Outpatient Treatment for Mild Acute Diverticulitis: Antibiotics vs. No Antibiotics

The incidence of diverticular disease is increasing: around 30% of the population over 45 years of age, and 60% of those over 85 years of age have diverticula. Approximately 10-25% of such patients will suffer an episode of acute diverticulitis (AD) at some point during their lifetime [1].

AD is one of the most prevalent reasons for consultation in emergency departments (ED) in Western countries [2]. However, 75% of these episodes do not present complications and the majority will have a good outcome with conservative treatment. [3].

Since its initial description, the treatment of AD has not had a solid scientific basis; has consisted of hospital admission and antibiotics, assuming an infectious etiology [3]. In recent years, 2 randomized controlled trials have shown no benefit of antibiotics in the treatment of uncomplicated AD in hospitalized patients [4,5].

These studies found no significant differences in relation to recovery time, complications, recurrences, and need for surgery between the groups with or without antibiotic treatment, and also suggested that antibiotics could be omitted in patients with a first episode of AD. uncomplicated.

Similarly, many systematic reviews and meta-analyses support non-antibiotic treatment of uncomplicated AD [6,7], and – in fact – that approach is included in the Guidelines of the American Society of Colon and Rectal Surgeons [8].

Outpatient treatment of uncomplicated AD has been shown to be safe and effective [9]. No differences have been reported in relation to treatment failure, and the total cost of medical care per episode is lower in the outpatient group.

The main objective of this study was to establish whether patients treated with and without antibiotics on an outpatient basis could present differences in terms of admission rates. The secondary objectives were to analyze the differences in relation to:

1) Repetition of consultations with the ED (and their reasons); 
2) Pain control at different times; and 
3) Complication rate.

A non-inferiority design was chosen for the study on the assumption that the new outpatient treatment regimen without antibiotics could be non-inferior to the standard treatment with antibiotics and hospital admission.

> Study design

The DINAMO study is a multicenter prospective open-label non-inferiority randomized controlled trial [10], with an intention-to-treat approach and parallel assignment, and with the participation of 15 colorectal surgical units, in secondary and tertiary acute care hospitals in Catalonia (Spain). ). All institutions belong to the Spanish Public Care System.

The trial was conducted in accordance with the Declaration of Helsinki, seventh revision [11], the Standard Protocol Items for Clinical Trials (SPIRIT 2013) [12], and Spanish laws and regulations for biomedical research. Authorization was obtained from the Spanish Agency for Medicines and Health Products (AEMPS).

The trial protocol, patient information, and informed consent documents were approved by the ethics committees of all participating centers, in accordance with Royal Decree 1990/2015 of December 4.

The trial was registered in the ClinicalTrials.gov database (ID: NCT02785549) and in the EU Clinical Trials Register database (EudraCT, no. 2016-001596-75). The study protocol has been previously published by the medical team [10].

> Participants

Inclusion criteria : age between 18 and 80 years (inclusive), AD stage 0 on computed tomography (CT), according to the modified Neff classification, absence of AD episode in the last 3 months, without antibiotic treatment for any reason in the last 2 weeks, without significant comorbidities, immunocompetence, written informed consent, adequate cognitive capacity, adequate family support, good control of symptoms in the ED, and a maximum of 1 of the following: T ≥ 38ºC, T < 36ºC, L > 12,000/µL, L < 4000/µL, HR > 90/min, RR > 20/min/, CRP > 15 mg/dL. (T: temperature. L: leukocytes. HR: heart rate. RR: respiratory rate. CRP: C-reactive protein).

Exclusion criteria : pregnant or breastfeeding woman, age < 18 or > 80 years, allergy to any of the study drugs, AD grade I or higher of the Neff classification, episode of AD in the last 3 months, antibiotic treatment for any reason in the last 2 weeks, presence of significant comorbidities, immunosuppression, lack of written informed consent from the patient, inadequate cognitive ability, inadequate family support, poor symptom control in the ED (visual analogue scale [VAS] ≥ 5), and/or systemic inflammatory response syndrome.

Significant comorbidities were defined as diabetes mellitus with organ involvement (retinopathy, angiopathy, nephropathy), emergency care for a cardiogenic event in the last 3 months (acute myocardial infarction, angina, heart failure), decompensation of chronic liver disease in the last 3 months (Child ≥ B), and end-stage renal disease.

Immunocompetence was defined as the absence, and immunosuppression as the presence, of any of the following: active neoplastic disease, hematological malignancy, human immunodeficiency virus with low CD4+ count (AIDS), prolonged treatment with corticosteroids, immunosuppressive therapy , transplant, splenectomy, and genetic immunodeficiency.

“Adequate cognitive capacity” was defined as the ability to read and understand the study description and sign an informed consent. “Family support” was defined as “adequate,” when the patient had someone capable of caring for him or her and providing the necessary help.

The diagnosis of AD was made by CT and the modified Neff (mNeff) classification was applied [13-15]. All eligible patients who were not included in the study were registered, and the reasons for their non-participation were explained in accordance with the Consolidated Standards for Reporting Trials statement , for randomized trials of non-inferiority and equivalence [16].

> Randomization and masking

Patients who presented to the ED with clinical findings compatible with AD underwent a blood test and an abdominal CT scan. Once the diagnosis of mild AD (grade 0 mNeff) was confirmed, patients who met the inclusion criteria were invited to participate in the study.

By agreeing to participate, they gave their written consent by signing a standardized informed consent document, and were randomized to one of the 2 arms of the study with a 1:1 allocation ratio.

Patients were randomized after successful symptom control in the ED. The study was open and there was no blinding of patients or surgeons. In the design stage, the possibility of conducting a single-blind trial of antibiotics vs. placebo was considered, but this option was rejected due to the multicenter design of the study.

> Procedures

In the experimental arm (group without antibiotics [ATB]), patients were given anti-inflammatory and symptomatic treatment, with 600 mg/8 hours of ibuprofen, alternating with 1 g/8 hours of paracetamol. In the control group (ATB group) the patients were treated with 875/125 mg/8 hours of amoxicillin/clavulanic acid, in addition to the same anti-inflammatory (NSAID) and symptomatic treatment.

Medical treatment was initiated in the ED. The route of administration was intravenous in the ED and oral at discharge. Patients were discharged with medical treatment and dietary recommendations when they achieved good symptomatic control in the ED. If good symptomatic control was not achieved after a maximum of 24 hours of observation in the ED, patients were admitted for intravenous treatment and therefore did not enter the study. The duration of medical treatment (with and without ATB) was 7 days.

> Results

The 2 groups underwent the same clinical controls at 2, 7, 30 and 90 days after the episode, carried out by surgeons from the coloproctological unit. At each control, a general evaluation was carried out through physical examination, clinical evolution was monitored, and adherence to treatment was controlled.

The degree of pain was recorded in each control using a VAS (0-10). If worsening or poor control of symptoms was detected at any time, the patient was referred to the ED. Patients also consulted the ED if, according to their own criteria and based on the information received, they presented any warning signs (temperature > 38ºC, or poor symptomatic control).

In the event of a repeat ED visit, blood tests and CT were repeated. The same monitoring was maintained. The protocol algorithm recommended by the investigators, to select the most appropriate treatment in the event of a repeat consultation, was described in the study protocol [10].

> Sample size

Based on the results of a previous study by the authors, the sample size was calculated taking hospital admission as the main factor [14].

A non-inferiority margin of 7% (D) for both groups (ATB and non-ATB), based on their previous study of outpatient treatment for AD, in which a success rate of 93% was obtained [14] .

A statistical power of 80% was used, and a 1-sided significance level of 0.025. With an estimated patient loss of 10%, it was concluded that a sample size of 230 patients was necessary for each arm of the study.

> Statistical analysis

The primary endpoint was the analysis both by intention to treat and by protocol, given that all randomized patients received treatment. The description of the factors and the statistical analysis was carried out using the Statistical Package for the Social Sciences (SPSS Inc., Chicago, IL) program, version 26.

Quantitative variables were described using mean and standard deviation values, when the distribution was considered normal (Kolmogorov-Smirnof test), otherwise median and interquartile range (IQR) values ​​were used. Categorical variables were described in absolute numbers and percentages.

The univariate statistical analysis of the quantitative variables, with independent groups, was carried out using the Student’s t test , if the conditions for its application were met; otherwise, the  Mann–Whitney U test was applied. For categorical variables, Pearson’s c2 test was used. Results of statistical tests are shown with a 95% confidence interval (CI), where possible.

Statistical significance was established with a P value below 0.05. A 95% CI of the difference was determined for the primary endpoint (5% 1-sided α level). Therefore, a conclusion of non-inferiority was reached if the lower limit of this interval was below the limit of non-inferiority (D=7).

The factor of repeat consultation in the ED was analyzed with the Kaplan-Meier estimation method and the log-rank test.

> Participant flow and recruitment

From November 2016 to January 2020, 849 patients diagnosed with mild AD (grade 0 mNeff) [13] were seen in the EDs of the hospitals participating in the study.

Four hundred and eighty patients with AD met the inclusion criteria and were randomly assigned to the non-ATB group (n = 242) and to the ATB group (n = 238). In the baseline characteristics of the patients, no statistically significant differences were found between the groups, except in the CRP values, but they were not clinically relevant.

> Primary objective: hospital admission

Repeat visits to the ED resulted in 22/480 (4.6%) admissions: 14/238 (5.8%) in the ATB group, and 8/242 (3.3%) in the non-ATB group, with a difference of 2.58% (95% CI: 6.32-1.17), ( P = 0.19). The non-ATB group showed non-inferiority when compared to the ATB group, with a D < 7%.

Most admissions were based on study protocol recommendations for repeat consultation. However, the final decision regarding admission was made by the doctors.

All patients who returned to the ED underwent abdominal CT. In the ATB group, of the 14 patients admitted for hospitalization, 78.6% (11/14) had mild grade 0 mNeff AD, and 21.4% (3/14) had grade Ib. In the group without ATB, of the 8 patients admitted for hospitalization, 87.5% (7/8) had grade 0 mNeff AD, and 12.5% ​​(1/8) had grade Ia.

In the ATB group, the same antibiotic, amoxicillin/clavulanic acid, was maintained in 57.1% (8/14), although in the other 42.8% (6/14), the antibiotic spectrum was expanded. In the non-ATB group, all admitted patients were treated with amoxicillin/clavulanic acid. The median duration of admission was 5 days in the ATB group (IQR: 3 days), and 2.5 days in the non-ATB group (IQR: 3 days) ( P = 0.002), a statistically significant difference. None of the patients in either group required emergency surgery.

> Secondary objectives

• Repetition of the query

During the study period, 447 patients did not return to the ED. However, 40/480 (8.3%) repetitions were recorded, corresponding to 33 patients: 16/238 (6.72%) in the ATB group, and 17/242 (7.02%) in the group without ATB. (mean difference: -0.3; 95% CI: 4.22 to -4.83). 

A total of 29 patients returned for consultation once (14 ATB group and 14 non-ATB group), 2 patients on two occasions (both non-ATB group), 1 patient from the ATB group on 3 occasions, and another from the ATB group on 4 occasions. .

In the ATB group, 21 repetitions corresponding to 16 patients were recorded, and in the group without ATB, 19 repetitions corresponding to 17 patients. In 11/19 (57.9%) there was no worsening in the complementary tests and the patients were able to continue with the same treatment (that is, NSAIDs on an outpatient basis).

No statistically significant differences were found between the groups in any of the patients who repeated the consultation in the ED.

Interestingly, there were more hospital admissions in the ATB group (14/21, 66%) than in the non-ATB group (8/19, 42.1%). A therapeutic change (either starting antibiotic treatment or expanding the spectrum) was needed in 6 of 21 (28.5%) patients in the ATB group, and in 8 of 19 (42.1%) patients in the non-ATB group, although the differences were not statistically significant.

The median time to repeat consultation was 17 days in the ATB group (95% CI: 0 to 36.4), and 13 days in the non-ATB group (95% CI: 4.5 to 25. 5), without statistically significant differences in the log-rank test ( P = 0.82).

• Monitoring, pain control and recovery

At the end of the study period, there were 22/238 (9.2%) lost to follow-up in the ATB group, and 19/242 (7.9%) in the non-ATB group.

Analyzes performed on the remaining patients showed no statistically significant differences between the groups in terms of clinical course, pain control, and physician-recommended ED consultation. At 30 days of follow-up, 212/216 patients in the ATB group (98.1%), and 218/233 in the non-ATB group (97.7%) presented good clinical progress, and were able to eat a normal diet.

Patients in the ATB group showed a higher degree of pain in the control on the 2nd day, 13/230 (5.7%) vs 5/221 (2.3%) in the group without ATB (mean difference 3.38 ; 95% CI: 6.96 to -0.18). Patients in the non-ATB group recorded higher pain scores at subsequent follow-ups, but the difference was not statistically significant.

• Complications

No patients in either group required emergency surgery during the study period.

Diverticular disease is highly prevalent, especially in the Western world, and 15% to 20% of the population with diverticulosis will present with complicated AD [1]. Three-quarters of AD cases are mild [2].

The study that is the gold standard for AD, in addition to the medical history and physical examination, is CT [3], which allows differential diagnosis and optimal classification. In the present study, the Neff classification, modified by the incorporation of substage Ia, was used to characterize AD with localized pneumoperitoneum [13]. This classification made it possible to differentiate between the initial stages of AD and establish an appropriate treatment for each stage.

“ Hospital admission” was considered the main factor, because it allowed us to evaluate the safety of outpatient care, and the probability of treatment failure in mild AD.

The delta margin was based on the results of a previous study by the authors, who achieved a success rate of 93% with the outpatient protocol, for the treatment of uncomplicated AD [14]. The doses and duration of treatment in the 2 arms (ATB group: antibiotic, and non-ATB group: anti-inflammatory), were prescribed using the outpatient treatment protocols in the ATB group applied in previous studies [14,15].

Two randomized controlled trials have already reported non-antibiotic treatment for uncomplicated AD [4,5], but none were conducted on an outpatient basis. With follow-up periods of 12 months, Daniels et al. [4], and Chabok et al. [5] demonstrated that antibiotic-free treatment does not worsen complications, recurrence of cases, or delay in complete recovery. As a result, in the design of the present study, a follow-up of more than 90 days was considered unnecessary to estimate recovery time.

The results of the present outpatient treatment regimen show that it is an effective and safe option. Sixteen of 238 patients (6.72%) in the ATB group returned to the ED, compared with 17/242 (7.02%) in the non-ATB group; and 14 of 238 (5.8%) in the ATB group required hospitalization, compared with 8 of 242 (3.3%) in the non-ATB group; The differences were not statistically significant, and the 95% CIs were far from the non-inferiority margin. These results are similar to those recorded in the DIVER study [10].

In that study, the primary endpoint was treatment failure, which was recorded in 4 of 66 patients (6.1%) in the admission group, and in 3 of 66 (4.5%) in the outpatient group ( P = 0.619). The overall number of admissions in the present study was 22/480 (4.6%), similar to rates in the DIVER study [8] (6.1%), and in the authors’ previous studies (64/480). 68.6%) [13].

The 40 repeat visits to the ED (a rate of 8.3%) were evenly distributed between the 2 groups: 21 (8.8%) in the ATB group, and 19 (7.8%) in the non-ATB group. . In all cases, analytical tests and control CT scans were performed to rule out worsening. Poor symptom control was recorded in 12/21 (57.1%) in the ATB group, and in 11/19 (57.9%) in the non-ATB group.

In the group without ATB, the same antibiotic was maintained in all cases and was administered in-hospital on 5 occasions; All those cases in the group without ATB were discharged again with anti-inflammatory treatment. Analytical or radiological worsening was detected in 9/21 (42.9%) in the ATB group, and in 8/19 (42.1%) in the group without ATB.

In the ATB group, all patients were admitted: in 3 cases the same treatment was maintained, and in the other 6 it was changed to a more powerful antibiotic.

In the group without ATB, all 8 cases were admitted and antibiotic treatment was started. Only 3 patients in the ATB group presented radiological deterioration to grade Ib of the mNeff classification, and 1 patient in the group without ATB progressed to grade Ia mNeff on the control CT, at the time of consultation.

None of the patients who consulted required emergency surgery or any measure other than admission or change of antibiotic. Therefore, the authors of this work can say that the action protocol described in the DINAMO study is safe and does not represent an increased risk.

None of the outpatient follow-up controls reached the non-inferiority limit for patient referral. Furthermore, there were no differences between the groups in terms of pain control up to 90 days, nor in relation to clinical evolution.

The main limitation of this study is the significant number of patients excluded, due to the application of strict selection criteria. This high proportion was due to the fact that patients had to be able to be selected for outpatient treatment and, therefore, to ensure high levels of safety, restrictive criteria had to be applied.

The DIABOLO study [4] also excluded a high number of patients: 323 of 893 possible candidates (36.2%). Furthermore, since doctors in the trial were involved in decisions related to hospital admission (the primary outcome factor), there may also have been some observer/selection bias. Another limitation has been the lack of use of a placebo. However, the high complexity of its control, in a multicenter study, led the authors to reject this procedure.

It is also true that some guidelines, such as those of the American Society of Colon and Rectum Surgeons [8], have already accepted the non-use of antibiotics in the cases of healthy patients with uncomplicated diverticulitis.

Some other studies [16] also performed outpatient treatment without antibiotics for AD. However, the research described in the present work is the first prospective randomized multicenter study carried out to try to demonstrate the non-inferiority of non-antibiotic treatment of mild diverticulitis.

The authors believe that their results can be extrapolated to populations of any type, and that episodes of uncomplicated AD can be treated on an outpatient basis and without antibiotics, as long as well-defined clinical and radiological criteria are applied.