Antibiotics for uncomplicated diverticulitis

Diverticulitis is a condition with inflammation of big bowel herniations termed diverticulae. Diverticulae are common in the elderly above age 70 and usually do not cause symptoms. However, in some cases inflammation cause a condition, diverticulitis, with pain in the abdomen and signs of infection like fever. Diverticulitis causes no complications in most cases, however, some develop complications and need surgery. The uncomplicated diverticulitis is the focus of this review. It has traditionally been viewed as an infection with bacterial overgrowth in the big bowel and has therefore been treated with antibiotics. We aimed to investigate if there existed any clinical research, evidence, on the effects of antibiotics for uncomplicated diverticulitis in this review.

We found 3 randomized controlled trials (RCTs) on the use of antibiotics for uncomplicated diverticulitis tested on hospitalised patients. The newest trial investigating the actual need for antibiotics when compared to no antibiotics, a second investigated two different antibiotic cures and a third investigated the length of IV antibiotic treatment. None of the studies found a statistical difference in the tested antibiotic regimes. The newest trial had the overall best quality and had the biggest groups of patients making it the overall best trial. It found no difference in the occurrence of surgery needing complications like abscesses and perforations of the big bowel.

Antibiotics can cause serious adverse events for patients like allergic reactions and can even cause other life threatening infections of the bowel. Ultimately, there is a growing antibiotic resistance meaning that the drugs loose their ability to function as bactericidals. This causes limitations in the clinical use of antibiotics when they are needed for treating patients with infections. Therefore there exists strong arguments for limiting the use of antibiotics. The trial that showed no effect of antibiotics is very new and needs confirmation from other similar trials. Ongoing trials will in the next few years be published on the subject.

Authors' conclusions: 

The newest evidence from one RCT says there is no significant difference between antibiotics versus no antibiotics in the treatment of uncomplicated diverticulitis. Previous RCTs have only suggested a non-inferiority between different antibiotic regimes and treatment lengths. This new evidence needs confirmation from more RCTs before it can be implicated safely in clinical guidelines. Ongoing RCTs will be published in the years to come and more are needed. The role of antibiotics in the treatment of complicated diverticulitis has not been investigated yet.

Read the full abstract...

Diverticulitis is an inflammatory complication to the very common condition diverticulosis. Uncomplicated diverticulitis has traditionally been treated with antibiotics with reference to the microbiology, extrapolation from trials on complicated intra-abdominal infections and clinical experience.


To assess the effects of antibiotic interventions for uncomplicated diverticulitis on relevant outcome.

Search strategy: 

Studies were identified by computerised searches of the The Cochrane Library (CENTRAL), MEDLINE and EMBASE. Ongoing trials were identified and reference lists of identified trials and relevant review articles were screened for additional studies.

Selection criteria: 

RCTs including all types of patients with a radiological confirmed diagnosis of left-sided uncomplicated diverticulitis. Interventions of antibiotics compared to any other antibiotic treatment (different regime, route of administration, dosage or duration of treatment), placebo or no antibiotics. Outcome measures were complications, emergency surgery, recurrence, late complications and duration of hospital stay and recovery of signs of infection.

Data collection and analysis: 

Two authors performed the searches, identification of RCTs, trial assessment and data extraction. Disagreements were resolved by discussion or involvement of a third part. Authors of trials were contacted to obtain additional data if needed or were contacted for preliminary results of ongoing trials. Effect estimates were extracted as relative risks (RR).

Main results: 

Three RCTs were identified. A qualitative approach with no meta analysis was performed because of variety in interventions between included studies. Interventions compared were antibiotics to no antibiotics, single to double compound antibiotic therapy and short to long IV administration. None of the studies found significant difference between the tested interventions. Risk of bias varied from low to high. The newest RCT overall had the best quality and statistical power.