Inflammatory bowel disease (IBD) represents a spectrum of diseases including ulcerative colitis (UC) and Crohn’s disease (CD). Both UC and CD indicate a chronic proinflammatory state of the gastrointestinal tract of largely unknown etiology with a relapsing and remitting course. There are approximately 1.6 million individuals diagnosed with IBD in the United States, with an increasing incidence of nearly 70,000 new diagnoses each year.1,2 Despite improvements in medical therapies to initiate and maintain long term disease control, only about half of these patients remain in clinical remission, which includes rare mild flares that do not require hospitalization.3–5 However, the other half do not maintain sustained disease control and experience severe flare-ups of chronic unremitting inflammation, which oftentimes lead to recurrent hospitalizations, difficult-to-control disease progression, and poor outcomes.3–6
Although available treatments for IBD have significantly improved since the advent of biologic and biosimilar therapies, IBD hospitalization rates have continued to rise, leading to significant health care costs and high health care utilization among these patients.7–10 Numerous studies show that the increase in hospitalization rates are one of the leading factors contributing to the medical costs of IBD patients.1,2,11,12 IBD currently ranks as one of the most costly gastrointestinal (GI) disorders despite it being one of the lowest in prevalence among all GI-specific disorders.13 The Crohn’s & Colitis Foundation of America extrapolated that the total annual financial burden of IBD in the United States to be $14.6 to $31.6 billion in 2014.12,14 A retrospective analysis evaluating factors contributing to the high financial burden among IBD patients performed by Limsrivilai et al15 demonstrated that higher hospitalization rates are strongly predictive of higher health care utilization and costs.
Despite significant advances in the medical management of IBD, the continued increase in IBD hospitalization rates is likely a product of disease progression because of inadequate maintenance of disease remission, associated complications, access to these medications, and comorbidities that may be present.16,17 Hospitalization outcomes, including charges, length of stay, and overall mortality can be affected by these complications and significant comorbidities. This is particularly important in IBD patients hospitalized for disease flares, as these flares predispose to comorbidities, such as increased risks of venous thromboembolism, bowel obstruction, and concurrent infections (eg, Clostridium difficile coinfection), all of which not only increase in-hospital mortality but contributes to higher health care resource utilization.2,17–20 Recurrent hospitalizations among both UC and CD patients are associated with significant morbidity and mortality and contributes to significant economic burden.3,4 These substantial financial burdens are largely dependent on the need for surgical intervention, expensive biologic therapies, frequent hospitalizations, and increased hospital length of stay.19–22
Access to disease-modifying agents for the treatment of IBD has been shown to vary by race/ethnicity, contributing to poorer outcomes and more severe disease in racial/ethnic minorities.23 For this reason, race/ethnic minorities may be expected to have more severe disease at presentation, potential delays in timely initiation of treatment, which may contribute to higher rates of hospitalization and health care resource utilization. The current study aims to specifically evaluate race/ethnicity-specific differences in hospitalization rates, in-hospital mortality, and hospitalization charges in the United States. We hypothesize that racial/ethnic minorities will experience higher rates of in-hospital mortality and will experience higher hospitalization charges.
We performed a retrospective analysis of data from the 2007 to 2013 Nationwide Inpatient Sample (NIS), a product of the Agency for Health care Research and Quality’s (AHRQ) Health care Cost and Utilization Project. NIS is the largest publicly available database of all-payer hospital inpatient stays in the United States; approximating a 20% stratified sample of community hospitals across the nation, NIS includes data from about 8 million hospital stays annually. Each unique hospital discharge record includes demographic and clinical data, as well as data on health care resource utilization, including expected primary payer, total hospital length of stay, and total hospital charges (not including professional charges). Lastly, the data set includes information about each participating hospital, including type of hospital (teaching vs. nonteaching), size, and location. NIS is not specifically designed to evaluate hospital outcomes in patients with IBD; thus, data specifically related to grading disease severity and treatment among IBD patients are lacking. However, NIS does include data on AHRQ comorbidity measures as well as methods to adjust for severity of illness and mortality risk overall. These measures of comorbidities and illness severity allow us to compare differences in health care resource utilization after correcting for comorbidity scores.
Adults (aged 18 y or older) with a primary diagnosis of IBD were identified using discharge diagnosis codes using the International Classification of Diseases, ninth revision, Clinical Modification (ICD-9-CM); the primary diagnoses included were Crohn’s disease (555.0, 555.1, 555.2, 555.9, 566, 569.81, 560.89) and ulcerative colitis (556.0, 556.1, 562, 556.3, 556.4, 556.5, 556.6, 556.8, 556.9). In addition, the presence of in-hospital comorbidities were identified using specific ICD-9 codes, including concurrent short-bowel syndrome, primary sclerosing cholangitis (PSC), HIV, and Clostridium difficile coinfection (all categorized as either present or absent). In addition, we included All Patient-Refined Diagnosis Related Group (APR-DRG) illness severity (categorized as minor loss of function, moderate loss of function, major loss of function, and extreme loss of function) and risk of mortality (categorized as minor likelihood of dying, moderate likelihood of dying, major likelihood of dying, and extreme likelihood of dying). Self-reported race/ethnicity included non-Hispanic white, African American, Hispanic, Asian, and Native American. Education and income level, although known to be correlated with race/ethnicity, were not included in the data set. Primary payer-specific evaluations included private and commercial insurance, Medicare, Medicaid, self-pay, and other types of insurance.
The analysis presented here compares in-hospital mortality, total hospital length of stay, and inflation-adjusted total hospitalization charges for CD and UC patients. We present the proportions (%) and frequencies (N) for categorical variables, and mean and SD, or median and interquartile range (IQR) for continuous variables. Hospitalization rates for each subgroup (eg, by race/ethnicity) were presented as hospitalizations per 1000 persons. Comparison of hospitalization rates within groups (eg, by race/ethnicity) utilized incidence density ratios (IDR), which were calculated by dividing the hospitalization rate of the comparator group with the hospitalization rate of the reference group. Univariate comparisons of crude in-hospital mortality between groups utilized χ2 testing, and comparisons of total hospital length of stay and total inflation-adjusted hospital charges were compared using Student t test or analysis of variance. Adjusted multivariate logistic regression models evaluated for independent predictors of in-hospital mortality among CD and UC hospitalizations, and multivariate linear regression models evaluated for independent predictors of total hospital length of stay and total inflation adjusted hospital charges. Given the large sample size of our cohort and the concern for over-fitting of the multivariate models, variables selected for inclusion in the final models were selected a priori based on what we hypothesized, based on clinical experience and knowledge of the literature, to be clinically significant in affecting the above outcomes among patients hospitalized for IBD. All statistical analyses were performed using STATA statistical software package (version 14.0, StataCorp LP, College Station, TX), and statistical significance was met with 2-tailed P-value<0.05. This study was granted exempt status from Alameda Health System Institutional Review Board.
Between 2007 and 2013, there were 224,500 hospitalizations with a primary diagnosis of IBD. Of these, 174,629 (77.8%) were for CD and 49,871 (22.2%) were for UC (Table 1).
Hospitalization Rates and Characteristics of the Study Cohort
Descriptive comparisons of the IBD cohort are stratified by CD versus UC (Table 1). The gender distribution among IBD hospitalizations were similar, with women representing 50.9% of CD and 52.7% of UC related hospitalizations. However, among both CD and UC, men had significantly higher hospitalization rates than women (CD IDR, 1.33; 95% CI, 1.32-1.33; P<0.001; UC IDR, 1.24; 95% CI, 1.23-1.24; P<0.001) (Table 2).
Non-Hispanic whites represented the largest race/ethnic group for both CD and UC hospitalizations (61.4% and 62.9%, respectively) and African Americans were the second largest race/ethnic group (Table 1). Compared with non-Hispanic whites with CD, significantly lower hospitalization rates were observed for Asians (IDR, 0.59; 95% CI, 0.57-0.61; P<0.001), Hispanics (IDR, 0.65; 95% CI, 0.63-0.65; P<0.001), and African Americans (IDR: 0.95, 95% CI, 0.94-0.95; P<0.001) (Table 2). Similar trends were seen when evaluating hospitalizations among UC patients (compared with non-Hispanic whites, Asians: IDR 0.60; 95% CI, 0.56-0.63; P<0.001; Hispanics: IDR, 0.78; 95% CI, 0.76-0.79; P<0.001; and African Americans: IDR, 0.70; 95% CI, 0.68-0.71; P<0.001) (Table 2).
Although the most common insurance type among both CD and UC hospitalizations was private/commercial insurance, a significantly greater proportion of patients who were hospitalized with CD had Medicare insurance compared with patients being hospitalized for UC (29.8% for CD, vs. 23.1% for UC; P<0.001). For both CD and UC, patients with Medicare and Medicaid had significantly lower hospitalization rates compared with patients with commercial insurance. UC patients with Medicare and Medicaid had about a 60% lower rate of hospitalization compared with UC patients with private/commercial insurance (Medicare IDR, 0.38; 95% CI, 0.39-0.39; P<0.05 and Medicaid IDR, 0.37; 95% CI, 0.37-0.38; P<0.001). Similarly, CD patients with Medicare and Medicaid had a 40% and 46% lower rates of hospitalization, respectively, compared with CD patients with private/commercial insurance (Medicare IDR, 0.60; 95% CI, 0.59-0.60; P<0.001 and Medicaid IDR, 0.54; 95% CI, 0.53-0.54; P<0.05) (Table 2).
From 2007 to 2013, overall number of CD hospitalizations decreased by 1.6% whereas the overall number of UC hospitalizations increased by 8.4% (Table 2). When stratified by age, there was a 6.7% increase in CD hospitalizations among patients age below 40 years, whereas CD hospitalizations decreased by 4.0% in those aged 40 to 59 years and decreased by 7.3% in those age 60 years and over. In contrast, there was an increase in hospitalizations among all age groups among UC patients, with the greatest increase seen in patients aged below 20 years (age below 20 y 14.7% increase; 20-39 y 13.5% increase; 40-59 y 7.4% increase; and above 60 y 0.8% increase) (Table 2). In 2013, the greatest hospitalization burden was seen among CD and UC patients age below 40 years, which represented 38.2% of all CD hospitalizations and 45.6% of all UC hospitalizations (Table 2). Although non-Hispanic whites represented the large majority of IBD hospitalizations between 2007 and 2013, the greatest increase observed in CD hospitalizations was seen among African Americans (40.7% increase; P<0.05), whereas the greatest increase in UC hospitalizations was seen among Hispanics (51.8% increase; P<0.05) during that same period. When stratified by insurance type, CD hospitalizations among patients with private/commercial insurance decreased by 11.6% (P<0.05) from 2007 to 2013, whereas CD hospitalizations among those with Medicaid increased by 23.6% (P<0.05). In the same time-period, UC hospitalizations among Medicaid patients increased by 41.2% (P<0.05), whereas UC hospitalizations in commercially insured patients increased by 4.2%, (P<0.01) (Table 2).
Hospitalization Charges and Hospital Length of Stay
Among all CD hospitalizations, the mean hospital LOS decreased from 5.8 days (IQR, 2-7) in 2007 to 5.4 days (IQR, 2-6) in 2013, whereas mean inflation-adjusted hospitalization charges increased from $29,632 (IQR, 9787-30,727) in 2007 to $41,484 (IQR, 14,194-44,478) in 2013 (Table 3). Similarly, for UC hospitalizations the mean hospital LOS decreased from 6.3 days (IQR, 3-7) in 2007 to 5.7 days (IQR, 3-7) in 2013, whereas mean inflation adjusted hospitalization charges increased from $31,449 (IQR, 10,965-35,715) in 2007 to $43,128 (IQR, 15,767-50,025) in 2013.
CD hospitalizations among patients aged 60 years and older had significantly longer mean hospital LOS (6.98 vs. 5.43 d; P<0.01) and significantly higher mean inflation adjusted hospitalization charges ($43,839 vs. $34,020; P<0.01) compared with those aged below 20 years (Table 3). For UC hospitalizations, patients aged less than 20 years had the longest mean hospital LOS (6.83 d; IQR, 3-8) and the highest mean inflation adjusted hospitalization charges ($41,902; IQR, $10,965-$48,424) among all age groups.
Hispanic and Asian patients hospitalized for CD had the shortest mean hospital LOS (5.24 d and 5.27 d, respectively), but the highest mean inflation adjusted hospitalization charges ($40,849 for Hispanics and $41,848 for Asians) compared with African Americans and non-Hispanic whites with CD. However, African American and Hispanic UC patients had the shortest mean hospital LOS (5.72 d and 5.73 d, respectively). Hispanic and Asian patients hospitalized for UC also had the highest mean inflation adjusted hospitalization charges ($43,901 and $43,801, respectively), whereas African American UC patients had the lowest mean hospitalization charges ($34,343). CD and UC patients with Medicare had disproportionately longer mean hospital LOS (6.9 d for CD and 6.75 d for UC) and mean hospitalization charges ($42,661 for CD and $42,450 for UC) compared with patients with Medicaid or private/commercial insurance (Table 3).
On multivariate linear regression, compared with CD patients age below 20 years, significantly shorter mean hospital LOS was observed among those aged 20 to 39 years (0.48 d shorter; 95% CI, –0.63 to –0.33; P<0.001) and among those 60 years and older (0.46 d shorter; 95% CI, –0.66 to –0.25; P<0.001) (Fig. 1). Compared with CD patients age below 20 years, significantly lower mean inflation adjusted hospitalization charges was observed among those aged 20 to 39 ($2669 lower; 95% CI, –$3964 to –$1434; P<0.001) and among those 60 years and older ($4745 lower; 95% CI, –$6569 to –$2920; P<0.001).
Compared with UC patients aged below 20 years, significantly shorter mean hospital LOS was observed among UC patients aged 20 to 39 (1.21 d shorter; 95% CI, –1.44 to –0.97; P<0.001), aged 40 to 59 years (1.11 d shorter; 95% CI, –1.36 to –0.85; P<0.001), and among those 60 years and older (1.48 shorter; 95% CI, –1.80 to –1.16; P<0.001) (Fig. 1).
When compared with non-Hispanic whites, Hispanic patients hospitalized for CD had a significantly longer mean hospital LOS (0.18 d longer; 95% CI, 0.04-0.31; P<0.05) and significantly higher mean hospitalization charges ($9302 higher; 95% CI, $7910-$10,694; P<0.001), representing the group with the highest hospitalization charges among all racial/ethnic groups. Although Asian patients did not have a statistically significant difference in mean hospital LOS, this group did have significantly higher mean hospitalization charges when compared with non-Hispanic whites ($7665 higher; 95% CI, $4859-$10,451; P<0.001). Among UC patients, there was no significant difference in mean hospital LOS between race/ethnic groups when compared with non-Hispanic whites. However, African American UC patients had $3551 lower charges (95% CI, –$5002 to –$2101; P<0.001), and Hispanic patients had $6910 higher charges (95% CI, $4623-$9197; P<0.001).
Patients hospitalized for CD with Medicare and no insurance had shorter mean hospital LOS (0.20 d shorter, P<0.002; 0.10 d shorter, P<0.05, respectively) and lower mean hospitalization charges compared with CD patients with private/commercial insurance (Medicare: $4784 lower; 95% CI, –$5970 to –$3598; P<0.001) (Fig. 1). Among UC patients, those with Medicare had significantly shorter mean hospital LOS (0.51 d shorter; 95% CI, –0.73 to –0.29; P<0.001) and had significantly lower hospitalization charges ($6773 lower; 95% CI, –$9049 to –$4497; P<0.001), compared with UC patients with private/commercial insurance. UC patients without insurance represented the lowest hospital LOS (0.58 d shorter; 95% CI, –0.76 to –0.39; P<0.001) and had the lowest hospitalization charges ($7740 lower; 95% CI, –$9752 to –$1757; P<0.001) compared with patients with any type of insurance.
The presence of comorbidities also had a significant impact on mean hospital LOS and mean hospitalization charges. Patients with concurrent CDCI or short bowel syndrome had significantly longer mean hospital LOS and significantly higher hospitalization charges compared with patients without these comorbidities (Fig. 1).
Overall, in-hospital mortality was low in both CD and UC hospitalizations (0.99% and 0.78%, respectively, Table 1). Among both CD and UC, patients aged 60 years and older had significantly higher odds of in-hospital mortality compared with patients below 20 years old (CD OR, 7.69; 95% CI, 2.39-24.74; P<0.001 and UC OR, 4.40; 95% CI, 1.25-15.41; P<0.05; Table 4). No significant differences in odds of in-hospital mortality were observed by sex. However, Hispanic UC patients had a trend towards higher odds of in-hospital mortality compared with non-Hispanic whites (OR, 1.54; 95% CI, 0.95-2.51; P=0.08). Hospitalized UC patients with CDCI had significantly higher mortality (OR, 1.65; 95% CI, 1.08-2.53; P<0.05) (Fig. 1).
IBD represents a chronic, costly, and increasingly morbid disease affecting all aspects of the United States health care delivery and utilization.24 Using a nationwide database of 224,500 IBD hospitalizations between 2007 and 2013, we evaluated race/ethnic disparities of IBD-related hospitalizations, with a focus on hospital length of stay, in-hospital mortality, and hospitalization charges.
Among both CD and UC patients, those of Asian, Hispanic and African American ethnicity had lower hospitalization rates when compared with their Non-Hispanic white counterparts. However, Hispanic CD patients had significantly longer mean hospital LOS as compared with non-Hispanic whites and significantly higher mean hospitalization charges. African American and Hispanic UC patients had significantly higher mean hospitalization charges compared with non-Hispanic whites. Interestingly, Hispanic UC patients demonstrated a trend towards higher in-hospital mortality when compared with non-Hispanic whites.
From 2007 to 2013, CD-related hospitalizations appeared to be declining, whereas UC-related hospitalizations were observed to be increasing. Similar findings were observed in a study by Gajendran et al,25 who observed a decrease in hospitalization of IBD patients from the emergency department from 2009 to 2011. Despite the decreasing hospitalization rates, the study reported significantly higher IBD-related hospitalizations in older patients.25 These findings differ from our current study observations where the greatest burden of hospitalizations in 2013 was seen in younger patients aged below 40 years. In addition, numerous international and national studies have shown an increased rate of inpatient IBD hospitalizations in the last decade.26,27 The rise in hospitalizations found in other countries could reflect an increase in prevalence of IBD and regional differences in the natural history of the disease and prevalence of usage of anti-inflammatory therapies. The decrease in CD hospitalizations observed in our study may be because of improved outpatient management of IBD patients, which reduces the need for inpatient hospitalization.
Non-Hispanic whites represented the major proportion of UC and CD hospitalizations in our analysis. This is supported by a 2018 systematic review and meta-analysis by Shi et al,23 who demonstrated that most patients were of Caucasian ethnicity. This can be explained by the higher incidence of family history and greater propensity of IBD in non-Hispanic whites compared with other ethnicities.4 However, these studies are largely based in the United States, and, although IBD has previously been thought to be a disease primarily of Caucasians, epidemiologic data is showing that it is becoming increasingly more common among non-White populations outside the United States.28 Our current study observed that Asian patients with CD had significantly higher mean hospitalization charges compared with non-Hispanic whites. This may be explained by phenotypic differences seen across different ethnicities, with Asian CD patients having a higher frequency of perianal involvement compared with non-Hispanic Whites, with such complications increasing the risk of hospitalization.23 Asian CD patients also had lower rates of surgeries compared with other ethnicities which could explain why they did not have increased mean hospital LOS.23 Previous studies evaluating ethnic disparities in IBD have observed lower utilization of primary care, lower utilization of gastrointestinal specialty care, higher risk of leaving against medical advice, and lower adherence to medication therapy either by lack of health literacy or cost, or both particularly among ethnic minority populations.29,30
In-hospital mortality was significantly increased in older patients, those of Hispanic ethnicity, and those with CDCI. Similar increases in mortality in IBD patients was observed in a 2013 systematic review looking at all-cause and cause-specific mortality in IBD patients as compared with the general population.23 Bitton et al28 found significantly higher all-cause mortality in UC and CD patients as well when compared with the general population. Bernstein et al31,32 evaluated risk factors for death in CD and UC patients and observed that the increased mortality in CD and UC patients were because of increased comorbidity burden. Importantly, lower socioeconomic status was also highly associated with mortality in CD and UC patients.31 Although our study observed certain race/ethnicity-specific differences in IBD outcomes, our observational study design only allows us to make associations without a definitive path for causation. Nevertheless, the findings do suggest the need for greater awareness and improved management of IBD among ethnic minorities, and quality improvement programs targeting patients and providers may help improve IBD management particular among these groups.
Although the use of a large national database of inpatient hospitalizations allows a more generalizable approach to evaluating IBD-related outcomes, certain limitations of this database should be acknowledged. The NIS data set only includes data from inpatient hospitalizations and thus disease characteristics, including mortality outcomes that occurred in the outpatient setting were not accurately captured. Although we attempted to use established ICD-9 coding based criteria for identification of our cohort and variables of interest, errors in coding or potential misclassification bias could have occurred during the development of this data set. Furthermore, the NIS data set was not specifically designed to evaluated IBD-related outcomes, and thus data necessary for evaluating IBD-related severity (eg, IBD-related symptoms, endoscopic findings) were not available for our analysis. In addition, treatment data was not available. However, we attempted to use the available APR-DRG illness severity and risk of mortality data to include in our multivariate models. Furthermore, one of our main study aims was to specifically evaluate race/ethnicity-specific differences in hospitalization outcomes among individuals with IBD. Although we did in fact observed race/ethnicity-specific disparities, it is important to acknowledge that race/ethnicity itself is often correlated with other socioeconomic factors such as household income of education level, which are equally important factors to assess when evaluating health outcomes. However accurate assessment of household income and education level was not available in the NIS data set, and thus the race/ethnicity-specific differences that we observed should be interpreted with this in mind. In addition, we also acknowledge that whereas it is generally accepted that there is some gene-related factor as it relates to IBD pathogenesis and natural history, it is also not quite clear what specific genetic testing is recommended in the evaluation of IBD patients and what specific role these tests have in disease prediction of prognostication. The NIS data set did not provide genetic data to be included in our analysis this limiting our ability to analyze this aspect. Despite these limitations, we believe that our findings provide important updates to understanding the epidemiology and clinical outcomes among adults with IBD in the United States.
In conclusion, among a large United States database evaluating IBD-related hospitalizations from 2007 to 2013, we observed increasing hospitalization rates for adults with UC. Significant race/ethnicity-specific disparities in IBD hospitalization outcomes were observed. Hispanic patients with CD had significantly longer hospital LOS and highest hospitalization charges among all groups after correcting for illness severity. In addition, Hispanic and African American UC patients both had significantly higher hospitalization charges compared with non-Hispanic whites. Overall, Hispanic UC patients had a trend towards higher in-hospital mortality compared with non-Hispanic whites. Although these differences in hospitalization outcomes may reflect disparities in timely and appropriate management of IBD, data are not available to address this hypothesis in this data set and more research is needed to more fully explore this hypothesis.
1. Kappelman MD, Moore KR, Allen JK, et al. Recent trends in the prevalence of Crohn’s disease
and ulcerative colitis
in a commercially insured US population. Dig Dis Sci. 2013;58:519–525.
2. Kappelman MD, Porter CQ, Galanko JA, et al. Utilization of healthcare resources by U.S. children and adults with inflammatory bowel disease. Inflamm Bowel Dis. 2011;17:62–68.
3. Monstad I, Hovde Ø, Solberg IC, et al. Clinical course and prognosis in ulcerative colitis
: results from population-based and observational studies. Ann Gastroenterol. 2014;27:95–104.
4. Solberg IC, Lygren I, Jahnsen J, et al. Clinical course during the first 10 years of ulcerative colitis
: results from a population-based inception cohort (IBSEN Study). Scand J Gastroenterol. 2009;44:431–440.
5. Solberg IC, Vatn MH, Hoie O, et al. Clinical course in Crohn’s disease
: results of a Norwegian population-based ten-year follow-up study. Clin Gastroenterol Hepatol. 2007;5:1430–1438.
6. Hovde O, Moum BA. Epidemiology and clinical course of Crohn’s disease
: results from observational studies. World J Gastroenterol. 2012;18:1723–1731.
7. Bewtra M, Su C, Lewis JD. Trends in hospitalization
rates for inflammatory bowel disease in the United States. Clin Gastroenterol Hepatol. 2007;5:597–601.
8. Bassi A, Dodd S, Williamson P, et al. Cost of illness of inflammatory bowel disease in the UK: a single centre retrospective study. Gut. 2004;53:1471–1478.
9. Feagan BG, Vreeland MG, Larson LR, et al. Annual cost of care for Crohn’s disease
: a payor perspective. Am J Gastroenterol. 2000;95:1955–1960.
10. Rawla P, Sunkara T, Raj JP. Role of biologics and biosimilars in inflammatory bowel disease: current trends and future perspectives. J Inflamm Res. 2018;11:215–226.
11. Kappelman MD, Rifas-Shiman SL, Porter CQ, et al. Direct health care costs of Crohn’s disease
and ulcerative colitis
in US children and adults. Gastroenterology. 2008;135:1907–1913.
12. Everhart JE. US Department of Health and Human Services PHS, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. The burden of digestive diseases in the United States. Volume NIH Publication No 09-6443. Washington, DC: US Government Printing Office; 2008:137–147.
13. Stone CD. The economic burden of inflammatory bowel disease: clear problem, unclear solution. Dig Dis Sci. 2012;57:3042–3044.
14. Gibson TB, Ng E, Ozminkowski RJ, et al. The direct and indirect cost burden of Crohn’s disease
and ulcerative colitis
. J Occup Environ Med. 2008;50:1261–1272.
15. Limsrivilai J, Stidham RW, Govani SM, et al. Factors that predict high health care utilization and costs for patients with inflammatory bowel diseases. Clin Gastroenterol Hepatol. 2017;15:385–392; e2.
16. Shaheen NJ, Hansen RA, Morgan DR, et al. The burden of gastrointestinal and liver diseases, 2006. Am J Gastroenterol. 2006;101:2128–2138.
17. Obi K, Hinton A, Sobotka L, et al. Hospital-acquired conditions are associated with worse outcomes in Crohn’s disease
-related hospitalizations. Digest Dis Sci. 2017;62:1621–1627.
18. Weizman AV, Nguyen GC. Quality of care delivered to hospitalized inflammatory bowel disease patients. World J Gastroenterol. 2013;19:6360–6366.
19. Ananthakrishnan AN, McGinley EL, Binion DG. Does it matter where you are hospitalized for inflammatory bowel disease? A nationwide analysis of hospital volume. Am J Gastroenterol. 2008;103:2789–2798.
20. Rezapour M, Galoosian A, Liu B, et al. Clostridium difficile co-infection in inflammatory bowel disease is associated with significantly increased in-hospital mortality
. Eur J Gastroenterol Hepatol. 2018;30:1041–1046.
21. Langholz E, Munkholm P, Davidsen M, et al. Course of ulcerative colitis
: analysis of changes in disease activity over years. Gastroenterology. 1994;107:3–11.
22. Hay JW, Hay AR. Inflammatory bowel disease: costs-of-illness. J Clin Gastroenterol. 1992;14:309–317.
23. Shi HY, Levy AN, Trivedi HD, et al. Ethnicity influences phenotype and outcomes in inflammatory bowel disease: a systematic review and meta-analysis of population-based studies. Clin Gastroenterol Hepatol. 2018;16:190–197; e11.
24. Sewell JL, Velayos FS. Systematic review: the role of race and socioeconomic factors on IBD healthcare delivery and effectiveness. Inflamm Bowel Dis. 2013;19:627–643.
25. Gajendran M, Umapathy C, Loganathan P, et al. Analysis of hospital-based emergency department visits for inflammatory bowel disease in the USA. Dig Dis Sci. 2016;61:389–399.
26. Bellolio Roth F, Gomez J, Cerda J. Increase in hospital discharges for inflammatory bowel diseases in chile between 2001 and 2012. Dig Dis Sci. 2017;62:2311–2317.
27. Jakubowski A, Zagorowicz E, Kraszewska E, et al. Rising hospitalization
rates for inflammatory bowel disease in Poland. Pol Arch Med Wewn. 2014;124:180–190.
28. Sewell JL, Yee HF Jr, Inadomi JM. Hospitalizations are increasing among minority patients with Crohn’s disease
and ulcerative colitis
. Inflamm Bowel Dis. 2010;16:204–207.
29. Nguyen GC, LaVeist TA, Harris ML, et al. Patient trust-in-physician and race are predictors of adherence to medical management in inflammatory bowel disease. Inflamm Bowel Dis. 2009;15:1233–1239.
30. Finlay DG, Basu D, Sellin JH. Effect of race and ethnicity on perceptions of inflammatory bowel disease. Inflamm Bowel Dis. 2006;12:503–507.
31. Bernstein CN, Nugent Z, Targownik LE, et al. Predictors and risks for death in a population-based study of persons with IBD in Manitoba. Gut. 2015;64:1403–1411.
32. Bernstein CN, Papineau N, Zajaczkowski J, et al. Direct hospital costs for patients with inflammatory bowel disease in a Canadian tertiary care university hospital. Am J Gastroenterol. 2000;95:677–683.