To the Editor.
—Mountain and Neff have written an excellent article pointing out the problems of intoxication from oral theophylline. However, their abstract contains a statement that is not corroborated in their article.They state that "serum theophylline levels were a poor predictor of serious toxic effects," an assertion barely mentioned in the main text. The authors cite two other studies that show that serious toxic effects are related to theophylline levels of 46.5 to 53.0 mg/L and contrast these levels with the mean maximum theophylline level of 38.2 mg/L in three of their patients who had seizures. However, they fail to point out that a level of 38.2 mg/L is generally considered to be a very high and potentially extremely toxic level. Moreover, the mean maximum theophylline level in their 21 patients was 42.8 mg/L, no patient who experienced toxic effects had levels below 22.4 mg/L, and only three