To the Editor: O’Leary et al1 are pessimistic about the value of antibiotic treatment for leptospirosis, based on an inconclusive Cochrane review2 and no proven mortality decrease. Yet we can be more optimistic, given the efficacy of prophylactic doxycycline (soldiers in Panama were 95% protected by 200 mg once-weekly3), the susceptibility of leptospira to various antibiotics in vitro (especially penicillin, but not erythromycin4), and the existence of a Herxheimer reaction with penicillin,5 which indicates in-vivo activity.

The Cochrane review2 was of randomised controlled trials of confirmed cases. It included 75 patients given antibiotics and 75 given placebo. Penicillin (61 patients) and doxycycline were not compared. The time from symptom onset to starting antibiotics was stated in only two trials (9 and 2 days). Nevertheless, the review concluded that penicillin or doxycycline may do more good than harm.

In the spirochetaemic phase of leptospirosis (Days 4–7), vasculitis causes multiorgan failure. Successful treatment with antibiotics seems likely if commenced within 2 days of symptom onset, before generalised vasculitis is irreversible. Unfortunately, early diagnosis and efficacy assessment is difficult because symptoms are protean and non-specific, no rapid laboratory test exists, the condition is mild and self-limited in most cases, and mortality varies from zero to 7%. A high index of suspicion is essential so that antibiotics can be commenced empirically.

The consensus is that antibiotics should be commenced within 4 days. In one study, starting antibiotics within 7 days was associated with shorter illness, and if antibiotics were started within 2 days the shorter duration was highly significant (P = 0.006).6 Another trial showed penicillin commenced on Day 9 was beneficial, but antibiotics had been used before entry to the trial.2

Various penicillins and tetracyclines have seemed useful, but only oral doxycycline and intravenous benzylpenicillin are recommended. These are the first and second preferences, respectively, in Therapeutic guidelines antibiotic.7

In the case described by O’Leary et al,1 antibiotics were commenced on Day 3, but the penicillin dose was only half the recommended daily dose for leptospirosis. The vasculitis was probably terminal before the patient was transferred to Concord Hospital.