ORIGINAL ARTICLE

A comparison of abdominal surgical outcomes betweenAfrican–American and Caucasian Crohn’s patients

Nicole Griglione & Shadi Yarandi & Jahnavi Srinivasan &

Thomas Ahearn & Tanvi Dhere

Accepted: 2 May 2014# Springer-Verlag Berlin Heidelberg 2014

AbstractBackground and objective Whether race affects the naturalhistory of Crohn’s disease is a matter of debate. The aim of thecurrent studywas to evaluate the differences in surgical outcomesbetween African–American (AA) and Caucasian (C) Crohn’spatients undergoing surgery at a tertiary care referral center.Methods With Institutional Review Board approval, the medi-cal records of our institutionwere queried to identify consecutiveAA and C patients who underwent surgery for Crohn’s diseasefrom December 1, 2009 to December 15, 2011. A retrospectivechart review was performed using electronic medical records.Results A total of 77 patients were included in this study,including 32 AA (41 %) and 45 C (59 %). No significantdifferences were seen with respect to age, gender, type ofinsurance, preoperative exposure to immunosuppressives,body mass index, or smoking history between the two popu-lations (p>0.05). There was a trend toward lower albumin in

AAs (p=0.09). AA and C patients who underwent their firstCrohn’s disease (CD)-related surgery had similar lag periodsbetween diagnosis and surgery. No significant differenceswere seen in location of disease, indication for operation,and need for open laparotomy over laparoscopy. No signifi-cant differences were seen in need for a repeat operationwithin 90 days of the original surgery or major postoperativecomplications. There was a trend toward higher rate of minorcomplications in the AA group (p=0.07).Conclusion No significant differences were noted in the cur-rent study in several preoperative variables and surgical out-comes between AA and C.

Keywords African–Americans . Crohn’s disease .

Caucasians . Surgery

Introduction

Crohn’s disease (CD) is a form of inflammatory bowel disease(IBD) that can be wrought with morbidity due to its associa-tions with complications such as bowel stricturing and perfo-rations. The etiology and pathogenesis of the disease is multi-factorial involving a complex interaction of genetics, intestinalmicrobiota, environmental factors, and innate immunity. Ini-tially felt to be a disease affecting primarily the Caucasianpopulation, the current literature suggests that the incidencerates of hospitalizations of white and African–American IBDpatients may be comparable [1]. . Data on overall incidence ofCD in AA come from an epidemiologic study performed in anarea with a limited minority population. Newer populationdata on incidence rates of CD in AA is missing from currentliterature. However, most recent literature on prevalence basedon data from a large health maintenance organization studywas two thirds that of whites [1]. Some studies have suggestedthat CD might be more prevalent than ulcerative colitis (UC)

N. Griglione : S. Yarandi : J. Srinivasan : T. Ahearn : T. Dhere (*)Department of Medicine, Emory University, 1365 Clifton Road, NEBuilding B, STE 1200, Atlanta, GA 30322, USAe-mail: tdhere@emory.edu

N. Griglionee-mail: ngriglione@yahoo.com

S. Yarandie-mail: shadi.sadeghi.yarandi@emory.edu

J. Srinivasane-mail: jsrini2@emory.edu

T. Ahearne-mail: tahearn@hsph.harvard.edu

J. SrinivasanDepartment of Surgery, Emory University, 1365 Clifton Road NE,Atlanta, GA 30322, USA

T. AhearnSchool of Public Health, Emory University, 1365 Clifton Road NE,Atlanta, GA 30322, USA

Int J Colorectal DisDOI 10.1007/s00384-014-1902-2

in AAs [2]. Despite this, there is a paucity of literature exam-ining the differences and outcomes of the disease in thispopulation [3]. Whether the presentation of CD is differentin AAs compared to Caucasians is not clear. Nguyen et al.reported that AA patients with CD were more likely to havecolorectal and perianal disease and less likely to have ilealinvolvement than Caucasians [4]. However, higher rates ofileal involvement and lower rates of perianal disease havebeen reported by several other studies [5, 6]. A qualitativeanalysis of eight studies of IBD in AAs showed no differencein disease location between AAs and Caucasians [7]. In addi-tion to the findings noted above, Nyugen et al. reported thatAAs are less likely to have abdominal perforating disease andmore likely to have strictures [4]. This finding has also beenchallenged by other studies which have shown similar diseasebehavior in AAs and Caucasians [6, 8–11].

The reason for potential differences between AA and Cau-casian CD patients may be related to genetics. However,genome-wide association studies have found that disease-conferring alleles found in AA CD patients is due to Europeanadmixture [12]. Additional factors are likely playing a role inthe disease as several studies noted above have shown pheno-typic differences between the two populations with differencesin disease location and perianal disease. The factors mayinclude differences in microbiota and environmental factors;however, these are yet to be defined. Socioeconomicsmay alsoplay a role in differences in disease outcomes. In the pediatricliterature, underinsured children were found to have moresevere IBD compared to privately insured patients [13]. Dataexamined in adults is missing in the current literature. Despitecontrolling for socioeconomic status, research in other diseasestates indicates that health care is inferior for AAs compared toCaucasians [14, 15]. It is unclear if the same is true for IBD.

Data regarding the course and the outcome of CD in AAs islimited and controversial. The number of surgeries performedin AA patients with CD has been reported to be lower [11],equal [5, 6], and higher [16] than Caucasians by differentstudies depending on the setting and the population of thestudy. The rate of reoperation in patients has also varied inthe current literature [5, 6, 16]. The study of Deveaux et al.showed similar disease severity between black and CaucasianCD patients undergoing surgery [8]. AAs with CD have beenreported to be less likely to have regular visits with gastroen-terologists and to also less likely receive treatment with bio-logic agents or other immunosuppressive agents [17]. It alsohas been reported that medication non-adherence is more com-mon amongAAs [18, 19]. Further studies have attributed thesefindings to socioeconomic status rather than ethnicity [20].

Despite multiple studies investigating racial disparities inCD, the current literature is lacking of data assessing potentialdifferences in AA and Caucasian CD patients undergoingsurgery. Emory University Hospital is a tertiary care referralcenter for CD-related surgery in a racially diverse area. In the

current study, we compared several preoperative variables andpostoperative outcomes between the two populations.

Materials and methods

Design and setting

We performed a retrospective chart review of patients with CDwhowere seen at the Emory Clinic and who underwent surgeryfor CD at Emory University Hospital between 2009 and 2011.Emory University Hospital is a tertiary care referral center thatevaluates patients from theAtlanta, Georgia, area and surround-ing areas in the southeastern USA. This study was approved bythe Institutional Review Board (IRB) at Emory University.

Inclusion and exclusion criteria

Patients were identified from the electronic medical recordssystem using the following ICD9 codes: 555.0, regional enter-itis of small intestine; 555.1, regional enteritis of large intes-tine; 555.2, regional enteritis of small/large intestine; and555.9, regional enteritis of unspecified site. Confirmations ofpatients’ diagnosesweremade following review of themedicalrecords, including radiographic and histologic findings. Onlythose patients who underwent surgery for a Crohn’s-relatedcomplication were included. Patients who underwent minorsurgery including perianal fistulotomy and ostomy takedown/repair were excluded. Patients with ulcerative colitis and inde-terminate colitis were also excluded from the study.

Race was defined by how patients identified themselves intheir medical records. All AA and Caucasian (C) patients whomet the above inclusion criteria were included in the study.

Demographic and clinical variables

We assessed several demographic and clinical variables in AAand C Crohn’s patients undergoing surgery for a CD-relatedcomplication. These included age, gender, and tobacco use.Additional information gathered included history of previousCrohn’s-related surgeries, number of years fromCD diagnosisto first surgery, type of medical insurance, and exposure toimmunosuppressives prior to surgery. Nutritional status ofpatients was assessed by examining body mass index andalbumin levels. Additional information including type of sur-gery, need for repeat surgery within 90 days of the originalsurgery, and major and minor postoperative complicationswere collected.

Data were extracted from electronic medical records. Con-tinuous variable were compared using Student’s t test, andcategorical variables were compared using Fisher’s exact test.All tests were two-sided, and a p value of less than or equal to0.05 was considered statistically significant.

Int J Colorectal Dis

Results

Demographics

A total of 77 patients were included in the study including 32AA (41 %) and 45 C (59 %). Twenty patients in the AA group(62.5 %) and 24 patients in the C group (51 %) were male (p=0.49). The average age of patients at the time of surgery was38.1 in the AA group and 41.9 in the C group (p=0.20). Thesefindings were not statistically significant. Within the AAgroup, 12 (37.5 %) patients had Medicare or Medicaid, 18(56.2 %) had private insurance, and 2 (6.3 %) had no insur-ance. In the C group, 17 (37.8 %) patients had Medicare orMedicaid, 28 had private insurance (62.2 %), and none had noinsurance (0 %). There was no statistically significant differ-ence between the status of insurance between these groups (p=0.49) (Table 1).

Preoperative data

We examined the clinical features, medication profiles,and nutritional status of the patients before undergoingsurgery. A summary of the data is presented in Table 1.No statistically significant differences in disease loca-tions between AAs and Caucasians undergoing surgerywere seen. Nine patients in the AA group (28.1 %) and14 patients in the C group (29.85 %) had involvement ofsmall bowel only while 8 patients in the AA group(25 %) and 7 in the C group (14.9 %) had only colonicdisease. The remainder of patients had both small boweland colon involvement. No statistically significant differ-ences in preoperative immunosuppressive exposures in-cluding steroids, biologics, and immunomodulator thera-py within 4 weeks of the surgery were seen between thetwo groups. History of prior surgery for CD was docu-mented in 15 AA patients and 21 C patients, which wasnot statistically significant (p=1.00). The average num-ber of years from CD diagnosis to surgery in thosehaving their first surgery was not different between thetwo groups (8 vs 7 years, respectively, p=0.66).

No statistically significant differences in indication forsurgery and nutritional status were seen among the twogroups. Nine AA and seven C patients had more thanone indication for need for surgery. The average BMI ofAA patients at the time of admission for surgery was22.1 which was comparable to the BMI of Caucasianpatients who had an average BMI of 23.7 (p=0.21).Although there was a trend toward lower albumin levels,the results did not reach statistical significance (2.90 vs3.19 g/dL, p=0.09).

Intra-operative data

We reviewed data regarding type of surgery (i.e., laparoscopicvs open surgery). More AAs underwent open laparotomyrather than laparoscopic surgery compared to Caucasians(72 vs 55 %); however, the difference did not reach statis-tical significance (p=0.16) (Table 2). No deaths occurred as aresult of surgery.

Table 1 A comparison of preoperative characteristics of patients includ-ed in the study

Preoperativecharacteristics

African–Americans(N=32)

Caucasians(N=45)

p value

Sex

Male 12 24 0.49

Female 20 21

Age (years) 38.1 41.9 0.20

Smoking 6 9 1.00

Insurance

Uninsured 3 0 0.49

Medicaid/Medicare 12 15

Private 17 30

Prior Crohn’s surgery 15 21 1.00

Preoperative BMI 22.1 23.7 0.21

Preoperative serumalbumin (g/dL)

2.90 3.19 0.09

Immunosuppressiveexposure

Steroid 10 13 1.00

Anti-TNF 9 10 0.60

6MP/AZA/MTX 10 19 0.47

Location of disease

Small bowel 9 14 0.81

Colon 8 7 0.38

Both 15 24 0.65

Surgical indication

Perforation 6 4 0.30

Obstruction 16 28 0.35

Fistula 12 12 0.11

Refractory disease 7 8 0.77

Number of years betweenCD diagnosis and firstsurgery

8 7 0.66

Table 2 A comparison of operative and postoperative variables in pa-tients included in the study

Operative andpostoperative variables

African–Americans(N=32)

Caucasians(N=45)

p value

Open laparotomy 23 25 0.16

Need for repeat ORwithin 90 days

7 7 0.38

Major complications 13 12 0.22

Minor complications 12 8 0.07

Int J Colorectal Dis

Postoperative data

No significant difference was seen in the need for repeatsurgery within 90 days of the original surgery (21.9 % AAvs 17.8 % C, p=0.38) (Table 2). Seven patients in each grouprequired surgery within 90 days of the original operation.Indications for the need for repeat surgery includedduodenectomy with fistula takedown, incisional hernia repair,debridement of chronic abdominal wall wound/abscess, co-lostomy obstruction, ileostomy stenosis, and small bowelobstruction. Additionally, we reviewed rates of major andminor complications within 60 days after surgery. End-organdamage, sepsis, acute kidney injury (AKI) requiring hemodi-alysis (HD), hemorrhage, pneumonia, central line complica-tions, and pulmonary embolism were considered major com-plications while urinary tract infection (UTI), complications ofincision site, AKI not requiring HD, and deep venous throm-bosis (DVT) were considered minor complications. Twelvepatients in each group suffered major complications (40.6 %in AAs vs 26.7 % in C, p=0.22). There was a trend towardhigher rates of minor complications in the AA group; howev-er, it did not reach statistical significance (37.5 % in AA vs10.4 % in C, p=0.07).

Discussion

Our understanding of CD as a disease that primarily affectsCaucasians is changing with an increasing body of evidenceshowing an increased incidence of the disease in other ethnicgroups including AAs. This change in our view of the diseasehas brought up an important question: Does CD present andbehave similarly in AAs and Caucasians? Several studies havetried to answer this question, but the results have been far fromconclusive [21]. Various aspects of racial disparity in CD thathave been studied thus far including prevalence, clinical pre-sentation, course, outcome, and health care utilization haveproduced incongruent results [3, 5, 6, 10, 11, 21–25]. Forexample, several studies have examined the phenotype ofCD in AAs. Initially, several small single-centered retrospec-tive studies reported that ileal and ileo-colonic involvement ismore common among AAs [5–8]. However, a multi-centerretrospective cohort study evaluating 830 Caucasians and 127African–Americans reported that African–Americans had lessileal disease but more colonic disease compared to Caucasians[4]. A qualitative analysis of eight studies of IBD in African–Americans reported that disease location subtypes did notdiffer between African–Americans and Caucasians [9].

Similar conflicting data has been reported for other aspectsof CD in AAs, but in general, data regarding natural historyand course and surgical outcomes of CD in AAs is sparse. Asmall study from 1986 suggested that disease manifestationsin the setting of surgery were more severe in blacks compared

to Caucasian CD patients [26]. Amore recent study from 2005contradicted this notion [8]. In the current study, we examinedseveral preoperative characteristics and surgical outcomes forCD in AAs and compared it to a C population. We did not findsignificant differences between the two populations whenassessing these factors. Both populations had similar insur-ance statuses as well as exposures to immunosuppressives.Although there was a trend toward lower albumin in the AApatients, the finding did not reach statistical significance. Inaddition, bodymass indices did not vary significantly betweenthe two populations. This seems to suggest that the nutritionalstatuses between the two populations were similar. Similarindications for surgery and locations of disease were seenbetween AA and Caucasian CD patients. No differences wereseen with respect to need for repeat surgery as well as rate ofmajor complications postoperatively. More AA patients hadminor complications postoperatively compared to Caucasians,although this finding did not reach statistical significance.Larger scale studies are warranted to determine if AA patientsare at higher risk of postoperative complications.

Differences in surgical outcomes between AAs and Cau-casians have been reported in other diseases. For example,AAs undergo fewer surgeries for treatment of various types ofcancers, despite surgery being recommended as the preferredmethod of treatment [22–24]. AAs have been reported to havehigher morbidity and mortality after a myriad of elective oremergent surgeries [25–30]. For example, Scarborough et al.found that blacks were more likely than whites to developserious postoperative complications such as surgical site in-fection, sepsis, and bleeding [31]. Causey et al. found thatblacks had a higher rate of complications but similar mortalitycompared to whites when undergoing emergent abdominalsurgery [28]. Another found that non-Hispanic blacks under-going bariatric surgery demonstrated higher in hospital mor-tality than whites [32]. Insurance and socioeconomic statushave been suggested as potential mitigating factors; however,studies that have controlled for insurance status did not reportthe same [33–35]. Our study did not show significant differ-ences in surgical outcomes in the setting of similar insurancestatuses between the two populations. Only 6.3 % of our AApatients had no insurance which is about one third of thegeneral US AA population. Future multi-center studies wouldbe useful to determine if insurance status and access to carewould influence outcomes.

Our current study does have limitations. The sample sizewas small, thereby limiting the power of statistical analysis.For example, a non-statistically significant trend toward ahigher rate of minor complications as well as lower albuminlevels was observed in AA patients; however, a larger samplesize may have revealed significant differences. Another limi-tation was that this study included data from a single institu-tion, thereby resulting in a referral and selection bias. Thebenefit of using only one center for the study is that there is

Int J Colorectal Dis

general uniformity in practice among the physicians and sur-geons at our institution. In addition, given that we are a tertiarycare academic referral center, we see a heterogeneous popu-lation from the entire state of Georgia as well as neighboringstates. We also accept patients without insurance, and thosepatients were included in our analysis.

Conclusion

In summary, our study indicates that preoperative characteris-tics and surgical outcomes in AA patients with CD may besimilar to Caucasian CD patients. Further larger scale multi-institutional studies are necessary to determine if these find-ings are reproducible.

Conflict of interest The authors declare that they have no conflict ofinterest.

References

1. Kurata JH, Kantor-Fish S, Frankl H et al (1992) Crohn’s diseaseamong ethnic groups in a large health maintenance organization.Gastroenterology 102:1940–1948

2. White JM, O’Connor S, Winter HS et al (2008) Inflammatory boweldisease in African American children compared with other racial/ethnic groups in a multicenter registry. Clin Gastroenterol Hepatol 6:1361–1369

3. Reddy SI, Burakoff R (2003) Inflammatory bowel disease in AfricanAmericans. Inflamm Bowel Dis 9:380–385

4. Nguyen GC, Torres EA, Regueiro M et al (2006) Inflammatorybowel disease characteristics among African Americans, Hispanics,and non-HispanicWhites: characterization of a large North Americancohort. Am J Gastroenterol 101:1012–1023

5. Basu D, Lopez I, Kulkarni A, Sellin JH (2005) Impact of race andethnicity on inflammatory bowel disease. Am J Gastroenterol 100:2254–2261

6. Cross RK, Jung C,Wasan S et al (2006) Racial differences in diseasephenotypes in patients with Crohn’s disease. Inflamm Bowel Dis 12:192–198

7. Nguyen GC, Bayless TM, Powe NR, Laveist TA, Brant SR (2007)Race and health insurance are predictors of hospitalized Crohn’sdisease patients undergoing bowel resection. Inflamm Bowel Dis13:1408–1416

8. Deveaux PG, Kimberling J, Galandiuk S (2005) Crohn’s disease:presentation and severity compared between black patients and whitepatients. Dis Colon Rectum 48:1404–1409

9. Paul H Jr, Barnes RW, Reese VE et al (1990) Crohn’s disease in blackpatients. J Natl Med Assoc 82:709–712

10. Mahid SS, Mulhall AM, Gholson RD, Eichenberger MR, GalandiukS (2008) Inflammatory bowel disease and African Americans: asystematic review. Inflamm Bowel Dis 14:960–967

11. Straus WL, Eisen GM, Sandler RS, Murray SC, Sessions JT (2000)Crohn’s disease: does race matter? TheMid-Atlantic Crohn’s DiseaseStudy Group. Am J Gastroenterol 95:479–483

12. Wang MH, Okazaki T, Kugathasan S, et al (2012) Contribution ofhigher risk genes and European admixture to Crohn’s disease inAfrican Americans.0 Inflamm Bowel Dis 18(12)2277-87

13. Spivak W, Sockolow R, Rigas A (1995) The relationshipbetween insurance class and severity of presentation ofinflammator bowel disease in children. Am J Gastroenterol90:982–987

14. Smedley BD, Stith AY, Nelson R (2002) Unequal treatment:confronting racial and ethnic disparities in health care. NationalAcademy Press, Washington, DC, pp 29–30

15. Mayberry R, Mili F, Ofili E (2000) Racial and ethnic differences inaccess to health care. Med Care Res Rev 57:108–145

16. Simsek H, Schuman BM (1989) Inflammatory bowel disease in 64black patients: analysis of course, complications, and surgery. J ClinGastroenterol 11:294–298

17. Nguyen GC, LaVeist TA, Harris ML et al (2010) Racial disparities inutilization of specialist care and medications in inflammatory boweldisease. Am J Gastroenterol 105:2202–2208

18. Jackson JF, Dhere T, Repaka A, Shaukat A, Sitaraman S (2008)Crohn’s disease in an African-American population. Am J Med Sci336:389–392

19. Flasar MH, Johnson T, RoghmannMC, Cross RK (2008) Disparitiesin the use of immunomodulators and biologics for the treatment ofinflammatory bowel disease: a retrospective cohort study. InflammBowel Dis 14:13–19

20. Jackson JF, Kornbluth A (2007) Do black and Hispanic Americanswith inflammatory bowel disease (IBD) receive inferior care com-pared with white Americans? Uneasy questions and speculations.Am J Gastroenterol 102:1343–1349

21. Joy GJ, Cross RK, Flasar MH (2009) Race and inflammatory boweldisease. Pract Gastroenterol 23–33

22. Hou JK, El-Serag H, Thirumurthi S (2009) Distribution and mani-festations of inflammatory bowel disease in Asians, Hispanics, andAfrican Americans: a systematic review. Am J Gastroenterol 104:2100–2109

23. Sewell JL, Inadomi JM, Yee HF Jr (2010) Race and inflammatorybowel disease in an urban healthcare system. Dig Dis Sci 55:3479–3487

24. Veluswamy H, Suryawala K, Sheth A et al (2010) African-Americaninflammatory bowel disease in a Southern U.S. health center. BMCGastroenterol 10

25. Goldman CD, Kodner IJ, Fry RD, McDermott RP (1986) Clinicaland operative experience with non-Caucasian patients with CD. DisColon Rectum 29(5):317–321

26. Chow WH, Shuch B, Linehan WM, Devesa SS (2013) Racial dis-parity in renal cell carcinoma patient survival according to demo-graphic and clinical characteristics. Cancer 119:388–394

27. Farjah F, Wood DE, Yanez ND 3rd et al (2009) Racial disparitiesamong patients with lung cancer who were recommended operativetherapy. Arch Surg 144:14–18

28. Causey MW, McVay D, Hatch Q et al (2013) The impact of race onoutcomes following emergency surgery: an American College ofSurgeons National Surgical Quality Improvement Program assess-ment. Am J Surg 206(2):172–179

29. Chu DI, Moreira DM, Gerber L et al (2012) Effect of race andsocioeconomic status on surgical margins and biochemical outcomesin an equal-access health care setting: results from the Shared EqualAccess Regional Cancer Hospital (SEARCH) database. Cancer 118:4999–5007

30. Curry WT Jr, Carter BS, Barker FG 2nd (2010) Racial, ethnic, andsocioeconomic disparities in patient outcomes after craniotomy fortumor in adult patients in theUnited States, 1988-2004. Neurosurgery66:427–437, discussion 37-8

31. Scarborough JE, Bennett KM, Pappas TN (2012) Racial disparities inoutcomes after appendectomy for acute appendicitis. Am J Surg 204:11–17

32. Nguyen GC, Patel AM (2013) Racial disparities in mortality inpatients undergoing bariatric surgery in the USA. Obes Surg 23(10):1508–1514

Int J Colorectal Dis

33. Nathan H, Frederick W, Choti MA, Schulick RD, Pawlik TM (2008)Racial disparity in surgical mortality after major hepatectomy. J AmColl Surg 207:312–319

34. Silber JH, Rosenbaum PR, Romano PS et al (2009) Hospital teachingintensity, patient race, and surgical outcomes. Arch Surg 144:113–120, discussion 21

35. Haider AH, Scott VK, Rehman KA et al (2013) Racial disparities insurgical care and outcomes in the United States: a comprehensive

review of patient, provider, and systemic factors. J Am Coll Surg 216(482–92):e12

This study was approved by the Emory Institutional Review Board andhas therefore been performed in accordance with the ethical standards laiddown in the 1964 Declaration of Helsinki and its later amendments.

Int J Colorectal Dis