When to extubate the croup patient: the "leak" test.

The reliability of "leak" as a criterion for elective extubation of children with croup treated with nasotracheal intubation was studied during the three month "croup epidemic" in Ontario of September 1, 1983 to November 30, 1983. Twenty-eight patients experienced 36 extubations; 31 planned and five accidental. Three of 23 (13 per cent) planned extubations with "leak" required reintubation whereas three of eight (38 per cent) children electively extubated after seven days of intubation without "leak" required reintubation. "Leak" is a helpful but not absolute prognostic indicator of a successful extubation.