To the Editor:—
We are greatly interested in the recent article by Faberowski et al. 1that analyzed the incidence of venous air embolism (VAE) in a small series of children undergoing repair of craniosynostosis. We were very impressed by the 82.6% incidence of VAE detected by precordial Doppler monitoring. As stated in their article, this percentage was much higher than any previously reported incidence of VAE during craniectomy in infants. Harris et al. 2used echocardiography for detection of VAE in infants undergoing craniectomy, some of whom were at risk of VAE because of major cranial malformations, and found what we considered to be a very high incidence of 66%. In our own experience, VAE occurred in only 3 of 130 children (2.6%) undergoing repair of craniosynostosis. 3In all cases, these patients had complex vault remodeling, heavy perioperative blood losses, and severe hypotension during VAE. None of these patients experienced postoperative consequences of VAE. Since that time, surgical prevention was enhanced, and no additional cases were noted in the previous 3 yr in more than 350 procedures.
It could be argued that the detection at our institution is only based on continuous end-tidal carbon dioxide monitoring, which is much less sensitive than Doppler monitoring. However, extrapolating from the results of Faberowski et al. , 1the incidence of VAE that is detectable by capnography could be more than 40%; 23 times greater than in our experience. Minimal venous air embolisms probably occur very frequently during vault resection before the surgeon can apply efficiently bone wax. This risk of air entry is increased in the presence of hypovolemia related to abrupt blood losses. If a very sensitive monitor is used, these minimal and short-lasting episodes of VAE will be detected. In these conditions, it is not surprising that only 30% of the children experiencing VAE had related hypotension; but the question of the clinical implications of detecting such a small amount of air entry is not answered. In the study by Cucchiara et al. , 436% of the adult patients in the sitting position and experiencing VAE had hypotension. In a similar pediatric population, we found an 85% incidence of cardiovascular variations related to VAE, 5which is in greater accordance with the reported incidence of hypotension related to VAE in pediatric patients.
The authors are to be congratulated for pointing out the problem of VAE during craniosynostosis repair. However, a possible conclusion drawn from this article could be that only 18% of the children undergoing craniosynostosis repair could be spared perioperative episodes of VAE that increase morbidity and mortality. This probably does not reflect the clinical practice of other centers with extensive experience with this type of surgery.