To the Editor: Recently, we provided anesthetic care for a 6-mo-old, 8-kg child during craniofacial reconstruction. Because of the 10 degrees head-up position used intraoperatively, a precordial Doppler probe was placed over the right atrium and checked for appropriate position by rapid infusion of intravenous fluid (rapid infusion resulting in a change in the quality of Doppler tones). The intraoperative course was smooth except for blood loss during the scalp incision and dissection of the skull flap; approximately one blood volume was lost intraoperatively. As blood was lost during the dissection, it was replaced with packed red blood cells reconstituted with saline, infused via a Y-type blood pump set connected to a Hotline Trademark fluid warmer (Level I Technologies, Rockland, MA). Occasionally, a change in quality of Doppler tones was noted intraoperatively--a change different from that related to rapid intravenous infusion alone. This short-lived change in Doppler tones was not associated with any change in vital signs or end-tidal CO2; the Doppler tone changes rapidly disappeared when fluid was not being infused. Close inspection revealed that visible air bubbles were forming in the tubing just distal to the Hotline Trademark Figure 1. The bubbles were so small that they could barely be noted unless a fluid bolus was being infused, at which time bubbles seem to coalesce into larger bubbles Figure 2. The entire system was reexamined to ensure appropriate assembly; no air was being entrained from the Y-type pump system hooked into the Hotline Trademark.Figure 1: Note bubbles within the intravenous tubing of the child described (arrows).Figure 2: Note coalescence of bubbles into larger bubbles within the intravenous tubing of the child described (arrows).We were able to replicate this bubble phenomenon at a later time (without a patient!) using infusion of cold fluids through the Hotline Trademark system. Presumably, these bubbles form because of the decreasing solubility of gases in liquid as they are warmed. Owing to their small size, these bubbles are unlikely to be of clinical hemodynamic significance, although they might pose a hazard of paradoxical embolization in patients with a potential for right-to-left intracardiac shunting. In addition, if a precordial Doppler is being used to detect early signs of operative air embolus, these bubbles are, at least, an annoyance and possibly the source of ill-advised therapeutic interventions for treatment of suspected air entry from the surgical field. A Hotline Trademark gas eliminator, L10 Model (Level I Technologies), is available as an optional addition of the Hotline Trademark fluid warmer system. This optional piece of equipment was not being used during the case described, since it comes with the warning "DO NOT USE WTTH HAND PUMPS," a component of our standard intravenous fluid administration set. Clinicians should be aware of this potential problem, especially if a precordial Doppler is being used during anesthesia for pediatric patients in whom an air embolus from the surgical field might occur. G. W. Stevenson, MD Michael Tobin, MD Steven C. Hall, MD Department of Anesthesia, Children's Memorial Hospital, Northwestern University Medical School, Chicago, IL 60614