Although it is lamentable that many herbal preparations of questionable safety are readily available, I think that Anderson and colleagues [1] wrongly characterize one of their cases as "the fatal case" of pennyroyal ingestion. In their article, we find that this patient was a young woman who had attempted to induce abortion and presented with rigors, which were followed by cardiopulmonary arrest, a "rigid abdomen," a "hemorrhagic ... ectopic pregnancy with indications of superinfection," "a substantial amount of old blood," sustained hypotension, and disseminated intravascular coagulation. Why was her death blamed on the pennyroyal ingestion rather than septic shock?

In addition to taking the pennyroyal and black cohosh, had the patient attempted other means of inducing abortion? Sepsis with vascular collapse is a complication of various methods of abortion that have been used by desperate women for centuries. Anderson and colleagues do not comment on the condition of the patient's cervix and vagina on autopsy, and they state that the uterus had "the anticipated ... findings," whatever that means. No blood cultures or intra-abdominal cultures are reported, despite the observed infection and a leukocyte count of 37 x 10⁹/L. Was the patient treated with antibiotics during her hospitalization? What were the opinions of the intensivists and gynecologists who actually cared for this woman? Their absence among the authors of Anderson and colleagues' article is notable.

Anderson and colleagues are baffled by the low menthofuran level found and suggest that "[m]etabolism in humans may differ [from that in animals] in that menthofuran is not a major metabolite." In their own study, patient 2, presumably human, had a menthofuran level of 40 ng/mL. A reasonable explanation is that patient 1, who already had hepatic necrosis from shock, was unable to metabolize the pulgenone to menthofuran.

Figure 2 of the article lacks a pertinent control-a liver sample from a patient who died of shock and disseminated intravascular coagulation without pennyroyal ingestion. Moreover, when lanes 2 and 4 of the blot are put side by side, the bands do not correlate well. The bulk of the hepatic necrosis could still have been caused by prolonged shock, a fact that is not changed by the finding of adducts in the liver.

Perhaps some details were eliminated for this article, but I remain unconvinced that pennyroyal oil was the root of this poor woman's death.