Does Hispanic ethnicity play a role in outcomes for diverticular surgery in the USA?

Abstract

Objective

The aim of this study is to investigate whether origins of ethnicity affect the outcomes of surgery for diverticulitis in the USA.

Design

The American College of Surgeons National Surgical Quality Improvement Programme database from 2008 to 2017 was used to identify patients undergoing colectomy for diverticulitis. Patient demographics, comorbidities, procedural details and outcomes were captured and compared by ethnicity status.

Results

A total of 375 311 surgeries for diverticulitis were included in the final analysis. The average age of patients undergoing surgery for diverticulitis remained consistent over the time frame of the study (62 years), although the percentage of younger patients (age 18–39 years) rose slightly from 7.8% in 2008 to 8.6% in 2017. The percentage of surgical patients with Hispanic ethnicity increased from 3.7% in 2008 to 6.6% of patients in 2017. Hispanic patients were younger than their non-Hispanic counterparts (57 years vs 62 years, p<0.01) at time of surgery. There were statistically significant differences in the proportion of laparoscopic cases (51% vs 49%, p<0.01), elective cases (62% vs 66%, p<0.01) and the unadjusted rate of postoperative mortality (2.8% vs 3.4%, p<0.01) between Hispanic patients compared with non-Hispanic patients, respectively. Multivariable logistic regression models did not identify Hispanic ethnicity as a significant predictor for increased morbidity (p=0.13) or mortality (p=0.80).

Conclusion

Despite a significant younger population undergoing surgery for diverticulitis, Hispanic ethnicity was not associated with increased rates of emergent surgery, open surgery or postoperative complications compared with a similar non-Hispanic population.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Diverticulitis is a common disease in the USA with identifiable differences in management and outcomes following diverticular surgery in African Americans relative to Caucasian Americans.

WHAT THIS STUDY ADDS

• Patients who identify as Hispanic tend to have diverticular surgery at a significantly younger age relative to non-Hispanic individuals; however, there is no difference in case acuity or outcomes between the two populations.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• Future studies should focus on identifying why Hispanic patients require surgical management at a younger age compared with non-Hispanic patients.

Introduction

Diverticulitis is a common disease in the USA, encompassing roughly 2.7 million outpatient visits and 200 000 inpatient admissions annually.^{1 2} The incidence of diverticulitis has long been associated with increased age, but interestingly, incidence among younger age groups (18–49) has increased significantly over the past 30 years.³ In addition to increasing age, other well-supported risk factors implicated in diverticular disease include consumption of a western diet, obesity, smoking and low physical activity.^4–6

While identifiable differences in the management and outcomes after diverticular surgery exist among African Americans, there is a paucity of data regarding the growing Hispanic population in the USA.⁷ Anecdotally, we have noticed trends of younger, Hispanic patients with significant diverticulitis requiring surgery. We hypothesised that Hispanic patients may have more emergent surgery for diverticulitis relative to non-Hispanic patients and aimed to answer this question using American College of Surgeons (ACS) National Surgical Quality Improvement Programme (NSQIP) database.

Materials and methods

Using the ACS NSQIP database, a retrospective review of all patients aged 18 or older undergoing colectomy for diverticulitis between 2008 and 2017 was conducted using Current Procedural Terminology (CPT) and International Classification of Diseases (ICD) 9/10 codes. The ACS NSQIP focuses on collecting preoperative and postoperative data (through 30 days) on randomly assigned patients at participating hospitals throughout the USA. Paediatric patients, pregnant patients and patients with incomplete data collection were excluded from this study. The primary outcome of interest is the proportion of patients requiring emergent surgery. Secondary outcomes of interest include rates of postoperative morbidity and 30-day mortality. Morbidity was defined as having any of the following complications: surgical site infection, wound disruption, pneumonia, unplanned intubation, pulmonary embolism/deep venous thrombus, on ventilator greater than 48 hours, progressive renal insufficiency, acute renal failure, urinary tract infection, stroke, cardiac arrest requiring CPR, myocardial infarction, bleed, sepsis and septic.

Normally distributed numeric variables were expressed as mean±SD and analysed via two-sample independent t-test. Non-normally distributed numeric variables were expressed as median (25th, 75th percentile) and analysed via Wilcoxon rank sum. Categorical data were expressed as a frequency (per cent) and analysed via χ² analysis. Multivariable logistic regression models were created to look at morbidity and mortality. These models were created using a random 70% sample of the observations from the overall dataset which incorporated the variables patient sex, ethnicity, age, body mass index (BMI), diabetes, smoking status, chronic obstructive pulmonary disease (COPD), hypertension, steroid use, weight loss, albumin level, wound class, American Society of Anesthesiologists (ASA) class, operation time, length of stay, procedure acuity and approach type (laparoscopic vs open). The final models reported used a backwards selection process to find optimal models. The remaining 30% of the observations from the overall data set were used to validate the models. Statistical significance is set with an alpha value of 0.05.

The ACS NSQIP and the hospitals participating in the ACS NSQIP are the source of the data used here; they have not verified and are not responsible for the statistical validity of the data analysis or the conclusions derived by the authors.

Results

A total of 407 003 surgeries for diverticulitis were identified. Of those, 36 patients were pregnant, 17 patients were less than 18 years of age and 31 639 patients had incomplete data resulting in 375 311 patients meeting inclusion criteria. Over the course of the study period, the average age of the annual patient population decreased slightly from 63 years of age in 2008 to 62 years of age in 2017 while the percentage of younger patients (18–39 years old) increased from 7.8% to 8.6%. Annually, most patients were female (53%) and Caucasian.

The proportion of Caucasians decreased from 80.4% to 73.1% over the course of the study period, whereas the incidence in under-represented races (black, Asian, American Indian, other/unknown) increased. Hispanic ethnicity increased from 3.7% in 2008 to 6.6% in 2017. Patients who identified as having Hispanic ethnicity were younger than non-Hispanic patients at time of diverticulitis-related surgery (57 years vs 62 years, p<0.01, table 1). Analysis of case approach (laparoscopic vs open) and case type (emergent vs elective) revealed a statistically significant difference in both, with Hispanic patients undergoing more laparoscopic surgeries (51% vs 49.1%, p<0.01) but fewer elective surgeries (61.9% vs 65.9%, p<0.01) than their counterparts. There was no statistically significant difference between Hispanic and non-Hispanic patients regarding the proportion of patients who received an end ostomy at time of surgery (11.4% vs 11.0%, respectively, p=0.12). Using a multivariate logistic model (table 2), ethnicity was not a statistically significant predictor of morbidity (p=0.13) and 30-day mortality (p=0.80).

Table 1.

Demographic and perioperative variables among Hispanic and non-Hispanic patients

Variable	Missing	Non-Hispanic (N=320 803)	Hispanic (N=19 219)	P value
Age	0	62.2±15.1	57.0±15.8	<0.01
BMI	4412	27.4(23.7, 31.9)	28.3(24.9, 32.7)	<0.01
Diabetes	0	48 639 (15.2)	4054 (21.1)	<0.01
Current smoker	2	59 091 (18.4)	2897 (15.1)	<0.01
History COPD	0	20 656 (6.4)	544 (2.8)	<0.01
Hypertension requiring medication	0	162 794 (50.8)	8536 (44.4)	<0.01
Steroid use	0	26 604 (8.3)	1037 (5.4)	<0.01
Weight loss	0	15 181 (4.7)	922 (4.8)	0.68
Wound class	0			<0.01
Clean		3171 (1.0)	206 (1.1)
Clean/contaminated		228 316 (71.2)	13 116 (68.2)
Contaminated		40 564 (12.6)	2673 (13.9)
Dirty/infected		48 752 (15.2)	3224 (16.8)
ASA class	41			<0.01
1		7367 (2.3)	618 (3.2)
2		131 651 (41.0)	8657 (45.0)
3		150 905 (47.1)	8363 (43.5)
4		28 670 (8.9)	1456 (7.6)
5		2172 (0.7)	122 (0.6)
Prealbumin	91 777	3.7±0.7	3.7±0.7	0.01
Surgical approach	0			<0.01
Open		163 245 (50.9)	9413 (49.0)
Lap		157 558 (49.1)	9806 (51.0)
Procedure acuity	52 859			<0.01
Neither		43 146 (16.0)	3228 (19.0)
Emergent		49 084 (18.2)	3254 (19.1)
Elective		177 928 (65.9)	10 523 (61.9)
Total operation time	0	149.0 (106.0, 208.0)	157.0 (111.0, 221.0)	<0.01
LOS	0	6.0(4.0, 10.0)	6.0(4.0, 10.0)	0.27
30-day mortality	0	10 833 (3.4)	535 (2.8)	<0.01

ASA, American Society of Anesthesiologists; BMI, body mass index; COPD, chronic obstructive pulmonary disease; LOS, length of stay.

Table 2.

ORs of variables utilised in the morbidity and mortality regression models

Effect	Morbidity regression model	Mortality regression model
	OR with 95% CI	OR with 95% CI
Sex (female vs male)	1.05 (1.02 to 1.08)	0.83 (0.78 to 0.89)
Ethnicity (Hispanic vs not)	1.00 (1.00 to 1.01)	---
Age	1.01 (1.00 to 1.01)	1.04 (1.04 to 1.05)
BMI	0.92 (0.88 to 0.95)	---
Current smoker (yes vs no)	1.21 (1.16 to 1.28)	1.10 (1.01 to 1.19)
History of COPD (yes vs no)	1.08 (1.05 to 1.11)	1.58 (1.46 to 1.70)
History of steroid use (yes vs no)	1.14 (1.08 to 1.21)	1.40 (1.29 to 1.52)
History of weight loss (yes vs no)	0.74 (0.73 to 0.76)	1.40 (1.27 to 1.55)
Preoperative albumin	1.05 (1.02—1.08)	0.48 (0.46 to 0.50)
Wound class (clean-contaminated vs clean)	1.10 (0.96 to 1.25)	0.73 (0.55 to 0.97)
Wound class (contaminated vs clean)	1.33 (1.16 to 1.52)	0.95 (0.71 to 1.27)
Wound class (dirty infected vs clean)	1.82 (1.59 to 2.08)	1.03 (0.77 to 1.36)
ASA classification (2 vs 1)	1.05 (0.94 to 1.17)	0.82 (0.44 to 1.56)
ASA classification (3 vs 1)	1.42 (1.26 to 1.59)	2.59 (1.38 to 4.85)
ASA classification (4 vs 1)	2.28 (2.02 to 2.57)	8.13 (4.33 to 15.27)
ASA classification (5 vs 1)	5.35 (4.39 to 6.51)	27.36 (14.43 to 51.86)
Operative time	1.00 (1.00 to 1.00)	1.00 (1.00 to 1.00)
Total hospital length of stay	1.10 (1.09 to 1.10)	0.96 (0.96 to 0.96)
Case acuity (elective vs neither)	1.03 (1.00 to 1.07)	0.47 (0.42 to 0.52)
Case acuity (emergent vs neither)	1.49 (1.43 to 1.55)	1.82 (1.69 to 1.96)
Case approach (laparoscopic vs open)	0.58 (0.57 to 0.60)	0.46 (0.42 to 0.51)

ASA, American Society of Anesthesiologists; BMI, body mass index; COPD, chronic obstructive pulmonary disease.

Discussion

The proportion of patients undergoing surgery for diverticulitis who identified as Hispanic increased by approximately 3% from 2008 to 2017, while the proportion who identified as non-Hispanic decreased by 7%. There were no clinically significant differences in management or outcomes. Ethnicity does not appear to be a prognostic variable for morbidity and mortality following surgical management of diverticulitis based on logistic models, which is consistent with findings from other studies.⁸ Of note, Hispanic patients were younger at time of surgical intervention compared with non-Hispanic patients, suggesting that Hispanic patients may require a quicker transition to surgical management than their non-Hispanic counterparts. This is also supported by the fact that fewer cases were elective for Hispanic patients compared with non-Hispanic patients. In addition, Hispanic patients in our study had an increased BMI compared with non-Hispanic patients, which may explain their younger age at time of surgical intervention. Previous studies have suggested a correlation between increased BMI and low dietary fibre intake,⁹ a known risk factor for diverticulitis, which may be part of dietary changes in a younger Hispanic population.

Another possible explanation for why Hispanic patients may undergo surgical intervention at a younger age compared with non-Hispanic patients may stem from racial disparity in healthcare utilisation. While our study did not capture socioeconomic status of the patient population, research on other gastrointestinal diseases such as irritable bowel syndrome has demonstrated that patients who identify with a racial/ethnic minority are more likely to receive procedural intervention than those who identify as white.¹⁰ As evident in this study, minorities with diverticulitis may not be referred to gastrointestinal specialists early in their disease course and ultimately receive consultation from a surgical service who offers surgical resection. Others have pointed to implicit racial bias and its effect on provider–patient communication as an explanation for healthcare utilisation disparity.^{11 12}

While the inherent limitations to large database studies include information bias and sampling bias still exist, the large numbers within NSQIP allow for identification of trends that are more difficult to ascertain in single institutions. In addition, NSQIP only captures data up to postoperative day 30, and thus any morbidities or mortalities including after this time frame would not be captured. Furthermore, the dataset used does not allow for more granular analyses regarding race and ethnicity to be conducted. Future studies should focus on identifying why Hispanic patients require surgical management at a younger age and non-elective setting compared with non-Hispanic patients.

In conclusion, in a growing population of Hispanic patients with diverticulitis, we did not identify an increase in emergency surgery, perioperative morbidity or mortality, suggesting no difference in care or access to care. However, Hispanic patients tend to have surgery for diverticulitis at a younger age compared with non-Hispanic patients.

Data availability statement

Data may be obtained from a third party and are not publicly available. All of our data were obtained from the National Surgical Quality Improvement Program database. It is only available for participating hospitals.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

This study was reviewed and approved by our institutional review board.