1.04)
|
0.189
|
|
Nadir bladder temperature
|
1.22 (1.06, 1.41)
|
0.006
|
Notes: Independent variables included race and the socioeconomic vulnerability index
(SVI ≥ 75%). Confounders included age, body mass index (BMI), baseline systolic and
diastolic blood pressure, comorbidities (hypertension, smoking, renal disease), aortic
presentation (aneurysm, dissection), operative urgency (urgent/emergent), and operative
variables (cardiopulmonary bypass [CPB] time, cross-clamp time, hypothermic circulatory
arrest [HCA] time, and nadir bladder temperature). Statistically significant variables
included urgent or emergent operative status (OR: 3.16, 95% CI: 1.44–6.93, p = 0.004), CPB time (OR: 1.02, 95% CI: 1.01–1.03, p < 0.001), cross-clamp time (OR: 0.99, 95% CI: 0.98–1.00, p = 0.005), and nadir bladder temperature (OR: 1.22, 95% CI: 1.06–1.41, p = 0.006). Age and BMI demonstrated borderline significance.
Postoperative readmission analysis revealed a complex interplay of demographic and
intraoperative factors ([Table 4]). Lower BMI was inversely associated with readmission risk (OR: 0.97, 95% CI: 0.94–1.00,
p = 0.025), whereas longer CPB times were linked to higher readmission rates (OR: 1.01,
95% CI: 1.00–1.01, p = 0.024). Additionally, race demonstrated significant associations, with African
American (OR: 0.24, 95% CI: 0.06–0.96, p = 0.043) and Caucasian (OR: 0.55, 95% CI: 0.31–0.97, p = 0.039) populations experiencing lower odds of readmission compared with other groups.
Notably, the modest protective effect of prolonged cross-clamp time (OR: 0.99, 95%
CI: 0.99–1.00, p = 0.034) warrants further investigation to clarify its clinical implications. Together,
these findings highlight the importance of demographic considerations and tailored
intraoperative strategies in reducing readmission risks, supporting a personalized
approach to postoperative care.
Table 4
Multivariate logistic regression analysis evaluating predictors of postoperative readmission
following aortic arch surgery
|
Variable
|
OR (95% CI)
|
p-Value
|
|
Patient characteristics
|
|
|
|
Age
|
1.00 (0.99, 1.02)
|
0.634
|
|
BMI
|
0.97 (0.94, 1.00)
|
0.025
|
|
Baseline systolic BP
|
1.00 (0.99, 1.01)
|
0.729
|
|
Baseline diastolic BP
|
0.99 (0.97, 1.00)
|
0.091
|
|
SVI ≥ 75%
|
1.18 (0.75, 1.85)
|
0.487
|
|
Race
|
|
|
|
African American
|
0.24 (0.06, 0.96)
|
0.043
|
|
Asian
|
0.26 (0.07, 1.01)
|
0.051
|
|
Caucasian
|
0.55 (0.31, 0.97)
|
0.039
|
|
Hispanic
|
0.73 (0.34, 1.58)
|
0.429
|
|
Comorbidities
|
|
|
|
HTN
|
1.34 (0.90, 2.00)
|
0.155
|
|
Smoking
|
0.95 (0.63, 1.41)
|
0.788
|
|
Renal disease
|
1.51 (0.90, 2.55)
|
0.120
|
|
Aortic presentation
|
|
|
|
Aortic aneurysm
|
0.82 (0.53, 1.28)
|
0.387
|
|
Aortic dissection
|
1.21 (0.75, 1.95)
|
0.441
|
|
Operative urgency
|
|
|
|
Urgent/emergent
|
0.92 (0.58, 1.46)
|
0.721
|
|
Operative variables
|
|
|
|
CPB time
|
1.01 (1.00, 1.01)
|
0.024
|
|
Cross-clamp time
|
1.00 (0.99, 1.00)
|
0.074
|
|
HCA time
|
1.00 (0.98, 1.02)
|
0.965
|
|
Nadir bladder temperature
|
1.06 (0.97, 1.16)
|
0.176
|
Notes: Independent variables included race and the socioeconomic vulnerability index
(SVI ≥ 75%). Confounders included age, body mass index (BMI), baseline systolic and
diastolic blood pressure, comorbidities (hypertension, smoking, renal disease), aortic
presentation (aneurysm, dissection), operative urgency (urgent/emergent), and operative
variables (cardiopulmonary bypass [CPB] time, cross-clamp time, hypothermic circulatory
arrest [HCA] time, and nadir bladder temperature). The table presents odds ratios
(OR) with 95% confidence intervals (CI) and p-values. Significant predictors of postoperative readmission included lower BMI (OR:
0.97, 95% CI: 0.94–1.00, p = 0.025), African American race (OR: 0.24, 95% CI: 0.06–0.96, p = 0.043), Caucasian race (OR: 0.55, 95% CI: 0.31–0.97, p = 0.039), and longer CPB time (OR: 1.01, 95% CI: 1.00–1.01, p = 0.024).
Discussion
The results of the study demonstrate a clear lack of access to care for underrepresented
groups. The demographics of the patient cohort are not consistent with the surrounding
city's demographics, with an underrepresentation of non-Caucasian ethnicities. Ethnicity
alone plays a statistically significant role in patient presentation, intraoperative
characteristics, and postoperative ED utilization. For example, African Americans
present with a higher baseline mean arterial pressure (MAP), present urgent or emergently,
and are more likely to present with dissection pathology. African American and Hispanic
individuals present at a younger age, and African Americans are also more likely to
have procedure-related ED presentations within 1 year of surgery, as compared with
other ethnicities. However, when investigating ethnicity alone, no statistically significant
differences are seen for in-hospital morbidity, in-hospital mortality, readmission,
or follow-up rates with cardiovascular providers.
Analysis based on SVI alone demonstrates that high SVI patients are vulnerable regardless
of ethnicity. Regarding patient presentation, high SVI patients are more likely to
present younger, present urgently or emergently, and have a history of smoking. Furthermore,
they are more likely to present with more extensive pathology requiring a total arch
replacement with longer operative times. Postoperatively, they more frequently require
mechanical circulatory support and utilize the ED more frequently postdischarge. However,
there is otherwise no significant difference in other morbidity or mortality for high
SVI patients.
The data further demonstrate that several of the differences between ethnicities become
statistically significant when the cumulative impact of SVI and ethnicity is considered,
particularly as it relates to patient presentation, circulatory arrest time, and ED
utilization postdischarge. This clearly establishes the importance of investigating
the combined effect of different demographic variables, as analyzing these variables
individually can hide important findings.
The results of this study were consistent with the findings of several prior studies.
The finding that non-Caucasian patients present at a younger age, more urgently or
emergently with a higher comorbidity burden has been demonstrated in previous research.[2]
[10] The current study supports those findings and indicates that specifically African
American and Hispanic individuals present at a younger age, regardless of SVI. Our
findings further support that high SVI patients are more likely to present younger,
present urgently or emergently, and have a history of smoking.[2]
[11] The current study also supports that African American patients present more urgently
or emergently and more frequently present with dissection pathology, regardless of
SVI. A new finding from this study demonstrates that African American and Asian patients
of all SVIs present with a higher baseline MAP. This study also demonstrated that
high SVI across all ethnicities is associated with increased urgent and emergent presentation.
The data further indicate that Asian patients with high SVI also present with dissection
pathology at a higher rate than other ethnicities. These differences at presentation
highlight the lack of access to care for certain ethnicities and high SVI groups.
Interventions that improve access to preventative care and increase screening for
aortic disease should be implemented and may have substantial positive impact.
Prior studies have shown that African American individuals were more likely to have
more complex procedures with longer intraoperative times and longer circulatory arrest
times.[10] However, this study shows that this difference is only statistically significant
at high SVI. Our data demonstrate that when analyzing SVI alone, patients with high
SVI are more likely to require total arch replacement leading to longer operative
times.
Consistent with prior data on other proximal aortic surgery, this study demonstrates
that SVI and ethnicity do not lead to significant difference for in-house morbidity
and mortality postoperatively.[2]
[10]
[11] To further that, this study was the first to investigate follow-up rates with cardiovascular
providers and rates of readmission, finding no significant differences. The lack of
significant differences in in-hospital morbidity, mortality, readmission, and follow-up
also demonstrates that once patients present to the hospital, they are receiving equitable
care. The study further indicates that regardless of SVI, ED utilization 1 year postoperatively
is higher for African Americans populations, but it does not lead to increased readmission
rates. Overall, patients with high SVI alone also utilize the ED more frequently in
the 1-year postoperative period. The increased ED utilization for these groups points
to inequities in access to care that force these individuals to present directly to
the ED for concerns not requiring readmission.
Our multivariate logistic regression analysis underscores the critical role of operative
urgency and intraoperative variables in shaping postoperative outcomes. Urgent or
emergent operative status significantly increased mortality risk, emphasizing the
importance of preoperative optimization and judicious patient selection in these high-risk
scenarios. Intraoperative factors also influenced readmission rates, with longer CPB
times associated with higher readmission risk. These findings suggest that efforts
to minimize bypass duration could mitigate postoperative complications and improve
recovery trajectories.
While race and SVI were not independent predictors of mortality or readmission, this
does not diminish the broader systemic and socioeconomic disparities highlighted in
prior research. These results instead reinforce the importance of focusing on modifiable
surgical and perioperative factors to improve outcomes universally, regardless of
demographic or socioeconomic differences. Future studies should explore additional
endpoints and potential confounders to better understand these complex relationships
in diverse populations.
Several limitations to the study need to be acknowledged. First, while the study had
a large overall sample size, specific cohorts within the dataset, such as the number
of Asian patients, still have a small sample size of 20. Also, this is still a single-center
study and thus subject to intrinsic limitations. The specific surgical techniques
at this institution, and the access to care in and around the hospital and state,
may limit the generalizability of the results. Furthermore, while all steps were taken
to ensure thorough data collection for all cardiovascular provider follow-up visits,
ED visits as well as unplanned readmissions within 1 year of surgery, including searching
CareEverywhere, given that not all institutions participate in CareEverywhere it is
possible that their frequency of visits and readmissions is underestimated. Additionally,
SVI is calculated based on street address and five-digit zip code and is limited to
neighborhood-level social vulnerability. It inherently generalizes the heterogeneity
within that region as it is unable to factor in individual patient considerations.
Given that SVI is calculated based on the census tract of an individual, it only provides
the social vulnerability of a patient at a specific point in time and cannot capture
any real-time changes. Despite its limitations, the use of neighborhood-level vulnerability
indices, such as SVI, is a validated methodology frequently used in studies investigating
health care disparities.[4]
[5]
[6]
[7]
[8] Another limitation of the study is the 11-year duration. Demographic and socioeconomic
changes likely occurred in each ZIP code over that period, but since the SVI reflects
the vulnerability of the neighborhood at the time the study was performed, it may
not truly reflect the social disparities at the time of surgery for the earlier data.
Lastly, it is important to consider that the study only captures patients who presented
to the institution for an intervention. It is unable to account for access disparities
that resulted in patient death or inoperable pathology.
Despite the limitations, this study emphasizes the importance of equitable access
to health care for all patients undergoing aortic arch surgery. It demonstrates the
importance of considering the cumulative effect of socioeconomic status and ethnicity,
given analyzing them as individual variables often hides significance. The data provide
insight into how social vulnerability affects different ethnicities and will help
develop more targeted approaches to improving access to care. Specifically, it highlights
the lack of access to preventative care faced by certain ethnicities and high SVI
populations, given the significant differences in initial presentation. The study
also sheds light on groups of individuals who may be at increased risk of certain
pathologies at a younger age, necessitating early, targeted screenings. Furthermore,
the data compel us to contemplate which interventions may improve postoperative access
to care, so certain populations will not be forced to utilize the ED as their primary
access point for health care. Further investigation is needed to determine specific
barriers to equitable access to care, and what interventions, if any, have worked
to mitigate these inequities.
Conclusion
Clear lack of access to care exists for underrepresented groups as demonstrated by
a patient population not reflective of city demographics, higher surgical acuity in
socially vulnerable patients, and trends in ED usage after discharge. Furthermore,
the ethnicity-only dataset hid significant differences within ethnicities between
normal and high SVI groups, emphasizing the importance of considering the cumulative
impact of both SVI and ethnicity. In contrast, patients belonging to minority groups,
such as African American patients, also had apparent differences in presentation regardless
of SVI, warranting further investigation to determine cause. Most importantly, approaches
to expanding care need to be racially sensitive, applicable at all levels of care,
and targeted toward high SVI groups.
Fig. 3 Forest plot depicting the odds ratios (OR) with 95% confidence intervals (CI) for
various risk factors, stratified by ethnicity, socioeconomic vulnerability index (SVI),
and their combination. The Ethnicity comparison demonstrates the differences in risk
between African American versus Caucasian patients for urgent/emergent procedures
and dissections. The SVI comparison depicts the differences between SVI ≥ 75 versus
SVI < 75 patients for urgent/emergent procedures, total arch surgeries, and smoking
status. The combination of ethnicity and SVI comparison demonstrates the differences
between African American patients with SVI ≥ 75 versus Caucasian patients with SVI < 75
for urgent/emergent procedures and dissections.
