Leptospirosis After the Adventure Race: When Does Oral Stop Being Enough
An ER nurse who competed in a multi-day Costa Rica adventure race develops classic leptospirosis, started on oral doxycycline — and by Day 2 her bilirubin and creatinine have both nudged upward. The debate is whether that's the early signature of Weil's disease demanding IV penicillin, or a mild trend the actual antibiotic evidence doesn't clearly say is worth escalating for.
M.S., a 39-year-old emergency department nurse, spent four days wading through jungle rivers and rice paddies as part of a multi-day adventure race across Costa Rica — the kind of event she trains for year-round with a small group of coworkers, half bonding exercise and half genuine obsession. She came home ten days ago and started feeling unwell four days later: fever, a headache that wouldn't quit, and calf pain severe enough that she initially wondered if she'd torn something during the race, until it was symmetric in both legs and out of proportion to any single injury she could name. As an ER nurse herself, she recognized her own conjunctival suffusion — redness without discharge, a sign she'd only ever seen described in textbooks — before the physician examining her did, and said so.
Leptospirosis is a biphasic illness in its classic form: the early "septicemic" phase she's now two days into, and a second immune phase in a minority of patients where the disease turns genuinely dangerous — jaundice, acute kidney injury, sometimes pulmonary hemorrhage, together known as Weil's disease. She started oral doxycycline the day she was diagnosed and has tolerated it well, fever trending down. But today's labs, drawn as part of routine daily monitoring rather than because anything new happened clinically, show her total bilirubin up from 1.1 to 2.3 mg/dL and her creatinine up from a normal baseline to 1.4 — small movements, not yet meeting most published severity thresholds for frank Weil's disease, but movement in the one direction nobody wanted to see. The actual evidence for switching antibiotic route at this point is thinner than the instinct to switch: the Cochrane review of antibiotic therapy for leptospirosis, by Brett-Major and Coldren, found insufficient evidence that treatment meaningfully changes mortality at all, let alone evidence pinpointing when oral therapy stops being enough.
A bilirubin that moved and a evidence base that hasn't
I'd switch her to IV penicillin now. A bilirubin that's doubled and a creatinine that's newly abnormal are the early signature of the transition toward Weil's disease, and I treat any new hepatic or renal involvement, not just frank jaundice or oliguria, as grounds for parenteral therapy — that's my own threshold, not something a guideline has fixed for me. If early aggressive treatment changes this disease's trajectory at all, it does it before organ dysfunction is advanced, not after.
I recognize she's clinically improving today. That's exactly the window I'd rather not wait past.
I want to be precise about what the evidence actually says here, because it's weaker than the instinct to switch. Brett-Major and Coldren's Cochrane review of antibiotic therapy for leptospirosis found insufficient evidence that treatment changes mortality at all — and her current numbers don't meet most published thresholds for frank Weil's disease. She's improving on a drug she's tolerating well.
Switching her to IV penicillin isn't a risk-free move either — there's a documented Jarisch-Herxheimer-like reaction associated with penicillin treatment of leptospirosis. I'm not willing to trade a real, if uncertain, risk for a switch the mortality data don't clearly support.
Two data points is not a trend, and both of you are treating it as one. A bilirubin drawn twice and a creatinine drawn twice cannot distinguish a curve from a wobble, and every argument either of you has made depends on which of those it is. What I'd do is admit her and keep her on the same oral doxycycline she's already tolerating, and let trending bilirubin, creatinine, and platelets over the next day or two actually answer the question neither of you can answer from two data points alone.
Observation isn't indecision here — it's the honest response to real uncertainty on both sides. If her numbers keep climbing tomorrow, we'll have a much easier answer than we do right now.
Agreed: admit for inpatient monitoring, continue oral doxycycline unchanged, and trend bilirubin, creatinine, and platelets every 12 hours rather than switch route today on the strength of two data points.
Not agreed, and explicitly left open rather than resolved: the infectious disease physician was not fully convinced by the observation plan and said directly that she would revisit pushing for IV penicillin sooner than 24-48 hours if the trend continued in the same direction at all, even before crossing a formal severity threshold. The clinical pharmacologist held that the same data, moving the same direction, would still not justify the switch absent a clearer sign of genuine organ failure, given how thin the underlying antibiotic-benefit evidence remains. Both plans are written into the chart as contingencies keyed to tomorrow's labs; which one actually gets followed depends on numbers nobody at the bedside has yet.