ORIGINAL CONTRIBUTION
Does Preoperative Immunosuppression Influence
Unplanned Hospital Readmission After Surgery in
Patients With Crohn’s Disease?
Evan C. White, M.D.1 • Gil Y. Melmed, M.D.2 • Eric Vasiliauskas, M.D.2
Marla Dubinsky, M.D.3 • Andrew Ippoliti, M.D.2 • Dermot McGovern, M.D.2
Stephan Targan, M.D.2 • Phillip Fleshner, M.D.1
1 Division of Colon and Rectal Surgery, Inflammatory Bowel Disease Center, Cedars-Sinai Medical Center, Los Angeles,
California
2 Department of Gastroenterology, Inflammatory Bowel Disease Center, Cedars-Sinai Medical Center, Los Angeles, California
3 Division of Pediatric Gastroenterology, Inflammatory Bowel Disease Center, Cedars-Sinai Medical Center, Los Angeles,
California
BACKGROUND:  Steroids, immunomodulators, and
biologics, often in combination with one another,
are frequently used in the treatment of Crohn’s
disease. Retrospective studies have yielded conflicting
results regarding the influence of preoperative
immunosuppressive therapy on postoperative
complications after surgery in Crohn’s disease.
Unplanned hospital readmission is considered to be an
index of quality surgical care.
OBJECTIVE:  The aim of this study was to examine the
association, if any, between the number of preoperative
immunosuppressive therapies and unplanned hospital
readmission after surgery in patients with Crohn’s disease.
DESIGN:  Consecutive patients with Crohn’s disease
requiring abdominal surgery were identified from
a prospectively maintained database. Preoperative
immunosuppressive therapy within 3 months before
surgery was categorized into 3 classes: steroids,
immunomodulators, and biologics.
MAIN OUTCOME MEASURES:  Unplanned readmission
occurring within 30 days of hospital discharge was
assessed. Trend analysis was performed with the use of
the Cochrane-Armitage test.
Financial Disclosures:  None reported.
Correspondence:  Phillip Fleshner, M.D., 8737 Beverly Blvd, Suite 101,
Los Angeles, CA 90048. E-mail: PFleshner@aol.com
Dis Colon Rectum 2012; 55: 563–568
DOI: 10.1097/DCR.0b013e3182468961
©The ASCRS 2012
Diseases of the Colon & Rectum Volume 55: 5 (2012)
RESULTS:  The study group included 338 patients.
Preoperative medical therapy included steroids (n 5 199;
59%), immunomodulators (n 5 162; 48%), and biologics
(n 5 59; 18%). Sixty-three patients (19%) were not
treated with any immunosuppressive medications
preoperatively, whereas 148 patients (44%), 108 patients
(32%), and 19 patients (6%) were treated with 1, 2, or 3
classes of immunosuppressive medications. Twenty-eight
patients (8.3%) had an unplanned readmission. The
incidence of unplanned readmission was similar among
patients treated with steroids (11%), immunomodulators
(9%), and biologics (12%). The incidence of unplanned
readmission was 3%, 7%, 11%, and 16% in patients
treated with 0, 1, 2, or 3 preoperative medication classes
(trend analysis p 5 0.02). No significant differences were
observed between patient groups treated with 0, 1, 2, or 3
preoperative immunosuppressive therapies with respect
to patient, disease, or surgical factors.
CONCLUSIONS:  Unplanned hospital readmission occurs
frequently (8.3%) after surgery for Crohn’s disease.
Combination immunosuppressive therapy before surgery
in patients with Crohn’s disease appears to be associated
with an increased incidence of postoperative unplanned
hospital readmission.
KEY WORDS:  Unplanned hospital readmission; Crohn’s
disease; Abdominal surgery.
C
rohn’s disease (CD) is a chronic, autoimmune in­­
flammatory condition of the gastrointestinal tract.
Although the exact etiology is unknown, it is thought
to result from an aberrant T-cell immune response to
563
564 White et al: Unplanned Readmission in Crohn’s Disease

microbial or environmental factors in a genetically sus- Assessment of Clinical Characteristics and Preoperative
ceptible individual.1 Medical management is thus directed Medical Therapy
against suppressing this overactive immune response, and Detailed clinical profiles assessing demographic informa-
typically relies on combination therapy with immunosup- tion and characteristics of the disease and its treatment
pressive medications. These medications include a spec- had been previously prospectively generated by one in-
trum of both broadly immunosuppressive and targeted vestigator (P.F.). Patient variables included age at surgery,
therapies and can be generally categorized into 3 classes: sex, previous abdominal surgery, and smoking history.
steroids, immunomodulators, and biologics.2 Patients who smoked at the time of surgery and/or after
Despite advances in medical therapy, most patients surgery were considered to be smokers. Disease character-
with CD will eventually require surgical intervention for istics included disease behavior and location, and the pres-
refractory disease.3 Studies analyzing the influence of pre- ence of a family history of IBD. CD behavior was classified
operative immunosuppressive therapy on postoperative as stricturing, internal penetrating, perianal penetrating,
outcomes in CD patients undergoing abdominal surgery and/or ulcerative colitis-“like.” Anatomic location of dis-
have revealed conflicting results. Whereas some studies ease included ileocecal or ileocolic, ileal, colorectal or co-
have reported that preoperative immunosuppression neg- lonic, continuous ileocolonic, or discontinuous ileocolic.
atively affects operative morbidity,4–8 others have shown Treatment characteristics included the indications for sur-
no effect.9–13 These conflicting results, in part, arise from gery, abscess at time of surgery, and nature of medical ther-
a number of factors, including retrospective study design, apy before surgery. Surgical indications were categorized
small patient numbers, suboptimal patient follow-up, and
referral center bias typical of these studies.
Unplanned hospital readmission (UR) after surgery
has been considered by many to be a reflection of quality TABLE 1.  Clinical features of the study cohort
of care.14–17 Recent federal legislation emphasizes the need
Variable Cohort (n 5 338)
to identify factors leading to preventable readmission, and
includes provisions intended to reduce preventable hospi- Sex (M/F) 176/162
tal readmissions by reducing Medicare payments to hos- Median age at time of surgery, y 33 (11–80)
Smoking history: never smoked 252 (75)
pitals with high preventable readmission rates. Although Family history of IBD 108 (32)
much has been written concerning complications in pa- First operation 221 (65)
tients undergoing colorectal surgery, UR after bowel sur- CD behavior phenotypea
gery remains a poorly evaluated aspect of postoperative   Stricturing 243 (72)
patient care. In this retrospective review of a prospectively   Internal penetrating 152 (45)
  Perianal penetrating 61 (18)
maintained database, we investigated the association be-   Ulcerative colitis-“like” 8 (2)
tween preoperative immunosuppressive therapy within 3 Location of disease
months before surgery and UR after surgery in a longitu-   Ileocecal/ileocolic 223 (66)
dinally followed cohort of consecutive CD patients oper-   Ileal 61 (18)
ated on by a single surgeon.   Colorectal/colonic 25 (7)
  Continuous ileocolonic 15 (4)
  Discontinuous ileocolic 14 (4)
MATERIALS AND METHODS Surgical indicationa
  Obstruction/stricture 201 (60)
Study Population   Perforation/abscess 49 (15)
A prospective database of CD patients requiring abdomi-   Fistula 45 (13)
nal surgery was queried. Consecutive operations done   Medically refractory disease 43 (13)
  Other 13 (4)
between December 1999 and July 2009 were included in Type of surgery
the analysis. All procedures were performed by one sur-   Laparoscopic 202 (60)
geon (P.F.) using an open or laparoscopic technique, and   Open 136 (40)
included resection, strictureplasty, or both. All patients Abscess at surgery 55 (16)
were included in a hospital-wide standardized fast-track Ostomy 7 (2)
Median length of postoperative stay, days 5 (3–10)
protocol. There were no significant changes in the periop- Preoperative medical therapya
erative care during the study period. All patients received a   Steroids 199 (59)
single dose of parenteral antibiotics at the time of incision,   Immunomodulators 162 (48)
and prophylactic sequential compression devices. Patients   Biologics 59 (18)
were fully ambulatory when discharged to home tolerat- The numbers in parentheses indicate percentage, with the exception of median
age and median length of postoperative stay (range).
ing a low-residue diet. All research-related activities were
CD 5 Crohn’s disease.
approved by the Cedars-Sinai Medical Center Institutional a
.100% because of multiple phenotypes, surgical indications, and/or medication
Review Board (no. 3358). class in the same patient.
Diseases of the Colon & Rectum Volume 55: 5 (2012) 565

as obstruction/stricture, medically refractory disease, fis-
tula, perforation/abscess, and other (including bleeding,
toxic dilation, dysplasia, or pseudopolyps). The use of a
temporary or permanent stoma was recorded. Preopera-
tive medical therapy within 3 months before surgery was
recorded and categorized into 3 classes: steroids, immuno-
modulators, and biologics. The steroid category included
treatment with prednisone or budesonide. Immunomodu-
lators included 6-mercaptopurine, 6-thioguanine, tacroli-
mus, thalidomide, purinethol, azathioprine, cyclosporine,
and methotrexate. The biologic category included inflix-
imab, adalimumab, and certolizumab.
Unplanned Hospital Readmission
UR was defined as readmission occurring within 30 days
of hospital discharge. Reasons for readmission were com-
piled. All URs were unanticipated and resulted from a
complication. Planned hospital visits for routine postop-
erative care such as stoma evaluation were excluded from
the study. In addition to review of Cedars-Sinai Medical
Center computer records, all patients were contacted to
rule out the possibility of readmission to a different hos-
pital after discharge. Hospital records of those admissions
were also reviewed. Readmission diagnosis and the inter-
val between discharge and readmission, as well as length of
hospital stay after readmission, were recorded.
Statistical Analysis
All data were entered into a standardized database com-
puter program (Microsoft Excel, Seattle, WA). Univari-
ate analysis for continuous or categorical variables was
performed with the use of the Mann-Whitney or Fisher
exact test. Trend analysis was performed with the use of
the Cochrane-Armitage test. All hypothesis testing was
2-sided with a p value of less than 0.05 considered statisti-
cally significant.
RESULTS
The study group included 338 CD patients who required
abdominal surgery during the study period. Clinical char-
acteristics of the study population are shown in Table 1.
The disease most commonly was the stricturing pheno-
type involving the ileocecal/ileocolic region. As expected,

TABLE 2.  Clinical features by number of classes of immunosuppressive medications

Variable 0 (n 5 63) 1 (n 5 148) 2 (n 5 108) 3 (n 5 19)
Sex (M/F) 33/30 79/69 56/52 8/11
Median age at time of surgery, y 39 (13–73) 33 (11–80) 28 (12–75) 28 (14–50)
Smoking history: never smoked 47 (75) 108 (73) 82 (76) 15 (79)
Family history of IBD 17 (27) 51 (35) 37 (34) 3 (16)
First operation 34 (54) 96 (65) 78 (72) 13 (68)
CD behavior phenotypea
  Stricturing 47 (75) 104 (70) 77 (71) 15 (79)
  Internal penetrating 24 (38) 64 (43) 59 (55) 5 (26)
  Perianal penetrating 14 (22) 24 (16) 19 (18) 4 (21)
  Ulcerative colitis-“like” 1 (2) 2 (1) 4 (4) 1 (5)
Location of disease
  Ileocecal/ileocolic 35 (56) 101 (68) 74 (69) 13 (68)
  Ileal 11 (18) 27 (18) 23 (21) 0 (0)
  Colorectal/colonic 7 (11) 12 (8) 4 (4) 2 (11)
  Continuous ileocolonic 6 (10) 6 (4) 3 (3) 0 (0)
  Discontinuous ileocolic 3 (5) 4 (3) 4 (4) 3 (16)
Surgical indicationa
  Obstruction/stricture 38 (60) 88 (60) 62 (57) 13 (68)
  Perforation/abscess 5 (8) 26 (18) 17 (16) 1 (5)
  Fistula 7 (11) 21 (14) 15 (14) 2 (11)
  Medically refractory disease 9 (14) 15 (10) 15 (14) 4 (21)
  Other 5 (8) 5 (3) 3 (3) 0 (0)
Type of surgery
  Laparoscopic 36 (57) 92 (62) 60 (56) 14 (74)
  Open 27 (43) 56 (38) 40 (44) 13 (26)
Abscess at surgery 6 (10) 30 (20) 17 (16) 2 (11)
Ostomy 0 (0) 4 (3) 3 (3) 0 (0)
Median length of postoperative stay, days 5 (4–9) 5 (3–10) 4 (4–9) 5 (3–10)
Unplanned hospital readmission 2 (3) 11 (7) 12 (11) 3 (16)
The numbers in parentheses indicate percentage, with the exception of median age and length of postoperative stay (range).
CD 5 Crohn’s disease.
a
.100% because of multiple phenotypes and surgical indications in the same patient.
566 White et al: Unplanned Readmission in Crohn’s Disease

the most common indication for surgery was obstruction. Twenty-eight patients (8.3%) had a UR. The median
Most patients (60%) had their surgery performed lap- time between hospital discharge and UR was 9 days (range,
aroscopically. Immunomodulators were used in almost 1–30). The median length of hospital stay after readmis-
one-half of patients and included 6-mercaptopurine (n sion was 4 days (range, 1–29). Readmission diagnoses
5 136), methotrexate (n 5 57), cyclosporine (n 5 26), included intra-abdominal abscess (n 5 10), small-bowel
6-thioguanine (n 5 12), and tacrolimus (n 5 9). Biologic obstruction (n 5 4), enterocutaneous fistula (n 5 3),
therapies included infliximab (n 5 38), adalimumab (n 5 bleeding (n 5 3), wound infection (n 5 3), dehydration
17), and certolizumab (n 5 4). (n 5 2), anasarca (n 5 1), pancreatitis (n 5 1), and kidney
Sixty-three patients (19%) were not treated with any stones (n 5 1). The incidence of UR was similar among
immunosuppressive medications preoperatively, whereas patients treated with steroids (11%), immunomodulators
148 patients (44%), 108 patients (32%), and 19 patients (9%), and biologics (12%).
(6%) were treated with 1, 2, or 3 classes of immunosup- Comparison of clinical factors between patients with
pressive medications (Table 2). No significant differences a UR and those without a UR is shown in Table 3. Patients
were observed between patient groups treated with 0, 1, 2, with a UR were more likely to have undergone previous
or 3 preoperative immunosuppressive therapies with re- abdominal surgery. Patients who received steroids had a
spect to patient, disease, or surgical factors. higher rate of UR than patients who did not receive ste-
roids (11% vs 4%; p 5 0.03). The incidence of UR (Fig. 1)
was 3%, 7%, 11%, and 16% in patients treated with 0, 1,
2, or 3 preoperative medication classes (trend analysis p 5
TABLE 3.  Clinical features of the unplanned readmission patient 0.02).
group
Unplanned No unplanned
readmission readmission DISCUSSION
Variable (n 5 28) (n 5 310) p
Steroids, immunomodulators, and biologics, often in com-
Sex (M/F) 18/10 158/152 0.24 bination with one another, are frequently used in the pre-
Median age at time of surgery, y 29 (15–66) 33 (11–80) 0.67
Smoking history: never smoked 17 (61) 235 (76) 0.11
operative treatment of CD. This study sought to determine
Family history of IBD 10 (36) 98 (32) 0.68 the influence of preoperative immunosuppressive therapy
First operation 12 (43) 209 (67) 0.01 on the incidence of UR after surgery in CD patients. On
CD behavior phenotypea 22 (79) 221 (71) 0.51 trend analysis, combination therapy with increasing num-
  Stricturing 13 (46) 139 (45) 1 ber of classes of immunosuppressive medications was as-
  Internal penetrating 8 (29) 53 (17) 0.13
  Perianal penetrating 0 (0) 8 (3) 1
sociated with an increased rate of UR.
  Ulcerative colitis-“like” This study is unique because it only included preop-
Location of disease 19 (68) 204 (66) 1 erative medical therapy within 3 months before surgery.
  Ileocecal/ileocolic 4 (14) 57 (18) 0.80 Other studies have considered preoperative medical treat-
  Ileal 1 (4) 24 (8) 0.71 ment at any time,5,6,11 although recent immunosuppres-
  Colorectal/colonic 3 (11) 12 (4) 0.12
  Continuous ileocolonic 1 (4) 13 (4) 1
sive therapy before surgery is more likely to have an effect
  Discontinuous ileocolic on postoperative complications. This study is also unique
Surgical indicationa 19 (68) 182 (59) 0.42 in that it categorized immunosuppressive ­therapy into 3
  Obstruction/stricture 4 (14) 45 (15) 1 separate classes and performed a trend analysis to ­analyze
  Perforation/abscess 1 (4) 44 (14) 0.15
  Fistula 2 (7) 41 (13) 0.55
  Medically refractory disease 2 (7) 11 (3) 0.29 Unplanned readmission (%)
Type of surgery 25
  Laparoscopic 17 (61) 186 (60) 0.91
  Open 11 (39) 114 (40) 0.90 20
Abscess at surgery 6 (21) 49 (16) 0.43
Ostomy 2 (7) 5 (2) 0.12 15
Median length of postoperative 5 (3–9) 5 (3–10) 0.80
   stay, days 10
Preoperative medical therapya
  Steroids 22 (79) 177 (57) 0.03 5
  Immunomodulators 15 (54) 147 (47) 0.56
  Biologics 7 (25) 52 (17) 0.30
0 1 2 3
The numbers in parentheses indicate percentage, with the exception of median
age and median length of postoperative stay (range).
Preoperative medical therapy (# of classes)
CD 5 Crohn’s disease.
a
.100% because of multiple phenotypes, surgical indications, and/or medication FIGURE 1.  Preoperative combination immunosuppressive therapy
class in the same patient. and postoperative unplanned readmission.
Diseases of the Colon & Rectum Volume 55: 5 (2012)
a­ dditive effects by number of classes of medications.
­Although other studies have examined the association
­between preoperative immunosuppressive medical ther-
apy and postoperative complications in CD patients,5–13
they did not analyze UR as an end point. Finally, this study
utilized a prospectively maintained database, and all oper-
ations were performed by a single surgeon. These aspects
of study design help minimize confounding factors such
as surgeon technique, which may influence UR.
Preoperative treatment with steroids has been associ-
ated with an increased rate of postoperative complications
in other studies.5-7 Furthermore, studies have shown a spe-
cific correlation between preoperative steroid treatment
and UR in IBD patients undergoing abdominal surgery.18,19
This study confirms these prior results, and also shows
that combination therapy with increasing numbers of
classes of immunosuppressive medications increases the
incidence of UR. In contrast to our results, Appau et al4
reported a higher incidence of UR following ileocolic re-
section in patients with CD who received infliximab in the
perioperative period. In our series, preoperative therapy
with immunomodulators or biologics (which included in-
fliximab as well as other biologics) alone did not increase
readmission rates, yet concomitant immunosuppressive
therapy did. The reasons for this observation are unknown,
but it is conceivable that increasing preoperative immuno-
suppression with combination therapy puts the patient at
increased risk of postoperative infectious complications
that may lead to UR. The most common reason for a UR
in this study was intra-abdominal abscess. Furthermore,
it is known that steroids decrease wound healing, and this
may also be a factor. With increasing number of classes of
immunosuppressive medications preoperatively, it follows
that the patient may be increasingly immunosuppressed
and thus be at increased risk for postoperative complica-
tions leading to UR.
Our study has several limitations. First, although data
were derived from a prospectively maintained database,
this study was retrospective in nature, and thus suffers
from limitations inherent to its retrospective design. As a
referral center, our patient population may have more ag-
gressive or advanced disease at the time of surgery, which
may have influenced the rate of UR. In addition, the study
did not account for how debilitated patients were at the
time of surgery, and there was no consistent recording of
laboratory factors such as albumin or complete blood cell
count that may shed light on the patient’s overall condi-
tion. Other factors that may have influenced UR such as
initial length of stay were not assessed. Furthermore, with
a UR rate of only 28 events, our study may have been un-
derpowered and at risk for a type II error. In addition, an-
other study limitation is our choice to look only at UR,
whereas other surgical morbidities such as postoperative
complications were not assessed. Finally, our results are
567
reported from the database of a single surgeon, and thus
may not be readily generalized.
CONCLUSION
This study suggests that treatment with steroids within 3
months before abdominal surgery increases the rate of UR
in CD patients. This study also shows that perioperative
treatment with immunomodulators or biologics alone
does not increase UR. However, combination immuno-
suppressive medical therapy does increase the rate of UR
in CD patients undergoing abdominal surgery. These
findings have clinical implications, and patients receiving
perioperative combination immunosuppressive therapy
before abdominal surgery should be counseled about the
increased risk of UR.
REFERENCES
1. Thoreson R, Cullen JJ. Pathophysiology of inflammatory bowel
disease: an overview. Surg Clin North Am. 2007;87:575–585.
2. Lichtenstein GR, Hanauer SB, Sandborn WJ. Manage-
ment of Crohn’s disease in adults. Am J Gastroenterol.
2009;104:465–483.
3. Cosnes J, Nion-Larmurier I, Beaugerie L, Afchain P, Tiret E,
Gendre JP. Impact of the increasing use of immunosuppres-
sants in Crohn’s disease on the need for intestinal surgery. Gut.
2005;54:237–241.
4. Appau KA, Fazio VW, Shen B, et al. Use of infliximab within
3 months of ileocolonic resection is associated with adverse
postoperative outcomes in Crohn’s patients. J Gastrointest Surg.
2008;12:1738–1744.
5. Yamamoto T, Allan RN, Keighley MR. Risk factors for intra-
abdominal sepsis after surgery in Crohn’s disease. Dis Colon
Rectum. 2000;43:1141–1145.
6. Post S, Betzler M, von Ditfurth B, Schürmann G, Küppers P,
Herfarth C. Risks of intestinal anastomoses in Crohn’s disease.
Ann Surg. 1991;213:37–42.
7. Alves A, Panis Y, Bouhnik Y, Pocard M, Vicaut E, Valleur P. Risk
factors for intra-abdominal septic complications after a first
ileocecal resection for Crohn’s disease: a multivariate analysis in
161 consecutive patients. Dis Colon Rectum. 2007;50:331–336.
8. Myrelid P, Olaison G, Sjödahl R, Nyström PO, Almer S, An-
dersson P. Thiopurine therapy is associated with postoperative
intra-abdominal septic complications in abdominal surgery
for Crohn’s disease. Dis Colon Rectum. 2009;52:1387–1394.
9. Tay GS, Binion DG, Eastwood D, Otterson MF. Multivari-
ate analysis suggests improved perioperative outcome in
Crohn’s disease patients receiving immunomodulator ther-
apy after segmental resection and/or strictureplasty. Surgery.
2003;134:565–572.
10. Indar AA, Young-Fadok TM, Heppell J, Efron JE. Effect of peri-
operative immunosuppressive medication on early outcome in
Crohn’s disease patients. World J Surg. 2009;33:1049–1052.
11. Marchal L, D’Haens G, Van Assche G, et al. The risk of post-
operative complications associated with infliximab therapy for
568
Crohn’s disease: a controlled cohort study. Aliment Pharmacol
Ther. 2004;19:749–754.
12. Kunitake H, Hodin R, Shellito PC, Sands BE, Korzenik J, Bor-
deianou L. Perioperative treatment with infliximab in patients
with Crohn’s disease and ulcerative colitis is not associated with
an increased rate of postoperative complications. J Gastrointest
Surg. 2008;12:1730–1736.
13. Colombel JF, Loftus EV Jr, Tremaine WJ, et al. Early posto­
perative complications are not increased in patients
with Crohn’s disease treated perioperatively with inflix-
imab or immunosuppressive therapy. Am J Gastroenterol.
2004;99:878–883.
14. Welch HG, Larson EH, Hart LG, Rosenblatt RA. Readmission
after surgery in Washington State rural hospitals. Am J Public
Health. 1992;82:407–411.
White et al: Unplanned Readmission in Crohn’s Disease
15. Healey MA, Shackford SR, Osler TM, Rogers FB, Burns E. Com-
plications in surgical patients. Arch Surg. 2002;137:611–617.
16. Courtney ED, Ankrett S, McCollum PT. 28-Day emergency sur-
gical re-admission rates as a clinical indicator of performance.
Ann R Coll Surg Engl. 2003;85:75–78.
17. Shively EH, Heine MJ, Schell RH, et al. Practicing sur-
geons lead in quality care, safety, and cost control. Ann Surg.
2004;239:752–760.
18. Kariv Y, Wang W, Senagore AJ, Hammel JP, Fazio VW, Dela-
ney CP. Multivariable analysis of factors associated with
hospital readmission after intestinal surgery. Am J Surg.
2006;191:364–371.
19. Kiran RP, Delaney CP, Senagore AJ, Steel M, Garafalo T, Fazio
VW. Outcomes and prediction of hospital readmission after
intestinal surgery. J Am Coll Surg. 2004;198:877–883.