In Response.

Abstract

We thank the editor for giving us the opportunity to respond to Dr Ansari et al,1 who claim that the association of exposure to general anesthesia for cesarean delivery with elevated risk of severe postpartum depression (PPD) might be due to residual confounding and that our conclusion implies a causality.1,2 Dr Ansari et al1 misread our study, and their concerns are unfounded. We discussed extensively in our article the limitations of our study, its proof-of-concept nature, and the need for confirmatory studies. More importantly, we explicitly stated that “[T]he reported association between exposure to general anesthesia and the risk of PPD and suicidality does not necessarily represent a causal relationship,” and “confounding by indication and residual confounding could explain the observed association between general anesthesia and postpartum mood disorders.”2 We also used the E value to evaluate the potential bias from residual confounding and acknowledged that “the exact indication for general anesthesia cannot be accurately identified in administrative data and the propensity score used to estimate the odds of receiving general anesthesia includes only known confounders or available confounders.”2 Dr Ansari et al1 assert that “General anesthesia is not generally performed without a clinical rationale…As such, women who do receive general anesthesia are likely to have had a more complicated peripartum course than women receiving regional anesthesia.” They attempted to back up this assertion by using the results from a quick chart review of 100 cesarean deliveries under general anesthesia performed in a single academic medical center. Based on a convenience sample, their result is not peer reviewed and cannot be taken as evidence to refute our findings. Our recent study in New York State indicates that about 40% of cesarean deliveries performed under general anesthesia did not have such a clinical rationale and could have been potentially avoided.3 Although postoperative depressive disorders are an understudied area in anesthesia outcomes research, there is emerging experimental and epidemiologic evidence linking general anesthesia to increased risk of postoperative depression, as alluded to in our study. Recent research in nonobstetric patients suggests that exposure to surgery under general anesthesia is associated with the development of newly onset postoperative depressive disorders and the acceleration of preexisting depressive symptoms.4 Studies in obstetric patients indicate that exposure to general anesthesia is partially responsible for heightened risk of postpartum depression. Acute or persistent postoperative pain, delayed first interaction with the newborn, and delayed breastfeeding have all been associated with the development of postpartum depression and are more frequent when general anesthesia is used for cesarean delivery.2,5,6 The antidepressant effect of ketamine does not necessarily contradict our findings, because the prophylactic administration of ketamine to surgical patients does not decrease the risk of newly onset postoperative depressive symptoms or the acceleration of preexisting depressive symptoms.4 Similarly, the therapeutic effect of electroconvulsive therapy on severe depressive disorders is due to the electrical current applied to the brain rather than the general anesthetic used for performing this traumatic procedure. Our preliminary findings should be viewed as positive signals that exposure to general anesthesia for cesarean delivery is a partially modifiable risk factor for PPD. More rigorous research is warranted to better establish the relationship between exposure to general anesthesia and PPD, assess the causative nature of this relationship, and elicit the mechanisms underlying this relationship. Jean Guglielminotti, MD, PhDDepartment of AnesthesiologyColumbia University Vagelos College of Physicians and SurgeonsNew York, New York[email protected] Guohua Li, MD, Dr PHDepartment of AnesthesiologyColumbia University Vagelos College of Physicians and SurgeonsNew York, New YorkDepartment of EpidemiologyColumbia University Mailman School of Public HealthNew York, New York