Antibiotics versus no antibiotic therapy for uncomplicated

sigmoid diverticulitis: a meta-analysis

Willmar Jayve M. Añoso, MD and Omar O. Ocampo, MD

Abstract
Introduction Antibiotics have been used customarily in the treatment of uncomplicated diverticulitis
since their introduction and have become the standard of care. The aim of this study is to compare
the effectiveness of antibiotic therapy versus no antibiotic therapy in the treatment of uncomplicated
sigmoid diverticulitis.
Methods An electronic search for randomized controlled trials comparing antibiotics versus no
antibiotic therapy for uncomplicated diverticulitis was conducted. The outcomes considered were
associated morbidity (abscess formation and sigmoid perforation); need for sigmoid colon resection,
and recurrence of diverticulitis. The included studies were evaluated for risk of bias. Meta-analysis
with Forest plot was performed using Review Manager Version 5.3.
Results Two trials, consisting of 1,151 subjects, were included in the meta-analysis. There was no
difference in the risk of sigmoid perforation (RR 1.02, 95% CI 0.30, 3.49). Abscess formation and
incidence of sigmoid resection were lower in the antibiotics groups (RR 2.24, 95 CI 0.51, 9.95 and
RR 1.59, 95% CI 0.75, 3.36, respectively) but the differences were not significant. There was no
difference in the recurrence of diverticulitis (RR 1.05, 95% CI 0.74, 1.48) between the two groups.
Conclusion There is no definite advantage in giving antibiotics to patients with uncomplicated
diverticulitis. Not giving antibiotics may be an acceptable treatment option for patients with acute
uncomplicated sigmoid diverticulitis.

Key words: diverticular disease, uncomplicated diverticulitis, conservative management, antibiotics,

treatment

D iverticulitis refers to inflammation and infection

associated with a diverticulum in the bowel.
pathophysiologic mechanism underlying sigmoid
diverticulitis is not well-understood. However, the
long-held belief that colonic diverticulitis is caused
The sigmoid colon is the most common site.¹´² The by microperforation and bacterial infection has been
challenged by the concept that diverticulitis may be
a primary inflammatory process.³
Uncomplicated sigmoid diverticulitis is the
Correspondence: presence of inflamed sigmoid colon diverticula in
Willmar Jayve M. Añoso, MD, Department of Surgery, University
the absence of complications such as perforation
of East Ramon Magsaysay Memorial Medical Center, 64 Aurora
Boulevard, Barangay Dona Imelda, Quezon City 1113; leading to peritonitis, fistula formation, obstruction
E-mail: [email protected] due to stricture, and/or bleeding. It is usually

VOL. 7 NO.1 • JANUARY- JUNE 2018 • UERM Health Sciences Journal 29

Antibiotics versus no antibiotic therapy for uncomplicated sigmoid diverticulitis

characterized by an acute onset of left lower quadrant to be stages 1a and 1b according to the Modified
abdominal pain and tenderness. Computed Hinchey's classification or "mild" diverticulitis
tomography findings include the presence of colonic according to Ambrosetti's criteria were included. The
diverticula with associated pericolic soft tissue a
trials should have determine at least one of the
stranding, colonic wall thickening, and/ or phlegmon following outcomes: morbidity, specifically sigmoid
formation.¹ perforation and abscess formation; incidence of
Most patients with uncomplicated diverticulitis surgical resection (sigmoidectomy or partial
in the sigmoid colon will respond to conservative colectomy); and recurrence. Trials that also looked
treatment in an out-patient setting.¹´² Conservative into length of hospital stay and abdominal pain and
treatment of mild colonic diverticulitis typically tenderness were considered.
includes careful observation while placing the patient All randomized clinical trials comparing
on a low-residue diet with broad-spectrum oral antibiotics versus no antibiotic therapy in patients
antibiotics.2 In a small number of patients with a more diagnosed with uncomplicated sigmoid diverticulitis
serious presentation such as severe abdominal pain, were identified by conducting an electronic search
tenderness, fever, and leukocytosis, in-hospital of the databases from Cochrane Library, PubMed
treatment with bowel rest and parenteral antibiotics and Google Scholar using the following keywords:
is done.² The majority of patients with uncomplicated diverticular disease, uncomplicated diverticulitis,
sigmoid diverticulitis improves with these sigmoid diverticulitis, conservative management,
conservative measures and recovers without surgery.²´⁴ antibiotics, treatment. A comprehensive hand search
Antibiotics have been used customarily in the of reference lists of published articles and review
treatment of uncomplicated left-sided colonic articles was performed to ensure inclusion of all
diverticulitis and have become the standard of care. possible studies and to exclude duplicates. Included
Recently, a Cochrane database review found were articles published in English up to May 2017.
contradictory results between two randomized Review articles, non-randomized trials, retrospective
controlled trials with regard to the use of antibiotics analyses, and abstracts were not considered.
in the treatment of uncomplicated sigmoid All potential trials were screened according to
diverticulitis.⁵ The DIABOLO trial investigated the the criteria specified in the research protocol. Three
cost-effectiveness of treatment with or without reviewers extracted data from each publication. The
antibiotics for uncomplicated acute sigmoid third reviewer served as the arbiter who resolved all
diverticulitis showed no significant difference in full discrepancies. The quality of included studies was
recovery between the two strategies.⁶ On the other assessed independently by three reviewers using the
hand, the AVOD study group showed antibiotic Cochrane Handbook for Systematic Reviews of
treatment neither accelerates recovery nor prevents Interventions criteria: random sequence generation,
complications or recurrence of diverticulitis.⁷ allocation concealment, blinding of the patient and
The aim of our study is to compare the the observer, blinding of outcome assessment,
effectiveness of antibiotic therapy versus no antibiotic incomplete outcome data, and selective outcome
therapy in the treatment of uncomplicated sigmoid reporting. The main comparison was stratified
diverticulitis. Specifically, it aims to compare the according to morbidity, incidence of surgery,
incidence of morbidity in terms of perforation and recurrence and length of hospital stay.
abscess formation, sigmoid resection, and recurrence. The relative risk or risk ratio (RR) was the
The study also compares the length of hospital stay primary measure of treatment effect or adverse events,
and clinical signs between the two treatment strategies. and 95 per cent confidence intervals (CI) for RR were
calculated. Heterogeneity was assessed by Q-square
(v2) and I -square statistics (I).⁸´⁹ The I² statistic
Methods indicated the degree of between-study or interstudy
Trials that were included in the study are those variability as opposed to within-study or intra-study
comparing antibiotic therapy versus no antibiotic variability. An I² value greater than 50% was
therapy for uncomplicated sigmoid diverticulitis considered as substantial heterogeneity.⁸´⁹ Studies were
published in English. Randomized controlled trials analyzed using the fixed-effects model; when
wherein the participants were diagnosed by CT scan heterogeneity was significant, the random-effects

30 UERM Health Sciences Journal•VOL.7 NO.1• JANUARY·JUNE 2018

----
 Antibiotics versus no antibiotic therapy for uncomplicated sigmoid diverticulitis
1.

model described by DerSimonian and Laird was group. The AVOD and DIABOLO studies followed
used.⁹ Meta-analysis using Forest plots was performed up their patients for 1 and 2 years, respectively. Both
with Review Manager Version 5.3. studies satisfied the criteria for assessment except for
blinding of patients and personnel. Assessment of
quality according to the Cochrane Collaboration's
Results tool for assessing risk of bias summary for RCTs is
The search strategy resulted in 491 studies from the reported in Figure 2.
online databases. After excluding duplicates and
articles that did not meet the inclusion criteria, seven
full texts were reviewed for eligibility. Five articles
were excluded because they were not randomized

Blinding of participants and personnel (performance bias)

trials, leaving two for inclusion in the meta-analysis
as seen in Figure 1.
Two trials were included in the meta-analysis and

Blinding of outcome assessment (detection bias)

their characteristics are summarized in Table 1. The

Random sequence generation (selection bias)

two studies included a total of 1,151 participants -

Incomplete outcome data (attrition bias)

Allocation concealment (selection bias)
571 in the no antibiotic group and 580 in the antibiotic

selective reporting (reporting bias)

491 records identified 486 records excluded
through database searching
Pubmed:312 Duplicate publications
Cochrane Library: 44 Complicated Diverticulitis
Google Scholar: 13S5 Inclusion cirteria not meet

other bias
7 full text articles assessed 5 full text articles excluded
for eligibility Not randomized design: 5
AVOD TRIAL + + - + + +
DIABOLO TRIAL + + - + + +

2 RCT studies included for

meta-analysis

Figure 1. Study selection flowchart. Figure 2. Risk of bias summary. (Gray - low risk of bias, Black
- high risk, white - unclear)

Table 1. Summary of included studies.

Characteristic AVOD trial, 20126 DIABOLO trial, 20167

Population 528
623
Follow up (months)
12 24
Antibiotics used Cefuroxime / cefotaxime plus Amoxycillin-clavulanic acid IV shifted to oral
metronidazole, or carbapenem,
or piperacillin-tazobactam,
shifted to oral ciprofloxacin/
cefadroxil plus metronidazole
Outcomes Complications (abscess formation, Recovery, days spent outside hospital (6 months),
perforation), emergency surgery, readmissions, complicated diverticulitis (abscess,
hospital stay, recurrence perforation, obstruction/stricture, bleeding, fistula),
ongoing diverticulitis, recurrence, need for resection/
surgery, other adverse events, mortality

VOL. 7 NO.1•JANUARY-JUNE 2018 • UERM Health Sciences Journal 31

Antibiotics versus no antibiotic therapy for uncomplicated sigmoid diverticulitis

Figure 3 demonstrates the incidence of sigmoid The sigmoid was resected in 17 out of 554
perforation between no antibiotics versus with subjects (3.0%) in the no antibiotics versus 11 out
antibiotics showing no statistical difference (overall of 569 subjects (1.9%) in the-antibiotics group. The
effect P = 0.98). The incidence was 5 out of 566 risk ratio of 1.59 (95% tends to favor with antibiotics
subjects (0.88%) for the no antibiotics group versus group but is inconclusive as seen in Figure 5. There
5 out of 575 subjects (0.86) for the antibiotics group. was recurrence in 56 out of 552 subjects (10.1%) in
The difference is not significant. As seen in Figure the no antibiotics group compared with 54 out of
4, abscess formation was seen in 5 out of 566 subjects 558 subjects (9.7%) in the antibiotics group. The
(0.9%) for the no antibiotics group compared with 2 overall risk ratio was 1.05 (95% CI, 0.74-1.48),
out of 578 subjects for the antibiotics group (0.3%). indicating no statistical difference in recurrence
The overall risk ratio was 2.24 (95% CI, 0.51-9.95) between the antibiotics and no antibiotics group
and tends to favor the antibiotics group but is (Figure 6).
inconclusive.

32 UERM Health Sciences Journal • VOL. 7 NO.1• JANUARY - JUNE 2018

 Antibiotics versus no antibiotic therapy for uncomplicated sigmoid diverticulitis

The median length of hospital stay in both groups by inspissated stool in a diverticulum. However, an
was comparable. In the AVOD trial, median length overlap between diverticulitis and inflammatory
of hospital stay was 3 days in the no antibiotics bowel disease has long been recognized, and recent
group versus 3 days in the antibiotics group.⁷ studies have postulated that all diverticular disease
Whereas in the DIABOLO trial, results could be a form of inflammatory bowel disease.¹⁸
demonstrated median length of hospital stay of 2 Altering the inflammatory response in cases of mild
days in the no antibiotic group versus 3 days in the diverticulitis may be a more logical step than giving
antibiotics group.⁶ Clinical bedside signs, such as antibiotics.⁵
abdominal pain, measured by VAS and tenderness Antibiotic resistance has become a worldwide
on abdominal palpation at admission, did not differ problem and the use of antibiotics has other
between the groups. potential side effects such as nausea and
vomiting, development of Clostridium difficile
colitis, and fatal allergic reactions, among others.⁵
Discussion
The possible development of such symptoms
In this meta-analysis, treatment strategies using
provides another important reason for reducing
antibiotics versus no antibiotic therapy for
the frequent use of antibiotics in these patients.⁷
uncomplicated sigmoid diverticulitis demonstrated
It could also be concluded that the risk of adverse
no statistical difference between the two groups in
effects and high costs warrant selective use
terms of complications such as sigmoid perforation
whenever possible.¹⁴
and abscess formation; incidence of sigmoid colon
A limitation of this study is the need for more
resection, and recurrence of diverticulitis. The risk
randomized controlled studies comparing the use
ratio for patients who had undergone sigmoid colon
of antibiotics versus no antibiotics for acute
resection (RR 1.59, 95% CI 0.75, 3.36) and patients
who had abscess formation (RR 2.24, 95% CI 0.51, uncomplicated sigmoid diverticulitis.
9.95) tends to favor the antibiotic group, but is In this study, treatment strategies using
inconclusive. antibiotics versus no antibiotic therapy for
Guidelines regarding treatment strategies for uncomplicated sigmoid diverticulitis demonstrated
acute uncomplicated sigmoid diverticulitis have comparable results in terms of complications
remained unchanged. According to Chabok, the (sigmoid perforation, abscess formation), incidence
recommendations for giving antibiotic therapy are of sigmoid colon resection, and recurrence. The risk
based on tradition and expert opinions, and not on ratio for patients who had undergone sigmoid colon
evidence derived from controlled clinical trials.⁷ resection and patients who had abscess formations
Current guidelines have stated bowel rest or intake tends to favor the antibiotic group, but is still
of oral fluids and a 7 to l0-day regimen of broad- inconclusive. Thus, there is no definite advantage in
spectrum antibiotics is recommended in patients with giving antibiotics to patients with uncomplicated
uncomplicated sigmoid diverticulitis.¹º⁻¹² Meanwhile, diverticulitis. From these results, it may be postulated
the likelihood that treatment of uncomplicated that not giving antibiotics may be an acceptable
sigmoid diverticulitis may not require antibiotics was treatment option for patients with acute
described and has been raised in literature.¹³ Two uncomplicated sigmoid diverticulitis.
observational studies have also indicated that
antibiotic therapy for uncomplicated sigmoid
diverticulitis showed no benefit.¹⁴`¹⁵ Furthermore, two References
2. Cameron JL, Cameron AM. Current Surgical Therapy
more retrospective cohort studies have indicated that
12th edition. Philadelphia: Elsevier, Inc.; 2017.
a no-antibiotic policy for acute uncomplicated 3. Brunicardi, FC. Schwartz's Principles of Surgery 10th
diverticulitis is feasible and safe. ¹⁶`¹⁷ Edition. New York: McGraw-Hill Education: 2015.
The rationale behind treating an episode of 4. De Korte N, Unlu C, et al. Use of antibiotics in
uncomplicated left-sided colonic or sigmoid uncomplicated diverticulitis. Br J Surg 2011: 98: 761-7.
diverticulitis with antibiotics lies in the fact that it 5. Shaikh S, Krukowski ZH. Outcome of a conservative
has long been believed that all forms of diverticulitis policy for managing acute sigmoid diverticulitis. Br J
are the result of a colonic microperforation caused Surg 2007, 94: 876-9.

VOL. 7 NO.1•JANUARY· JUNE 2018 • UERM Health Sciences Journal 33

Antibiotics versus no antibiotic therapy for uncomplicated sigmoid diverticulitis

5. Shabanzadeh DM, Wille-Jorgensen P. Antibiotics for 15. Hjern F, Josephson T, Altman D, et al. Conservative
uncomplicated diverticulitis. Cochrane Database Syst treatment of acute colonic diverticulitis: Are antibiotics
Rev 2012; 11: CD009092. always mandatory? Scand J Gastroenterol2007; 42: 41-
6. Daniels L, Unlu C, et al. Randomized clinical trial of 7. ~
observational versus antibiotic treatment for a first 16. Isacson D., Kalle Andreasson K., et al. No antibiotics
episode of CT -proven uncomplicated acute in acute uncomplicated diverticulitis: Does it work?
diverticulitis. Br J Surg 2017; 104: 52-61. Scand J Gastroenterol 2014; 49: 1441-6.
7. Chabok A, Pahlman L, et al. Randomized clinical trial 17. Brochmann N, Schultz JK, Jakobsen GS, 0resland T.
of antibiotics in acute uncomplicated diverticulitis. Br J Management of acute uncomplicated diverticulitis
Surg 2012; 99: 532-9. without antibiotics: A single centre cohort study.
8. Higgins JPT, Green S. Cochrane Handbook for Colorectal Dis 2016; 18(11): 1101-7.
Systematic Reviews of Interventions Version 5.1.0. 18. Hinchey EJ, Schaal PG, Richards GK. Treatment of
Copenhagen: The Cochrane Collaboration; 2011. perforated diverticular disease of the colon. Adv Surg
9. DerSimonian R, Laird N. Meta-analysis in clinical trials. 1978; 12: 85-109.
Control Clin Trials 2006; 7: 177-88. 19. Kaiser AM, Jiang JK, Lake JP, et al. The management
10. American Society of Colon and Rectal Surgeons. of complicated diverticulitis and the role of computed
Available from: https:/ /www.fascrs.org/education/ tomography. Am J Gastroenterol 2005; 100: 910-7.
core-subj ects. 20. Wasvary H, Turfah F, Kadro 0, Beauregard W. Same
11. National Health Service (NHS). Available from: http:/ hospitalization resection for acute diverticulitis. Am
/ wwwnhs.uk / Conditions /Diverticular-disease-and- Surg 1999; 65(7): 632-5.
diverticulitis/ Pages/Treatment.aspx. 21. Ambrosetti P, Grossholz M, Becker C, Terrier F, Morel
12. Feingold D, Steele SR, Lee S, et al. Practice parameters P: Computed tomography in acute left colonic
for the treatment of sigmoid diverticulitis. Dis Col Rect diverticulitis. Br J Surg 1997; 84: 532-4.
2014; 57: 284-94. 22. Klarenbeek BR, de Korte N, et al. Review of current
13. McDermott FD, Collins D, Heeney A, Winter DC. classifications for diverticular disease and a translation
Minimally invasive and surgical management strategies into clinical practice. Int J Colorect Dis 2012; 27(2): 207-
tailored to the severity of acute diverticulitis. Br J Surg 14.
2014; 101: e90-e99. 23. Rankin F. Diverticulitis of the colon. Surg Gynecol
14. De Korte N, Kuyvenhoven JP, van der Peet DL, Felt- Obstet 1930; 50: 836-47.
Bersma RJ, Cuesta MA, Stockmann HB. Mild colonic 24. Thorisson A, Smedh S, et al. CT imaging for prediction
diverticulitis can be treated without antibiotics: A case- of complications and recurrence in acute uncomplicated
control study. Colorectal Dis 2011; 14(3): 325-30. diverticulitis. Int J Colorect Dis 2016; 31: 451-7.
25. Isacson D, Thorisson A, et al. Outpatient, non-antibiotic
management in acute uncomplicated diverticulitis: A
prospective study. Int J Colorect Dis 2015; 30: 1229-34.

34 UERM Health Sciences Journal • VOL. 7 NO.1•JANUARY· JUNE 2018