Acute Cellulitis: Meta-Analysis Does Not Reveal Optimal Antibiotic Regimen
A systematic review and meta-analysis was undertaken to ascertain optimal route and duration of antibiotic treatment of acute cellulitis
Studies that might have informed appropriate antibiotic regimens for acute cellulitis were likely to be flawed by biased comparisons, inadequate follow-up periods, and/or lack of adverse event documentation, according to the conclusions from a recently published meta-analysis.
Frequent disparities between practice and guideline recommendations for antibiotic treatment of acute cellulitis that result in excessive antibiotic use and questionable choice of intravenous (IV) route of administration prompted the investigators to review the literature and conduct a meta-analysis of pertinent studies to determine an optimal route and duration of antibiotic treatment.
"Decisions about whether to initiate IV or oral treatment should be based on severity of illness but 30-50% of patients eligible for oral therapy receive IV antibiotics, and then many remain on them for longer than necessary," Martin Llewelyn, PhD, Department of Global Health and Infection, Brighton and Sussex Medical School, University of Sussex, Falmer, East Sussex, UK, and colleagues observed.
"International recommendations for treatment duration in cellulitis are inconsistent and range from 5 to 10-14 days, but in practice durations commonly exceed 2 weeks," they noted.
Llewelyn and colleagues identified 47 interventional and observational studies of antibiotic treatment for cellulitis from over 8,000 articles screened, incorporating data from 11 trials comprising a total of 1,855 patients, in the period from the inception of the MEDLINE, EMBASE and trial registries to December 11, 2019. Randomized controlled trials were restricted to comparisons of shorter versus longer durations, or oral versus IV routes of administration. Studies were excluded if they investigated antibiotic prophylaxis or solely topical route of administration; and case series and case reports were not included.
Clinical response was the primary outcome measure, with secondary measures of treatment effectiveness including recurrence, microbiological cure/bacteriological response, mortality, biochemical response, and patient-focused outcomes. In addition, the investigators considered balancing outcomes with such factors as length of hospitalization, adverse events, and development of antibiotic resistance.
Finding few studies that adequately addressed the question of optimal regimens, Llewelyn and colleagues found no evidence of difference in clinical response rates between IV and oral treatment groups, or between shorter and longer durations of treatment. They also considered the timing of switching from IV to oral, but here too encountered insufficient studies to ascertain optimal timing or to identify markers for changing the route of administration.
In addition, while patient factors emerged as the key determinants of clinical response across studies, there was no evidence of associations between the duration of therapy and outcome. They also noted that few studies assessed treatment-associated harms, and none reported on the development of antibiotic resistance.
"Currently, the available evidence suggests that for adults with cellulitis, clinical response rates are similar for initial oral and IV antibiotic treatment," the investigators report. "While this is the consistent finding of 5 trials, all are of low or very low quality," they added.
Of the 10 randomized trials that provided evidence on the duration of therapy, 8 of the trials used short-course treatment with agents anticipated to have longer durations of action compared to the longer-course agent used. The investigators note that the common guidelines recommending 5 days treatment duration are based on a single trial with levofloxacin, which is not a recommended first-line treatment.
"The comparisons are therefore severely flawed and their findings of limited relevance to clinical decision making," Llewelyn and colleagues concluded.