Disparities in postoperative outcomes among diverse patient groups with inflammatory bowel disease

Introduction

Inflammatory bowel disease (IBD), comprising Crohn’s disease and ulcerative colitis, represents a significant burden of disease in the United States, with prevalences of 505 per 100,000 and 322 per 100,000, respectively, and is associated with a high healthcare cost and utilization [1]. In the past, the incidence of IBD was thought to have been highest in white and Jewish racial and ethnic groups [2,3]. Overall, the incidence of IBD is increasing [4]. This increase in diagnoses is changing the epidemiology of IBD, with a higher prevalence in non-White populations [5]. Studies estimate that up to one-third of patients with IBD are Black [6]. This shift in epidemiology presents an opportunity to determine whether there are disparities among those suffering from IBD.

There are a number of treatment options for those suffering from IBD. The majority of IBD patients are treated medically [7]. A deeper understanding of the pathophysiology of IBD has broadened the options for medical treatments in IBD patients [8]. Despite an increasing number of medical therapies for IBD, a large portion of IBD patients require surgery [9]. Disparities in surgical outcomes have been seen in the literature, with Hispanic and Black patients suffering more complications than White patients [10,11]. However, despite this increasingly diverse set of patients with IBD, the literature regarding their surgical outcomes remains limited [12-15].

A better understanding of the disparities in post-surgical outcomes allows targeting of areas for improvement in patient care. This can lead to changes in preoperative management and optimization, potentially decreasing postoperative complications. This study aimed to evaluate disparities in postoperative outcomes among a diverse set of patients undergoing surgery for IBD.

Materials and methods

Results

A total of 107,375 patients (weighted) underwent surgery for IBD. The majority were White (81.7%), rather than Black (10.1%) or Hispanic (8.2%); 56,630 (52.7%) of the patients were female. There was roughly an equivalent number of males and females with IBD among each race. Black and Hispanic patients were younger than White patients (44.1 vs. 44.6 vs. 50.5 years, P<0.001). Forty-seven percent of patients were diagnosed with Crohn’s disease, 24% had ulcerative colitis, while the remainder had unspecified IBD. Ulcerative colitis was the least frequent diagnosis for Black patients (14.7%) compared to the other races. A full description of the patients and demographics within the sample can be found in Table 1.

Table 1 Demographics

Most patients, 64275 (59.9%), underwent their operations at large hospitals. Nearly 80% of patients had their surgeries at urban teaching facilities. This trend was seen in all races. Close to half of all patients had private insurance, though private insurance was slightly less common in Black patients (37.9%).

In a pairwise comparison, Black and Hispanic patients had significantly longer hospital stays compared to White patients, with Black patients having the longest overall (12.6 vs. 11.2 vs. 9.6 days, P<0.001). Black patients had higher rates of postoperative infections than White patients, while Hispanic patients had the lowest rates (4.2% vs. 3.2% vs. 2.7%, P=0.0137). Black patients had significantly higher morbidity alone, as well as morbidity and mortality combined, compared to White and Hispanic patients (19.1% and 20.1% vs. 16.2% and 17.1% vs. 16.9% and 17.9%, P=0.0031 and P=0.0029) (Table 2).

Table 2 Outcomes by race

On univariate analysis, Black patients, females, patients with Medicare, patients being treated in medium sized hospitals, patients in rural hospitals, patients in the South, those with indeterminate inflammatory disease/other and those undergoing surgery on the left colon had the highest complication or death rates. The comparisons among variables (race, sex, primary payer, hospital size, location/teaching status, hospital region, calendar year, diagnosis, procedure type) in all categories were statistically significant. The univariate analysis was used to determine the variables for the multivariate analysis.

Using a multivariate logistic regression model controlling for covariates (Tables 3,4), Black patients were more likely to suffer from postoperative morbidity or mortality than White patients (odds ratio [OR] 0.762, 95%confidence interval [CI] 0.675-0.861). Those with Medicare and Medicaid had higher odds of morbidity and mortality than those with private insurance (OR 1.584, 95%CI 1.426 and OR 1.594, 95%CI 1.418-1.793). Patients undergoing surgery in rural or urban nonteaching hospitals had higher odds of postoperative morbidity and mortality than those at urban teaching facilities (OR 1.377, 95%CI 1.163-1.630 and OR 1.346 95%,CI 1.224-1.480). Age was not a significant predictor of complications or death. Black patients had significantly longer hospital stays on univariate analysis. A separate multivariate analysis was performed, controlling for covariates, which demonstrated that Black patients’ hospital stays were long regardless of other factors.

Table 3 Logistic regression for complication/death

Table 4 Linear regression on length of hospital stay

Discussion

Using a large and diverse database, our study has demonstrated significantly different rates of post-surgical complications across various races. Overall, rates of complications were slightly higher for Black patients compared to White or Hispanic patients. Similarly to previously reported data, Black patients had significantly higher rates of postoperative infections, morbidity and mortality, the cause of which remains unknown [13,15,18,19]. Our study adds to the literature, as it specifically addresses patients with IBD as the diversity in these patients continues to expand.

Notably, along with postoperative infections, Black patients experienced significantly longer hospital stays than White or Hispanic patients, by almost 3 days. This has further implications of increased costs to the patients and hospitals. A number of studies in the literature similarly describe an extended length of stay for Black patients after bariatric surgical procedures [5,14,20,21]. Dotson et al reported that Black children with Crohn’s disease had longer hospital stays (6.8 vs. 6.3 days) compared to White children. Similarly, Gunnells et al reported that Black adults with IBD had a longer index length of stay as well as a higher readmission rate than patients of other races [22,23]. Our study contributes to the growing evidence that Black patients, and specifically those with IBD, have longer hospital stays, a finding that warrants further investigation. In our study, this significant difference in length of stay was despite Black patients with IBD having fewer comorbidities. A number of factors other than clinical status can lead to a longer hospital stay. It may be difficult to discharge patients because they do not have social support at home, they struggle with transportation or live far from the hospital. Health literacy can impact patients’ ability to manage wounds or ostomy care. These factors are hard to capture within a database, but are an area that can be explored preoperatively with multidisciplinary allies to shorten the length of hospital stay and decrease costs.

As reported above, White patients in our study had more medical comorbidities than Black patients. Despite this, Black patients still suffered from more postoperative complications than White patients. This finding is not supported in the literature, where in general, a higher MCI was associated with worse postoperative outcomes in surgical oncology patients and orthopedic patients [24,25]. Focusing on Black patients, a study of patients undergoing knee and hip arthroplasty demonstrated similar results, with Black patients suffering from more postoperative complications despite having lower MCI scores [26]. It is unclear why the MCI failed to predict outcomes in this study, since it has often been used to accurately predict outcomes in the literature [17]. This suggests that the MCI may not be valid for diverse patient populations. There may be other important factors, in addition to patient comorbidities, that contribute to the length hospital of stay, such as social determinants of health and access to specialized care, such as gastroenterology and colorectal surgeons. High social vulnerability has been shown to increase morbidity following colorectal surgery [27]. These factors are hard to uncover in large database studies that have limited access to personal health information. This information is beyond the scope of this paper, but can guide future studies focusing on patients limited to one hospital or region to identify patients with high social vulnerability preoperatively.

The nature of the database limited our study to patients’ information from the initial hospital stay. This does not capture their entire post-surgical course and could leave out a number of post-surgical complications and hospital readmissions after the initial discharge. Complication rates, though statistically significantly different, were similar in percentage values. The clinical significance of these small differences is unclear, but is likely to be relevant given the large number of patients included in the database. Capturing the entire postoperative course for patients through 90 days or even a year could widen these percentages, as has been shown in previous publications [10]. This database also does not include such social determinants of health as social support at home, food security, transportation limits and patient’s distance from the hospital. This limits the interpretation of the results; however, our study adds to the growing body of literature on disparities in IBD and highlights the need to focus on other social determinants of health for this vulnerable population. In addition, this study was limited by a lack of specific IBD treatment information, such as steroids or type of biologics being used preoperatively. This limits the ability to determine disease severity or degree of immunosuppression, both of which could contribute to outcomes.

In conclusion, racially diverse patients with IBD suffer from higher rates of mortality and complications such as postoperative infection rate, and longer hospital stays after colorectal surgery compared to White patients. Further research is needed to determine whether the disparities still exist or widen after a patient’s initial hospital discharge. This information could potentially change preoperative management to improve patient outcomes and decrease the length of hospital stay.