A Childhood Penicillin Rash and an Infection That Won't Wait for Skin Testing
A validated low-risk score and a rising creatinine push a decades-old penicillin label toward same-day testing instead of the slower, traditional skin-test-first pathway.
Harold P., a 67-year-old man, has told every doctor who's asked for the past fifty-plus years that he's allergic to penicillin, going back to a rash his mother described after a course of it when he was 9 — a detail he knows secondhand, not from memory, and has never had reason to test since. He was admitted three days ago with rapidly spreading cellulitis of his lower leg, and blood cultures have now grown methicillin-susceptible Staphylococcus aureus and group A streptococcus, both reliably treatable with a narrow-spectrum beta-lactam. Because of his penicillin label, he was started empirically on vancomycin, and his creatinine, stable at admission, has crept up over the past 48 hours in a pattern his team recognizes as an early vancomycin-related trend rather than progression of his underlying infection.
Applying the PEN-FAST clinical decision rule to what he can actually recall — a rash roughly fifty-eight years ago, well outside the rule's own five-year window, without hives, anaphylaxis, or any treatment beyond stopping the drug — scores him a zero on all three criteria Trubiano's derivation and validation cohorts identified, the very-low-risk band carrying under a 1% chance of a positive penicillin allergy test rather than the 5% that a score of one or two would imply. Skipping skin testing on that basis is a separate finding from a separate trial: PALACE, Copaescu's randomized comparison, is what showed direct oral challenge to be non-inferior to skin testing first in patients scoring under three — a difference that matters directly for a man whose kidneys are already telling the team how much longer vancomycin is a good idea.
A rising creatinine putting a deadline on how long this can wait
Both of his organisms are exactly what a narrow-spectrum beta-lactam treats best, and every day he stays on vancomycin is a day of unnecessary levels, unnecessary nephrotoxicity risk, and broader-spectrum coverage he doesn't need. His creatinine is already trending the wrong direction. I want him off vancomycin and onto penicillin or a related beta-lactam as soon as it's safe — not eventually, this admission.
I'm not asking anyone to skip a safety step to get there faster. I'm asking that whatever safety step we use, we use it today rather than defaulting to the slowest available option out of habit.
Applying PEN-FAST to what he can actually recall — a rash, no hives, no anaphylaxis, fifty-eight years before this admission and so nowhere near the rule's five-year cutoff — scores him zero, Trubiano's very-low-risk band, under 1 percent predicted risk of true immediate hypersensitivity rather than the 5 percent people tend to quote for the low-risk group generally. And the permission to skip skin testing doesn't come from that derivation study at all — it comes from PALACE, where Copaescu randomized patients scoring under three and found direct oral challenge non-inferior to the skin-test-first sequence. I'd recommend the direct challenge today, under observation, rather than the traditional skin-test-first pathway, which would cost him another day or more he may not have given his kidneys.
Skin testing first isn't wrong, exactly — it's simply not the standard the evidence supports for a patient who scores this clearly low-risk, and defaulting to it here would cost him time the stewardship argument already told us matters.
I agree with the direction, and I don't want to be the reason this takes longer than it needs to. What I want before we proceed is the actual monitoring plan spelled out for the floor nurses doing this challenge, not just the decision to do it — he's 67, and even a low-probability reaction in a patient with rising creatinine and an active infection is one I want caught in the first ten minutes, not found on a routine check an hour later. Give me a written observation protocol and I'll get it started this afternoon rather than waiting for tomorrow's allergy consult schedule.
Agreed: direct oral amoxicillin challenge performed that afternoon under nursing observation, using a written monitoring protocol the allergist provided at the hospitalist's request. He tolerated the challenge without any reaction, his penicillin allergy label was formally removed from his chart, and he was transitioned to IV cefazolin that same evening.
His creatinine began trending back down within 24 hours of stopping vancomycin. No disagreement carried past the monitoring-protocol request — once that was in hand, all three specialties considered the plan settled.