In Response.

Abstract

We thank Dr Frisch et al1 for their interest in our studies2,3 and for proposing an intriguing explanation for the lower estimated hazard ratios observed in the boys compared to the girls. In our study, we conceded that effects of the anesthetic could not be distinguished from the surgical procedure, inflammation from the procedure, or the potentially stressful perioperative experience. In their correspondence, Dr Frisch et al1 suggest that it is perhaps the painful and traumatic experience of surgery that may account for the observed increased attention deficit hyperactivity disorder (ADHD) risk in children exposed to surgery and anesthesia. One potentially painful procedure they highlight is neonatal circumcision, which is commonly performed in boys in the United States. In our study, we considered boys who were circumcised during the neonatal period as being “unexposed” because prior studies have found that a vast majority of these procedures are done without general anesthesia. Frisch et al1 raise the concern that if painful procedures are detrimental, by considering boys with circumcisions in the neonatal period as unexposed, we may be placing children who are at increased risk for ADHD in our control group, which could result in an underestimate of the hazard ratios for ADHD in the anesthetic-exposed boys. We evaluated 4 common procedures in our study, pyloromyotomies, inguinal hernia repairs, circumcisions outside the perinatal period, and tonsillectomies and/or adenoidectomies. If painful procedures themselves pose neuropsychiatric risks, the hypothesis of an underestimate in hazard ratios in the anesthetic-exposed boys due to neonatal circumcision is certainly possible. The impact would be particularly profound on our results evaluating boys who received circumcisions outside the perinatal period because they are compared to a control group that includes a large proportion of boys who received neonatal circumcisions and who, in theory, would also be at increased risk for ADHD. However, when evaluating children who were exposed to the remaining procedures (pyloromyotomies, inguinal hernia repairs, and tonsillectomies and/or adenoidectomies), we would expect that neonatal circumcisions would be balanced between the exposed and unexposed children, and, thus, neonatal circumcisions would have a limited impact on biasing the hazard ratios. In addition to suggesting this possible explanation, Frisch et al1 also propose a number of additional analyses. These suggestions are sensible and constructive. However, they are also complex because neonatal circumcisions are elective in nature and highly associated with the religion or culture of a child’s family. This family information is not available in the Medicaid data set, so evaluating the long-term impact of circumcisions while accounting for these demographic differences is challenging and beyond the scope of this correspondence. The concept that procedural pain or inflammation early in life may contribute to impaired brain development is one that has been evaluated by others and has not been ruled out as a potential mechanism behind neurodevelopmental deficits after exposure to surgery and anesthesia. Interestingly however, the general anaesthesia or awake-regional anaesthesia in infancy (GAS) trial randomly assigned infants to receive either sevoflurane anesthesia or a regional anesthetic for inguinal hernia repair.4 In this study, all children were exposed to the pain and trauma from surgery, with the difference being the type of anesthetic that was used. Compared to children who had a regional anesthetic, those exposed to sevoflurane had worse executive function scores. While these results should be interpreted with caution because executive function was one of a host of secondary outcomes, this type of study presents a methodologically sound way to independently evaluate the impact of a general anesthetic. We again thank Dr Frisch et al1 for their interest and constructive comments and would be pleased to take their suggestions into consideration in future studies. Caleb Ing, MD, MSDepartments of Anesthesiology and EpidemiologyColumbia University College of Physicians and Surgeons and Mailman School of Public HealthNew York, New York[email protected] Melanie M. Wall, PhDDepartments of Psychiatry and BiostatisticsColumbia University College of Physicians and Surgeons and Mailman School of Public HealthNew York, New York Mark Olfson, MD, MPHDepartments of Psychiatry and EpidemiologyColumbia University College of Physicians and Surgeons and Mailman School of Public HealthNew York, New York Guohua Li, MD, DrPHDepartments of Anesthesiology and EpidemiologyColumbia University College of Physicians and Surgeons and Mailman School of Public HealthNew York, New York