Systemic Antibiotics After Incision, Drainage of Abscesses
Systemic Antibiotics After Incision, Drainage of Abscesses
We conducted a systematic search of the MEDLINE, Embase, Scopus databases and the Cochrane registry of clinical trials (last accessed 15 Dec 2012) using the terms 'abscess', 'antibiotics', 'clinical trial' and 'outcome'. We performed an additional search using Google Scholar and a list of bibliographies. We also searched trial registries and abstracts of major emergency medicine (EM) congresses. We included all studies that were placebo-controlled, randomised trials that compared any oral antibiotic (intervention) with a placebo (comparator) in ED patients after incision and drainage of their abscesses (patients) and provided outcome data on clinical cure (outcomes). We excluded animal trials and crossover trials. Two investigators independently screened studies by title and abstract to evaluate whether the trial fitted the inclusion criteria.
Two authors extracted data independently on a predefined data extraction form. A two-by-two summary table was completed for each outcome and for each trial, summarising the number of patients who experienced the event or outcome for each group. These data were double-checked. If there was any disagreement, then the source data were evaluated jointly. The primary outcome was clinical cure at 7 or 10 days, and a secondary outcome was abscess recurrence at 30 or 90 days.
For the randomised trials, two authors independently collected information from all studies to assess the risk of bias using the Cochrane risk of bias tool. We collected information on random sequence allocation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias), and any other obvious bias.
We assigned 'low risk', 'high risk' or 'unable to determine' to each item of the Cochrane risk of bias tool. Data are reported in accordance with the PRISMA guidelines for systematic reviews. Since this was a meta-analysis of published articles, it was exempt from institutional review board (IRB) review.
For each randomised trial and the aggregate analysis, we used the OR as the summary measure of association. We evaluated statistical heterogeneity among studies using Cochrane Q and I statistics. We evaluated publication bias by means of visual inspection of the funnel plot. Data were analysed using REVMAN5 using weighted analysis where the weights were the inverse of the variance of the effects estimates. All reported p values are two-sided, and p < 0.05 was considered to be statistically significant.
Methods
Data Selection and Extraction
We conducted a systematic search of the MEDLINE, Embase, Scopus databases and the Cochrane registry of clinical trials (last accessed 15 Dec 2012) using the terms 'abscess', 'antibiotics', 'clinical trial' and 'outcome'. We performed an additional search using Google Scholar and a list of bibliographies. We also searched trial registries and abstracts of major emergency medicine (EM) congresses. We included all studies that were placebo-controlled, randomised trials that compared any oral antibiotic (intervention) with a placebo (comparator) in ED patients after incision and drainage of their abscesses (patients) and provided outcome data on clinical cure (outcomes). We excluded animal trials and crossover trials. Two investigators independently screened studies by title and abstract to evaluate whether the trial fitted the inclusion criteria.
Two authors extracted data independently on a predefined data extraction form. A two-by-two summary table was completed for each outcome and for each trial, summarising the number of patients who experienced the event or outcome for each group. These data were double-checked. If there was any disagreement, then the source data were evaluated jointly. The primary outcome was clinical cure at 7 or 10 days, and a secondary outcome was abscess recurrence at 30 or 90 days.
For the randomised trials, two authors independently collected information from all studies to assess the risk of bias using the Cochrane risk of bias tool. We collected information on random sequence allocation (selection bias), allocation concealment (selection bias), blinding of participants and personnel (performance bias), blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), selective reporting (reporting bias), and any other obvious bias.
We assigned 'low risk', 'high risk' or 'unable to determine' to each item of the Cochrane risk of bias tool. Data are reported in accordance with the PRISMA guidelines for systematic reviews. Since this was a meta-analysis of published articles, it was exempt from institutional review board (IRB) review.
Statistical Analysis
For each randomised trial and the aggregate analysis, we used the OR as the summary measure of association. We evaluated statistical heterogeneity among studies using Cochrane Q and I statistics. We evaluated publication bias by means of visual inspection of the funnel plot. Data were analysed using REVMAN5 using weighted analysis where the weights were the inverse of the variance of the effects estimates. All reported p values are two-sided, and p < 0.05 was considered to be statistically significant.
Source...