Retiring Steroids for Hives

This is one of those perfectly unglamorous, yet infinitely practical sorts of topics we encounter in everyday Emergency Medicine. I must see a patient with urticaria, almost always without known underlying trigger or etiology, nearly every other shift. They are itching furiously, and, well – it’s an Emergency!

In true “don’t just stand there, do something!” fashion, I’ve done what I can to help. This typically means “something stronger”, something not over-the-counter, and is usually a dose of dexamethasone to augment antihistamine therapy.

This small trial of 100 patients randomized patients with uncomplicated urticaria to levocetirizine (a H1 receptor-blocker) plus 40 mg of prednisone for four days, or levocetirizine plus placebo. Patients were assessed at several subsequent time points for “itch score”, rash recurrence, and other adverse events – and the winner is: placebo! There was no obvious difference or trend favoring those patients receiving steroids.  There is, however, always the potential for Type II error with such a small sample, but when a positive outcome is difficult to demonstrate, the magnitude of effect is not likely to be large.

Interestingly, they screened 710 patients in order to enroll 100, with 412 not meeting inclusion criteria. These exclusions were mostly evenly distributed between the following criteria: angioedema or anaphylaxis, use of antihistamines or glucocorticoids prior to the ED visit, and rash of greater than 24 hours duration. These limitations do limit the generalizability of these findings, considering their study cohort was ultimately only about one-fifth of all comers. It is probably still reasonable to suggest from a Bayesian sense, at least, steroids should be assumed not to have value in somewhat wider a population than explicitly testing here, but this is not definitive.

“Levocetirizine and Prednisone Are Not Superior to Levocetirizine Alone for the Treatment of Acute Urticaria: A Randomized Double-Blind Clinical Trial”

https://www.ncbi.nlm.nih.gov/pubmed/28476259