Mavros, 2019¹³

US

Includes multi-centre RCTs published between 1996 and 2009

Literature searched up to March 2018

Studies that compared regimens that are not recommended for the treatment of patients with cIAI based on the 2017 Infection Society Revised Guidelines on the Management of Intra-Abdominal Infection were excluded

N = 4125

Adult (≥18 yers) patients with cIAI, as defined in individual studies

Other population characteristics not provided by review author

Intervention:

Fluoroquinolone-based regimen (initiated as IV; could be switched to oral when patient in stable condition)

Comparator:

Beta lactam-based regimen (initiated as IV; could be switched to oral when patient in stable condition)

Outcomes

Clinical Effectiveness:

Treatment success

Defined as sustained resolution or improvement of clinical symptoms and signs attributed to the cIAI, with the absence of surgical site infection requiring systemic antimicrobial treatment at time of test of cure visit

AEs:

all-cause mortality; Incidence of treatment-related AEs; severe AEs; withdrawal due to AEs

as defined in each study

Measurement tools and properties not reported

Follow-up

Time to test of cure visit not reported

De Waele, 2014¹⁴

Belgium

Data were from pooled studies collected in US, Asia, and Europe (specific countries not identified)

Four non-inferiority RCTs were pooled; 3 double-blind, 1 open-label; non-inferiority margin of 10% in 3 studies and 15% in 1 study.

Studies conducted between 2000 and 2009

N = 1,229

Eligibility:

Age ≥18 years

Eligibility for the pooled analysis:

-cIAI (i.e., secondary peritonitis) requiring surgery and supportive management;

-no other systemic antimicrobial agent administered concomitantly with the study drug unless treatment failed;

-documented compliance with ≥80% of the study medication given;

-no protocol violations influencing treatment efficacy;

-successful completion of an assessment at the test-of-cure visit

Gender: female, 422 (34.3%); male, 807 (65.7%)

Mean age: 47.6 years

Mean APACHE II score (severity of illness) (SD): 7.0 (5.0)

Intervention:

IV moxifloxacin 400 mg once daily for up to 14 days in 4 studies; option to switch to per os moxifloxacin 400 mg after a ≥3 days in 2 studies

Comparators:

i)

Piperacillin-tazobactam, 3.0g/0.375g IV, 4 times/day followed by amoxicillin/clavulanic acid, 800 mg/114mg orally, 2 times/day

ii)

Ceftriaxone, 2.0, 4 times/day + metronidazole, 500mg IV 3 times/day, followed by amoxicillin clavulanic acid, 500mg/125mg orally 3 times/day

iii)

Ceftriaxone, 2.0g daily + metronidazole, 500mg IV, 2 times/day

iv)

Ertapenem, 1.0g IV, 4 times/day

Outcomes

Clinical effectiveness outcome:

success rates at test-of-cure, n (%); defined as “continued resolution or improvement of clinical signs and symptoms related to the infection not requiring any antibiotic therapy, and without occurrence of any surgical infection requiring a systemic antibiotic treatment at test-of-cure.” (p569)

AEs

All AEs and subcategories: Nausea, diarrhea, abdominal pain, constipation, surgical site infection, post-operative surgical site infection, drug-related AEs occurring in >5 patients in either group, serious AEs, drug-related serious AEs, premature discontinuations due to AEs, deaths

Defined according to MedDRA Preferred Terms

Follow-up:

Test-of-cure conducted between day 10 and 45 after antibiotic therapy

Wirth, 2018¹⁵

MOXIPEDIA study

Germany

Multi-centre, double-blind, double-dummy, active-controlled parallel group, phase 3 RCT

Participants recruited between January 2010 and 2015

N = 451

Median (range) = 13 years (3 months to 17 years); All patients had undergone an initial surgical or interventional radiology procedure with or without postoperative drainage of abdominal cavity at baseline

Eligibility:

Pediatric patients with cIAI (includes intra-abdominal abscess or macroscopic intestinal perforation with peritonitis)

Intervention:

IV moxifloxacin

Dose: age and body weight-scaled infusions corresponding to the 400 mg adult dose were administered over 60 minutes; For those aged 3 months to <18 years with BW <45kg, doses ranged from 4mg/kg to 6 mg/kg BW twice daily for those with BW <45kg not exceeding 400mg daily; For those aged ≥12 years with BW ≥45kg dose = 400 mg once daily

Duration of treatment: Mean (range) = 8.7 (1–24) days

Switching to oral dose was an option for those aged ≥2 yers with body weight ≥20kg, doses were 4mg/kg as 50mg tablets twice daily not exceeding 400mg or 400mg once daily

Comparator:

IV ertapenem followed by oral amoxicillin/clavulanate

Dose: as per dosing instructions on the label (not further described by Wirth et al.)

Duration of treatment: Mean (range) = 8.7 (1 to 14) days

Outcomes

Clinical Effectiveness:

Clinical and bacteriologic efficacy at test of cure

AEs:

Cardiac AEs

Measured by standard 12-lead ECG; recordings transmitted to a specified ECG core laboratory for semiautomated and manual verified analyses of ECG parameters; verified by physician expert

Musculoskeletal AEs

Standardized assessment of shoulder, elbow, wrist, hip, knee, ankle, Achilles tendon and patellar tendon at both sides; Parent-report medical history questionnaire

AEs defined according to MedDRA version 17.1

Follow-up

Assessed at baseline, day 1, day 3–5, day of switch from IV to oral therapy if applicable, end of treatment and test of cure (28 to 42 days after end of treatment);

Patients with musculoskeletal AEs at 1 year after end of treatment were followed up until resolution or up to 5 years

Xu, 2017¹⁶

China

Open-label, parallel-arm, single-centre RCT

Conducted between 2012 and 2016

N = 80

Eligibility:

Age >18 years; continuous ambulatory peritoneal dialysis patients; Diagnosis of acute peritonitis according to the ISPD guideline

Intervention:

Intraperitoneal vancomycin, 1g every 5 days + oral moxifloxacin, 400mg daily

(Moxifloxacin taken 2 hours before or 4 hours after aluminum or magnesium containing antacids or iron supplements)

Comparator:

Intraperitoneal vancomycin, 1g every 5 days + intraperitoneal ceftazidime, 1g daily

Outcomes

Complete cure

Defined as complete resolution of peritonitis with antibiotics alone without relapse within 4 weeks of completion of treatment

Primary treatment failure

Defined as presence of fever, abdominal pain, or turbid peritoneal dialysate, plus total peritoneal white blood cell count > 50% of pretreatment value after 3 days of treatment

Secondary treatment failure

Defined as “treatment failure despite adjustment of antibiotics or changing to second-line antibiotics for at least 3 days in patients with primary treatment failure.” (p32)

Relapse, recurrent, or repeat peritonitis

episodes Defined according to ISPD guidelines

Peritonitis-related death

Defined as death of patient with active peritonitis, admitted with peritonitis, or within 2 weeks of peritonitis episode

Follow-up

1, 3, 5, 7, 10, 14, and 21 days following initiation of treatment

SIS (Mazuski),¹⁷ 2017

US

Task force comprised of members of the Therapeutics and Guidelines Committee and additional individuals from SIS with subject matter expertise

Intended users: clinicians

Subjects were selected from the 2010 guideline for updating

Task force chair and working group leaders were appointed by SIS Executive Council. Working groups consisted of additional content experts and Therapeutics and Guidelines Committee

The task force chose subjects to update from the previous guideline, developed specific questions, and reviewed RCT evidence published between 2000 and 2016 identified by systematic search of Medline. Additional study designs were considered (not systematic) where gaps existed

“Because of the limited quantity of methodologically rigorous studies investigating key questions in the management of IAI, the task force did not undertake a detailed statistical analysis for most of these recommendations, but relied on a process of iterative consensus among task force members to develop the recommendations and their final grading.” (p9)

Recommendations were drafted and reviewed by working groups over teleconferences, a face-to-face meeting, and email until consensus was achieved

Recommendations were graded according to GRADE

Guideline document was reviewed by experts from the SIS, and modified according to reviews by consensus of the task force; it is expected the guideline was externally peer reviewed based on its publication in a scientific journal

Task force responded to critiques and forwarded final guideline and supporting materials to SIS Executive Council for final approval.

EKMUD (Avkan-Oguz),⁵ 2016

Turkey

The consensus report development group was led by EKMUD, and comprised of 15 experts in IAI

Intended users: physicians in Turkey involved in diagnosis and management of IAIs

Committee met to discuss issues and common practices, delegate subareas among themselves, and screen relevant literature.

Literature searches of databases were conducted (unclear inclusion criteria).

A supplemental questionnaire was circulated to assess surgeons’ attitude and ability regarding microbiologic sampling in the diagnosis of intra-abdominal infections

Recommendations from existing guidelines were discussed (unclear how identified)

“all guidelines do not use the same evidence strength table. Therefore, only the evidence grades which were consistent, recommended in both of these guidelines and were also considered appropriate by our expert panel were included in this consensus report with references to the relevant guidelines.” (p307)

Recommendations were formulated over 4 meetings, using data from Turkey (unclear what type, as no RCTs were identified and no other data or inclusion criteria were discussed) and referring predominately to 2 guidelines, as well as 7 other guidelines

Strength of recommendations and quality of evidence was reported based on an unspecified grading system

GRADE system

1. Strong recommendation:

“The task force concluded that the intervention is a desirable approach for the care of those patients to whom the question applies. This rating is generally based on moderate to high quality evidence. The conclusion is unlikely to be changed with future research.

The magnitude of the effect is also sufficient to justify the recommendation. A strong recommendation was also used to describe interventions that are likely to have a significant effect on patient outcome, even if based on weak evidence. These recommendations are prefaced as “We recommend.”.”

2. Weak recommendation:

“The task force concluded that the intervention is a reasonable approach for the care of patients. Not all patients and clinicians, however, would necessarily want to follow the recommendation. A decision not to follow the recommendation is unlikely to result in a major adverse outcome. This rating was generally based on weak to moderate quality evidence. Both the magnitude of the treatment effect and its direction might be altered by future research. These recommendations are prefaced as “We suggest.”.

3. No recommendation:

“The evidence was considered inadequate or too inconsistent to allow any meaningful conclusion to be reached.”

(p9)

GRADE system

A. High quality evidence:

“The evidence was primarily obtained from RCTs, meta-analyses of such trials, or methodologically sound epidemiologic studies. If the preponderance of evidence is based on studies that do not directly address the question being posed, the overall grade is downgraded to B or C. If there are conflicts in Class A data, the evidence grade is lowered to B or C, depending on the degree of conflict.”

B. Moderate quality evidence:

“The evidence was obtained from lower quality prospective studies, retrospective case control studies, and large observational, cohort, or prevalence studies, and was based on clearly reliable data. If there are significant conflicts in Class B data, the evidence grade is lowered to C.”

C. Weak quality evidence:

“The evidence was obtained from smaller observational studies, studies relying on retrospective or less reliable data, authoritative opinions expressed in reviews, or expert opinions of task force members.”

D. Insufficient evidence:

“There was little or no relevant evidence to address a question, or the evidence reviewed was highly conflicting.”

(p9)

System of unclear origin

Grade A:

“Good evidence supporting a recommendation for use”

Grade B:

“Moderate evidence supporting a recommendation for use”

Grade C:

“Poor evidence supporting a recommendation for use”

(p307)

System of unclear origin

Level I

“Evidence from at least one well-designed randomized controlled trial”

Level II

“Evidence from at least one non-randomized clinical trial”

Level III

“Evidence from opinions of respected authorities, based on clinical studies, descriptive studies, or reports from expert committees”

(p307)

• Research questions and inclusion criteria for the review included the components of PICO
• Review contained an indirect statement suggesting that the review methods were established prior to the conduct of the review – the reason for not registering the methods was a technical glitch
• Review authors used a comprehensive literature search strategy
• Study selection and extraction were performed in duplicate
• Reasons for exclusion were reported in PRISMA flow diagram
• Interventions, comparators, outcomes, and research designs were described in adequate detail
• Appropriate methods for statistical combination of results in meta-analysis were performed
• Review authors did not account for RoB in individual studies when interpreting the results of the review, however predominately high quality RCTs were included
• Adequate investigation of publication bias occurred
• RoB was assessed using modified Jadad scale and Cochrane risk of bias tool (reported in online supplement only)

• Review authors did not explain their selection of the study designs for inclusion in the review
• A list of excluded studies was not provided
• Population characteristics were described in limited detail
• Review authors did not report on the sources of funding for the studies included in the review
• Review authors did not assess the potential impact of RoB in individual studies on the results of the meta-analyses
• Review authors did not discuss the heterogeneity observed in the results of the review
• Review authors indicated potential conflicts of interest but did not indicate how conflicts of interest were managed

• Research questions were clearly described
• Included studies were described in adequate detail
• Statistical methods used to combine studies were reported
• Authors assessed statistical heterogeneity using exploratory methods
• Pooled studies were homogenous
• Authors disclosed potential conflicts of interest, but did not report how conflicts were managed

• Inclusion criteria were not clear. Outcomes were those commonly assessed by the selected RCTs
• Review methods and study selection process were not described
• Authors did not explain selection of the study designs for inclusion in the review, and it was not likely to have been systematic
• Review authors did not use a comprehensive literature search strategy
• It is unclear how the studies were identified or selected for inclusion
• RoB in individual studies was not assessed by study authors
• Sources of funding for included studies was not reported
• Authors used appropriate methods for statistical combination of results
• Review authors did not assess the potential impact of RoB in individual studies on the results of the meta-analysis
• Authors did not investigate publication bias. It is presumed to present a high RoB given the apparent absence of a systematic search
• Authors disclosed potential conflicts of interest, but did not report how conflicts were managed

• Allocation sequence was random
• No apparent baseline differences between groups

• Participants, carers, and people delivering the interventions were not aware of the assigned interventions during the trial
• Appropriate analyses were used to estimate the effect of assignment to intervention

• Data were available for nearly all participants randomized

• Method of measuring outcomes was appropriate
• Measurement of outcomes unlikely to have differed between intervention groups
• Outcome assessors were blinded

• Blinded outcome data were un-blinded when data-collection was complete
• The numerical results were not likely to have been selected on the basis of results from multiple outcome measurements or multiple analyses of the data

• Authors did not report allocation concealment

• Unclear if an analysis plan was pre-specified

• 26 patients in the intervention group vs 4 in the comparator group did not complete study medication. Reasons in the intervention group were largely due to the outcome of interest, i.e., 15 experienced an AE, parents of 2 patients withdrew consent, 2 protocol violations and 5 for other reasons; reasons in the comparator group were 2 experienced AE and 2 for other reasons

• Unclear if an analysis plan was pre-specified

• Unclear if an analysis plan was pre-specified

• Study was funded and conducted by Bayer. There was no statement included in the publication indicating that the funder’s interests did not influence the study conduct or results.

• Allocation sequence was random
• No important differences between intervention groups were identified

• Authors reported no deviations from the intended intervention
• An appropriate analysis was used to estimate the effect of assignment to intervention

• Outcome data were available for all patients

• Ascertainment of the outcome is not likely to have differed between groups
• Assessment of the outcomes was not likely to have been influenced by knowledge of the intervention received

• This trial was registered a priori
• The numerical results were not likely to have been assessed on the basis of multiple outcome measurements or multiple analyses of the data

• Allocation was not concealed from patients or researchers

• Participants were aware of their assigned intervention during the trial
• Carers and people delivering the intervention delivering the interventions were aware of participants’ assigned intervention during the trial

• Method of measuring the outcomes was not described in the article.
• Outcome assessors were not blinded to intervention assignment

• The analysis plan was not included in the a priori registered protocol

Fluoroquinolone-based regimens vs beta lactam-based regimens

Clinical effectiveness (ITT)

n = 3055

RR = 0.97, 95% CI, 0.94 to 1.01

AEs

All-cause mortality

N = 3614

RR = 1.04, 95%CI, 0.75 to 1.43

Overall treatment related AEs

n = 2801

RR = 0.97, 95% CI, 0.70 to 1.33

Withdrawal due to AEs

n = 3380

RR = 1.07, 95%CI, 0.86 to 1.33

Note:

(Ps not reported)

Analyses were also conducted by those clinically evaluable and microbiologically evaluable for clinical effectiveness outcome; data not extracted

“Overall, the 2 classes of antimicrobials were equally effective and safe. Limited evidence from subgroup analyses suggested slightly lower clinical effectiveness of moxifloxacin when compared to a beta lactam-based regimen in the overall population (based on 4 trials), as well as in the subset of patients with complicated appendicitis (3 trials).”

(p9)

“In conclusion, our meta-analysis suggests that ciprofloxacin/metronidazole is as effective as ceftriaxone/metronidazole and carbapenems for the treatment of cIAIs, while moxifloxacin monotherapy had slightly inferior outcomes, especially in complicated appendicitis. Empiric antimicrobial treatment of patients with cIAIs should be selected in light of the local bacterial epidemiology and patterns of resistance.” (p12)

Clinical Effectiveness (ITT)

Moxifloxacin n/N (%) vs comparator n/N (%)

449/609 (73.7%) vs 482/620 (77.7%)

Point estimate for the difference in success rates (%): −3.96; 95% CI, −7.06 to −1.05, P = 0.25

AEs (MedDRA Preferred Terms Event) (ITT)

Moxifloxacin, n (%) vs comparator, n (%)

All AEs

410 (67.3%) vs 371 (59.8%)

 Nausea

 49 (8.0%) vs 28 (4.5%)

 Diarrhea

 39 (6.4%) vs 49 (7.9%)

 Abdominal pain

 26 (4.3%) vs 19 (3.1%)

 Constipation

 21 (3.4%) vs 19 (3.1%)

 Surgical site infection

 65 (10.7%) vs 51 (8.2%)

 Post-operative surgical site infection

 10 (1.6%) vs 6 (1.0%)

Drug-related AEs occurring in >5 patients in either group

127 (20.9%) vs 124 (20.0%)

Serious AEs

110 (18.1%) vs 88 (14.2)

Drug-related serious AEs

19 (3.1%) vs 6 (1.0)

Premature discontinuations due to AEs

31 (5.1%) vs 25 (4.0%)

Deaths

26 (4.3%) vs 21 (3.4%)

Note: per protocol analysis were also reported for clinical effectiveness outcome; data not extracted

Note: statistical significance was not assessed for any outcome

Efficacy – Clinical response

(modified ITT i.e., all patients valid for safety with had at least one pretreatment causative organism for the primary site of infection or from blood cultures).

(n = 246 vs n = 133)

Moxifloxacin n (%) vs ertapenem and amoxicillin/clavulanate n (%)

Bacteriologic success and clinical cure at time of cure

208 (84.6%) vs 127 (95.5%)

Note: analysis for safety population not reported by study author. Author reported “Similar results were found in the safety population.” (pe210)

Any clinical failure

38 (15.4%) vs 6 (4.5%)

Infectious failure

20 (8.1%) vs 2 (1.5%)

Reoperation

1 (0.4%) vs 1 (0.7%)

Wound infection

7 (2.8%) vs 1 (0.7%)

Other (includes QT interval prolongation and administration of other antimicrobials)

1/246 (0.4%) vs 0 (0%)

Incidence rate of AEs (ITT)

(n = 301 vs n = 150)

Moxifloxacin n (%; 95%CI) vs comparator n (%; 95%CI)

Any AE

175 (58.1%; 95% CI, 52.6% to 63.7%) vs 82 (54.7%; 95% CI, 46.7% to 62.6%)

Any drug-related AE

43 (14.3%; 95% CI, 10.3% to 18.2%) vs 10 (6.7%; 95% CI, 2.7% to 10.7%

Any serious AE

20 (6.6%) vs 6 (4.0%)

Severe AE

12 (4%) vs 3 (2.0%)

Any drug-related serious AE

0 (0%) vs 0 (0%)

Death

0 (0%) vs 0 (0%)

“No relevant differences in incidence rates of the most common AEs were observed between the 2 treatment arms, except for ECG QT prolongation” (pe209)

ECG QT/QTc AEs:

QT prolongation in ECG recordings

28 (9.3%) vs 4 (2.7%) (95%CI not reported)

Drug related QT prolongation in ECG recordings

7%; 95% CI, 4.4 to 10.3 vs 1.3%; 95% CI, 0.2 to 4.7 (n not reported)

Note: one event in intervention group assessed as severe AE by investigator

QTc interval prolongation-related morbidity or mortality

none

“Ten patients (3.3%) discontinued MXF early because of uncorrected QT interval prolongation. A case of QTcB (+24 ms) and QTcF (+13 ms) prolongation after 400 mg MXF infusion at Day 1 resolved on the same day after MXF withdrawal.” (pe210)

Musculoskeletal AEs:

Musculoskeletal AE

13 (4.3%; 95% CI, 2.3 to 7.3) vs 5 (3.3%; 95% CI, 1.1 to 7.6)

Forearm fracture

1 (0.3%) vs 0 (95% CIs not reported)

Joint injury

0 vs 1 (0.7%) (95% CIs not reported)

Ligament sprain

1 (0.3%) vs 1 (0.7%) (95% CIs not reported)

Muscle strain

0 vs 1 (0.7%) (95% CIs not reported)

Arthralgia

9 (3.0%) vs 1 (1.3%) [unclear if the reporting error is in absolute number or percentage for comparator group] (95% CIs not reported)

Joint swelling

0 vs 1 (0.7%) (95% CIs not reported)

Musculoskeletal pain

3 (1.0%) vs 0 (95% CIs not reported)

Myalgia

1 (0.3%) vs 0 (95% CIs not reported)

Note: no musculoskeletal events were reported by the investigators as being related to intervention or comparator treatments

Clinical effectiveness (n = 40 vs n = 40)

Intervention n (%) vs comparator n (%); OR, 95% CI

Complete cure

31 (78%) vs 32 (80%);

OR = 0.86, 95% CI, 0.30 to 2.52; P = 0.8

Primary treatment failure

13 (33%) vs 8 (20%);

OR = 1.93, 95% CI, 0.70 to 5.34; P = 0.2

Secondary treatment failure

4 (10%) vs 5 (13%);

OR = 0.78, 95% CI, 0.78, 0.19 to 3.14; P = 0.7

Peritonitis-related death

1 (3%) vs 1 (3%);

OR = 1.00, 95% CI, 0.06 to 16.56; P = 0.9

Successive episodes of peritonitis during 3-month follow-up

7 (18%) vs 3 (8%);

OR = 2.62, 95% CI, 0.63 to 10.95; P = 0.2

Outcome of successive episodes of peritonitis

OR = 1.40, 95% CI, 0.88 to 2.24; P = 0.3

-Transfer to hemodialysis

2 (29%) vs 0 (0%)

-Maintenance of PD

5 (71%) vs 3 (100%)

4.7a.

Strong recommendation; high quality evidence

4.7b.

Strong recommendation; high quality evidence

4.7c.

Weak recommendation, weak quality evidence

6.1.

Strong recommendation; high quality evidence

6.3.

Strong recommendation; moderate quality evidence

6.5.

Weak recommendation; moderate quality evidence

Managing community-acquired intra-abdominal infections of mild to moderate severity in adults

20. For adult patients with mild to moderate IAI in Turkey, the use of ertapenem, moxifloxacin or tigecycline monotherapy or combinations of metronidazole with cefazolin, cefuroxime, ceftriaxone, cefotaxime, levofloxacin or ciprofloxacin are preferable to regimens with anti-pseudomonal activity (IDSA, A-I). Because data are scarce concerning the ESBL positivity rate of organisms involved in community-acquired IAIs in Turkey, empiric antibiotics should be selected on an individual hospital basis according to the ESBL rate in local epidemiologic data (10% or higher) or the ESBL risk factors. Furthermore, although Pseudomonas species are isolated in approximately 8% of IAIs, they are less likely to be the causative agent.” (p313–314)

Managing high-risk community-acquired intra-abdominal infection in adults

26. “The empiric use of antimicrobial regimens with broad-spectrum activity against gram-negative organisms is recommended for the treatment of patients with high-risk IAIs as defined by an APACHE II score >15 or the presence of other factors. These agents include meropenem, imipenem-cilastatin, piperacillin-tazobactam, and ceftazidime as single agents or combinations of metronidazole with cefepime, ciprofloxacin or levofloxacin (IDSA, A-I).” (p315)

Monitoring antimicrobial therapy in patients with intra-abdominal infection

Clinical monitoring: “During the recovery period, amoxicillin-clavulanic acid and oral quinolone (moxifloxacin, ciprofloxacin, levofloxacin) or oral cefalosporins in combination with metronidazole can be used for patients able to accept oral medications and whose infections are due to bacteria susceptible to such agents (IDSA, B-II).” (p318)