Antibiotics for treating leptospirosis

Abstract

Background: Leptospirosis is a parasitic disease transmitted by animals. Severe leptospirosis may result in hospitalisation and about five per cent of the patients die. In clinical practice, penicillin is widely used for treating leptospirosis.

Objectives: To evaluate the effectiveness and safety of antibiotics versus placebo or other antibiotic regimens in treating leptospirosis. We addressed the following clinical questions: a) Are treatment regimens with antibiotics more efficient than placebo for leptospirosis? b) Are treatment regimens with antibiotics safe when compared to placebo for leptospirosis? c) Which antibiotic regimen is the most efficient and safest in treating leptospirosis?

Search strategy: Electronic searches and searches of the identified articles were combined.

Studies: Randomised clinical trials in which antibiotics were used as treatment for leptospirosis. Language, date, or other restrictions were not applied.

Participants: Patients with clinical manifestations of leptospirosis.

Interventions: Any antibiotic regimen compared with a control group (placebo or another antibiotic regimen).

Data collection and analysis: Data and methodological quality of each trial were independently extracted and assessed by two reviewers. The random effects model was used irrespective of significant statistical heterogeneity.

Main results: Three trials met inclusion criteria. Allocation concealment and double blind methods were not clearly described in two. Of the patients enrolled, 75 were treated with placebo and 75 with antibiotics: 61 (81.3%) penicillin and 14 (18.6%) doxycycline. The patients assigned to antibiotics compared to placebo showed: a) Mortality: 1% (1/75) versus 4% (3/75); risk difference -2%, 95% confidence interval -8% to 4%. b) Duration of hospital stay (days): weighted mean difference 0.30, 95% confidence interval -1.26 to 1.86. c) Prolonged hospital stay (> seven days): 30% (7/23) versus 74% (14/19); risk difference -43%, 95% confidence interval -70% to -16%. Number needed-to-treat 3, 95% confidence interval 2 to 7. d) Period of disappearance of fever (days): weighted mean difference -4.04, 95% confidence interval -8.65 to 0.58. e) Leptospiruria: 5% (4/75) versus 40% (30/75); risk difference -46%, 95% confidence interval -88% to -3%. Number needed-to-treat 2, 95% confidence interval 1 to 33.

Reviewer's conclusions: Antibiotic regimens for treatment of leptospirosis is a form of care for which the evidence is insufficient to provide clear guidelines for practice. The randomised trials suggest that antibiotics could be a useful treatment for leptospirosis. Because of the questionable quality of two of the three trials, the indication for general use of antibiotics is uncertain. However, the evidence suggest that penicillin may cause more good than harm.