To the Editor:-The article by Bennett et al., (Anesthesiology 1997; 87:1070–4) was interesting and enlightening. I have observed during my personal practice and while working with residents that small doses of sufentanil (1 [micro sign]g/kg or less) will, commonly and unfortunately, produce violent coughing in young children and even in adolescents. Premedication with midazolam has a modest sparing effect. These children were clearly in distress, and some will, if one is not careful, become hypoxic without bradypnea or apnea. When I first noted this reproducible pattern, some of my colleagues, experts in opioid anesthesia, proposed that such coughing may merely reflect reduced chest wall compliance. Dr. Bennett has shed some light on this issue by demonstrating vocal cord closure in adults receiving sufentanil. The authors make no mention that boluses of sufentanil produced paroxysms of coughing in the patients they studied. Possibly this is just one more age-related difference. This phenomenon also may be of interest as an increasing number of patients, children and adults, are sedated with narcotics, given by non-anesthesiologists outside of the operating room. Encouraged by the work in this study, it would certainly seem appropriate to study this phenomenon in other patient age groups and with other, natural occurring and synthetic, narcotics. I am certainly curious how medication commonly believed to be cough-suppressing may have such a paradoxical effect on the vocal cords.

Terrance A. Yemen, M.D.

Head; Department of Anesthesia; Montreal Children's Hospital; Associate Professor of Anesthesia; McGill University; Montreal, Quebec; Canada