Version July 2025
INTRODUCTION —
The possibility that anesthetic and sedative drugs may be neurotoxic to the developing brain is of widespread importance to patients, parents or caregivers, and the clinicians who care for them. This topic will discuss the animal studies that initially questioned the safety of anesthetics on the developing brain, the proposed mechanisms of damage, as well as the clinical studies and the limitations of studies in humans.
THE ISSUE —
Studies in animals have shown that anesthetic and sedative drugs can cause neuronal damage to the developing brain, resulting in lasting behavioral and cognitive deficits. Some human studies also suggest an association between exposure to anesthesia in early childhood and subsequent neurodevelopmental deficit, but causation has not been established, and specific anesthetic and patient factors that may affect risk remain unclear [1].
Any degree of neurotoxicity of anesthetic and sedative drugs in humans may have widespread implications. In the United States, over 1.3 million children under the age of five years undergo surgical procedures each year [2,3]. Most of these children require anesthesia to tolerate these procedures, and most of these procedures cannot be delayed without adding substantial surgical risk. In addition, an increasing number of young children receive sedatives or general anesthesia for neuroimaging studies [4], with many more requiring procedural sedation in the emergency department.
In 2016, the US Food and Drug Administration (FDA) issued a warning about potential risks of negative effects on the developing brain from administration of anesthetics for children under age three, as well as pregnant women during their third trimester, especially for repeated exposures or procedures lasting more than three hours (see 'The FDA warning' below). Most children who undergo general anesthesia are exposed for less than the FDA defined high-risk threshold of three hours. In a retrospective cohort analysis of approximately 1.5 million pediatric (<18 years of age) anesthetic billing records from the National Anesthesia Clinical Outcomes Registry, the median anesthetic duration was 57 minutes, with six percent of anesthetics lasting longer than three hours, but in children under one year old, 13.7 percent of anesthetics lasted longer than three hours [5]. In another retrospective study of approximately 21,000 children born between 1994 and 2007 in Minnesota, 14.9 percent of children were exposed to anesthesia before three years of age. Of the children exposed to anesthesia, 26.3 percent had either one exposure longer than three hours or multiple anesthetic exposures [6].
PROPOSED MECHANISMS —
Much of the research on the mechanisms of anesthetic neurotoxicity has focused on apoptosis, though there is evidence that other mechanisms may be involved.
●Oxidative stress-associated apoptosis – Anesthetic agents cause neuronal oxidative stress by inducing reactive oxygen species (ROS) formation within mitochondria during exposure [7,8]. ROS produced in this setting permits cytochrome c mobilization from the inner mitochondrial membrane [9,10]. Simultaneously, the outer mitochondrial membrane becomes permeable following anesthetic-mediated gamma-aminobutyric acidA (GABAA) receptor stimulation and N-methyl-D-aspartate (NMDA) receptor antagonism [11]. The combination of these events allows cytochrome c to be released into the cytosol and sets in motion a cascade of cellular events that ultimately leads to apoptotic neuronal cell death.
●Other mechanisms – Laboratory studies involving human embryonic stem cells and rodent neurons have suggested that anesthetics may impair neural development during critical periods through abnormalities of NMDA and/or GABA receptor expression, activity, and signaling; neuroinflammation; and dysregulation of other growth and survival pathways [7,10-16]. Furthermore, other downstream developmental processes, such as synaptogenesis, neurogenesis, and dendritic branching, may be targeted and disrupted.
NEUROTOXICOLOGIC EFFECTS IN ANIMAL STUDIES —
Animal studies in multiple species from worms to nonhuman primates have reported neuronal apoptosis, and some studies have reported behavioral changes associated with exposure to anesthetics [17-19]. By contrast, a handful of preclinical studies failed to identify deleterious neurobehavioral effects of anesthetics [17]. Virtually all of the commonly used anesthetic agents (eg, propofol, midazolam, nitrous oxide, isoflurane) have been shown to elicit a degree of a neurotoxicologic response in some studies.
Rodent studies — An initial study on anesthetic neurotoxicity reported that exposure of immature rodents to a combination of commonly used anesthetic agents (ie, midazolam, nitrous oxide, and isoflurane) resulted in deletion of neurons, defects in synaptic function in the hippocampus, and cognitive impairment as juveniles and young adults [20]. Subsequent studies have corroborated these findings in both mice and rats, using a variety of anesthetic agents, durations of exposure, and a range of outcome measures [21].
Translation of the results of rodent studies to humans is limited by a number of factors, including the following:
●Most importantly, the greater complexity of the human brain and longer period of development in humans may result in a greater capacity for neuroplasticity and recovery from any insult.
●The differential timing and rate of brain maturation across species. (See 'Vulnerable age periods' below.)
●Dose and duration of anesthesia in animal studies that often exceed clinically relevant equivalents in humans.
●Difficulty monitoring physiologic parameters in rodents (eg, hypoxia, hypercarbia) that could affect neurologic outcome.
Nonhuman primate studies — Several studies have reported that both single and multiple exposures to anesthetic agents in nonhuman primates (NHPs) result in long term neurodevelopmental defects. The duration of a single exposure was as long as 24 hours. These studies are likely of greater translational value than rodent studies because the neurodevelopmental stage of NHPs is similar to humans at birth. In addition, in contrast with rodent studies, the conditions of human anesthesia can be mimicked in NHPs, and physiologic parameters can be measured and homeostasis maintained during anesthesia, as they would be in humans.
●In one study, infant rhesus monkeys that were exposed to sevoflurane for four hours on days of life 6 to 10, followed by subsequent exposures on days of life 14 and 28, demonstrated increased anxiety compared with controls [22].
●In another study, infant rhesus macaques that were exposed to isoflurane for five hours on days of life 6, 9, and 12 demonstrated motor reflex deficits at one month of age and increased anxiety at 12 months of age [23].
●Studies have reported neuronal apoptosis in the fetal and neonatal NHP brain induced by prolonged exposure (5 to 24 hours) to isoflurane [24-26], ketamine [27-29], or propofol [30]. In one study, an increase in astrocyte activation was found in these same brain regions two years after a five-hour neonatal exposure to isoflurane [31].
●In a post hoc analysis of imaging data obtained in NHPs at either 12 or 18 months of age, repeated exposure to ketamine or isoflurane was associated with widespread reduction in white matter integrity [32]. There were no evaluations to assess the functional effects of these changes, or long-term follow-up for potential resolution. However, an imaging study of two year old rhesus macaques exposed to isoflurane for five hours on days of life 6, 9, and 12 found changes in resting-state functional connectivity in several brain regions that correlated with social behavior abnormalities and increased astrogliosis [33].
Translation of the results of NHP studies to humans is limited by several factors:
●Assessment of neurodevelopmental outcome in other animal species cannot address the complexity of higher-order cognitive function in children [34-37].
●Controlled experimental exposure of animals to anesthesia does not model or account for the confounding indications for surgery and anesthesia exposure known to occur in infants and children. While studies in humans almost universally evaluate anesthetic exposure associated with surgery, the surgical exposure and any resulting inflammation are not typically addressed in animal models of anesthetic exposure.
●The prolonged administration of anesthetics used in some of these studies may not be relevant for many common procedures in humans.
●It has not been possible to prove that anesthesia-induced neuronal cell death or other histopathology actually occurs in humans, since histologic assessment cannot be performed on the brains of relatively healthy children after exposure to anesthesia [34].
SUMMARY OF THE EVIDENCE IN HUMAN STUDIES —
While animal studies have reported neurotoxic effects due to anesthetic medications, human studies have been more difficult to interpret due to ethical and practical considerations.
Children generally receive anesthesia to tolerate a surgical or diagnostic procedure, therefore in most clinical studies it is difficult to distinguish the effect of the anesthetic medication from other factors, including the medical problem requiring treatment. However, there has been increased recognition of the specific neurodevelopmental outcomes that differ in children with early exposure to surgery and anesthesia, the scale of these domain specific differences, and the fact that children with multiple exposures have worse outcomes than those with single exposures.
Several relevant meta-analyses have been performed. While meta-analyses of observational studies do not mitigate the fundamental limitations of potential bias and confounding inherent to such studies, pooling data from multiple studies has allowed for the assessment of more subtle differences than individual studies. These meta-analyses have found worse neurodevelopmental outcomes in children exposed to surgery and anesthesia compared with children with no exposure, though the differences are modest, particularly after single exposures. Larger effects of exposure have been observed in some outcomes (eg, behavioral problems), whereas no or limited effects on intelligence have been observed.
Examples of relevant meta-analyses include the following:
●A 2021 meta-analysis of three studies (general anaesthesia compared to spinal anesthesia [GAS] [38], Pediatric Anesthesia Neurodevelopment Assessment [PANDA] [39], Mayo Anesthesia Safety in Kids [MASK] [40]; total 1644 children) that compared prospectively collected neurodevelopmental outcomes in children who had previously had a single exposure to GA versus unexposed children, intelligence quotient (IQ) was similar in the two groups [41]. However, parent-reported externalizing, internalizing, and total behavioral scores were worse in the exposed compared with unexposed children. The GAS, PANDA, and MASK studies are described below.
●A 2022 systematic review included studies that compared the association between exposure to surgery and anesthesia and outcomes in each of nine different neurodevelopmental domains [42]. Meta-analysis was performed for studies in each of the domains. All differences in outcomes in exposed children were small. There were associations between exposure and more behavioral problems and neurodevelopmental disorder diagnoses, as well as deficits in executive function, nonverbal reasoning, motor function, and, to a lesser extent, language, general development, and academics. The weakest association with exposure was in cognition scores. Studies that evaluated single versus multiple exposures to surgery and anesthesia found larger differences in several domains in children with multiple exposures. However, all conclusions implying causation are limited by potential confounding by indication and the presence of comorbid conditions that require multiple procedures requiring anesthesia.
STATE OF THE LITERATURE
Limitations of the literature — There has only been one published randomized clinical trial of the effects of anesthesia on neurodevelopment. (See 'Randomized trial' below.)
All the other clinical studies are observational; most have the following limitations:
●The major limitation of the observational studies is the inability to determine whether an observed detrimental effect is due to the anesthetic medication or other factors such as the underlying medical condition (known as confounding by indication), the surgery itself, or psychological trauma in the child resulting from hospitalization.
●Existing studies use a variety of outcome measures, with variable sensitivity for assessment of anesthetic effects.
●Existing studies have involved varied timing of exposure to anesthesia and neurodevelopmental assessment of the children.
●In most studies, there is heterogeneous or unknown anesthetic drug administration.
●Many studies are limited to children in one geographic area or institution.
Some studies have tried to mitigate the impact of these factors by using specific study designs and analytic methods discussed below, including one randomized trial, a sibling matched study, and a study of children who underwent appendectomy (which is typically not associated with underlying medical conditions). Studying prenatal exposure to anesthesia for maternal surgery may reduce confounding by pre-existing fetal conditions but may still be affected by maternal physiologic changes related to the need for surgery. (See "Anesthesia for nonobstetric surgery during pregnancy", section on 'Effects of anesthetics on the fetus and the pregnancy'.)
Randomized trial — Randomized controlled trials (RCTs) are generally the preferred approach for evaluating causality because they are uniquely able to balance measured and unmeasured confounders in exposed and unexposed subjects. However, randomized trials are challenging and time-consuming to perform. The general anaesthesia compared to spinal anesthesia (GAS) trial is the only published RCT evaluating anesthetic neurotoxicity in children [38]. The GAS trial was an international assessor-masked RCT including 722 infants who were randomly assigned to receive general anesthesia (GA) with only sevoflurane versus awake regional anesthesia (spinal and/or caudal) for inguinal hernia repair [38]. The median duration of anesthesia in the GA group was 54 minutes. Intelligence at five years of age as measured by the Wechsler Preschool and Primary Scale of Intelligence Third Edition Full Scale Intelligence Quotient (WPPSI-III FSIQ) was similar in the two groups. The mean FSIQ score was 99.1 in the awake regional anesthesia group, versus 99 in the GA group, mean difference 0.2 (95% CI -2.6 to 3.1). The majority of secondary behavioral and neurocognitive outcomes were also similar between groups.
A limitation of the GAS trial is that it may not be generalizable to children who undergo prolonged anesthesia, or who receive multiple anesthetic agents. In addition, 84 percent of the infants in the GAS trial were male, and it is unknown whether sex affects neurotoxicity.
GAS trial data was subsequently used for a secondary analysis that evaluated children with multiple exposures who were enrolled in the study [43]. (See 'Repeated exposure to anesthesia' below.)
Sibling-matched study — It possible that children who receive anesthesia and surgery differ from those who do not from a sociodemographic or familial standpoint. The Pediatric Anesthesia Neurodevelopment Assessment (PANDA) study tried to address this issue by evaluating 105 sibling pairs in which only one sibling had a single exposure to anesthesia for inguinal hernia repair before age 36 months [39]. The primary outcome was intelligence quotient (IQ) scores, while a variety of secondary outcomes at age 8 to 15 years were also evaluated. This sibling cohort design minimized confounders (eg, genetics, parental education level, and socioeconomic factors) and the study reported no difference between exposed and unexposed siblings in IQ or neurodevelopmental domains including language, executive function, or memory.
NEURODEVELOPMENTAL OUTCOMES EVALUATED —
The vast majority of studies utilized pre-existing datasets. These studies were therefore reliant on the existing patient population and outcome measures that were assessed. Some representative studies are discussed here, according to the outcome measures that were tested.
Academic and teacher evaluation — A number of studies have used population wide educational databases to evaluate academic achievement testing and teacher ratings in children exposed to anesthesia. Most studies evaluating academic achievement or teacher evaluation in children exposed to anesthesia have reported either no difference or a marginal decrease compared with unexposed children.
A limitation of these studies is that academic achievement testing and teacher evaluations may be an insensitive measure of the effects of anesthesia. In studies in which multiple parameters were evaluated, there has been no difference in academic test results between study groups, whereas neuropsychologic testing showed an effect of anesthesia [44]. It should be noted that academic achievement tests often vary among states in the United States and among countries in the world.
●In three studies from Canada, the results of the early developmental index (EDI), a standardized developmental assessment teacher questionnaire evaluating five to six year olds across Canada, were compared for children who had surgery and anesthesia and those who had not, with results ranging from showing small differences in EDI scores [45,46] to no difference when performing a sibling-matched study [47].
●In a study evaluating a large proportion of children in Sweden, in children with surgery and anesthesia, small differences in academic achievement scores were observed as well as slightly lower IQ scores in a subset of boys evaluated for military enlistment [48].
Learning disability — Development of learning disability has been associated with exposure to anesthesia and surgery, particularly with multiple exposures. Learning disability is defined as a combination of academic achievement and IQ testing [49].
In one population-based, retrospective birth cohort study that included 5357 children from Minnesota born between 1976 and 1982, the 593 children who received general anesthesia for surgery before age four had an increased risk of learning disability with two anesthetics (hazard ratio [HR] 1.59, 95% CI 1.06-2.37) and three or more anesthetics (HR 2.60, 95% CI 1.60-4.24), but no difference was found with a single exposure to anesthesia [50]. This study was repeated with 1057 children from Minnesota born between 1996 and 2000, which also found an increased risk of learning disability in children with two or more exposures (HR 2.17, 95% CI 1.32-3.59), but no increased risk in children who had a single exposure [51].
Neurodevelopmental and psychiatric disorders — Exposure to surgery and anesthesia at a young age has been associated with an increased risk of subsequent clinical diagnoses for psychiatric and neurodevelopmental disorders. In some studies, increased risk was associated only with multiple anesthesia exposures, while other studies have reported such an association with a single exposure.
●In a 2022 meta-analysis of four observational studies, a single exposure to surgery and anesthesia was associated with an increased risk of subsequent diagnosis of attention deficit hyperactivity disorder (ADHD; HR 1.30, 95% CI 1.25-1.36) compared with no exposure to surgery and anesthesia [42].
●In a 2024 meta-analysis of five observational studies including children who had two or more exposures to general anesthesia at <5 years of age, exposure to anesthesia was associated with an increased risk of diagnosis of ADHD prior to age 19 (HR 1.71, 95% CI 1.59-1.84) [52].
Neuropsychologic assessments and parental reports of behavior — A wide range of neuropsychologic assessments and parental reports have been used in studies of children exposed to anesthesia with varying results.
●In a study of children in an Australian birth cohort, those exposed to anesthesia before age three were found to have increased deficits in receptive language, expressive language, and abstract reasoning. [53]. In a study of children from Singapore with minor surgery, no differences in cognitive, language, motor, or socioemotional parameters between children exposed to anesthesia were observed, but parent-reported behavior scores were worse at 24 months [54].
●In the Mayo Anesthesia Safety in Kids (MASK) matched cohort study, 997 children who had no anesthesia, a single exposure to general anesthesia, or multiple exposures prior to age three, were tested for general intelligence and neuropsychologic domains [40]; 380 children had one anesthetic, 206 children had multiple anesthetics, and 411 had no anesthesia. There was no difference in intelligence quotient (IQ) among groups. Multiple anesthetic exposures were associated with modest decreases in processing speed, fine motor coordination, and increased parent-reported behavioral problems compared with unexposed children. Parents also reported more difficulty with executive function and reading in both singly- and multiply-exposed children.
●In evaluating children from the Avon Longitudinal Study of Parents and Children (ALSPAC), a birth cohort from England, those with single and multiple exposures had lower scores in social communication and manual dexterity, children with multiple exposures also had lower scores in dynamic balance and an increased number of behavioral difficulties. Differences however were not seen in a number of other outcomes including general cognitive ability, attention, memory, language, and academic achievement [55].
VULNERABLE AGE PERIODS —
The age at which the developing human brain may be most vulnerable to the neurotoxic effects of anesthetic drugs is unclear. Given that this vulnerable period is unknown, and whether anesthetic medications cause neurodevelopmental deficits in children also remains unresolved, there is little support for delaying necessary minor surgery in young children to reduce neurodevelopmental risks.
Whereas brain development extends from the embryonic period until and through adolescence, neurodevelopment may be more susceptible to environmental insult during specific periods depending on the toxicant [56]. In rodent studies, the vulnerable period to anesthesia is thought to correspond to the period of maximal synaptogenesis (ie, approximately two days before birth until two weeks after birth) [57], based solely on findings of induced apoptosis and neurodegeneration during this time period [21,58,59]. Translation of this data to humans is debated as synaptogenesis in children peaks at under six months of age in the sensorimotor cortex, at approximately 10 months of age in the parietal and temporal cortices, and at approximately age three years in the prefrontal cortex [60].
Studies on neurotoxicity have involved children exposed to anesthesia from as early as birth to as late as 10 years of age. While some studies have found varying risk of neurodevelopmental deficit based on the age at exposure to anesthesia, most have not accounted for the fact that children tend to receive different procedures at different ages.
●In one study, children exposed to anesthesia before age three had an increased risk of deficits in language and reasoning, but the same deficits were not present in children exposed between ages three and five, and 5 and 10, suggesting a potential vulnerability at a younger age [61].
●The opposite result, however, was found in one of the studies from Canada, finding that children exposed between ages two and four years were more likely to have deficits than children exposed from age zero to two years, suggesting vulnerability at an older age [46]. In this study, however, the majority of children exposed at ages two to four years received anesthesia for dental procedures. It is possible that children who require anesthesia for dental procedures have a higher incidence of underlying behavioral problems compared with children who do not require anesthesia for these procedures.
●One other large retrospective study compared the risk of a mental health diagnosis or developmental delay for children in 11 separate age cohorts (from <28 weeks to 5 years) undergoing minor common surgical procedures with matched controls [62]. Although there was an increased risk for a mental health disorder, developmental delay, or attention deficit hyperactivity disorder (ADHD), in children with anesthesia exposure (hazard ratio [HR] 1.26-1.31), there was no difference in risk between the different age cohorts.
PRENATAL EXPOSURE TO ANESTHESIA —
Data on the neurotoxic effects of in utero exposure to anesthesia are limited. While some studies have reported associations between prenatal exposure to general anesthesia and worse neurodevelopmental scores, potential confounders may be present. Given the risk of complications to mothers and children, necessary surgery during pregnancy should not be avoided or delayed.
Fetal effects of anesthesia and surgery during pregnancy are discussed separately. (See "Anesthesia for nonobstetric surgery during pregnancy", section on 'Effects of anesthetics on the fetus and the pregnancy'.)
POTENTIALLY TOXIC EXPOSURE DOSES, REPEATED EXPOSURE, AND SPECIFIC DRUGS —
The minimum potentially neurotoxic doses of sedatives and anesthetics in humans are unknown. While subanesthetic doses of propofol [63], ketamine, midazolam [64], and isoflurane [16] have been found to cause neuronal apoptosis in rodent brains, exposure to specific doses and agents has not been well studied in humans.
Repeated exposure to anesthesia — Most human studies evaluate exposures to general anesthesia but lack information on the specific drugs and doses administered, or even the duration of anesthesia exposure. Thus, the number of exposures to anesthesia has commonly been used as a surrogate for dose, with a single exposure considered to be a lower dose, and multiple exposures a higher dose. However, this approach may not account for the fact that some children undergo single complex procedures with long exposures, while others undergo multiple short straightforward procedures. Whether the number of exposures has an effect independent of the total duration of anesthesia is also unknown.
●CF-GAIN enrolled 97 children with cystic fibrosis who had previously participated in a randomized trial comparing repeated bronchoalveolar lavage-directed therapy (BAL) under general anesthesia versus standard care without BAL, up until the age of five years [65]. The children were assessed at a mean age of 12.8 years with neurocognitive and neurobehavioral tests and brain MRI; neurodevelopmental testing and brain MRI results were similar in the two groups. The standard treatment group received a median of zero anesthetic exposures prior to age two years, two exposures prior to age four years, and four exposures at the age of assessment, while the BAL group received a median of two exposures prior to age two years, six exposures prior to age four years, and 10 exposures prior to age of assessment. Cumulative anesthesia exposure times were available for a subset of children in each group and were a median of 180 minutes in the BAL group and 48 minutes in the group who received standard care. This study is important as, apart from the general anesthesia compared to spinal anesthesia (GAS) trial, it is the only study in which exposure was driven by randomization. As a result, it provides evidence of lack of effect of repeated exposure to anesthesia on neurodevelopment with more effective mitigation of confounding. However, the number of subjects is fairly small, limiting the power to rule out minor differences. In addition, results of this study may not be directly comparable to other studies of repeated exposures, as it compared children with a higher number of exposures to a second group with a lower number of exposures, rather than comparing repeated exposure to no exposure.
●The original GAS trial compared infants who were randomly assigned to an awake-regional anesthetic versus those with a sevoflurane general anesthetic. A subsequent observational study evaluated multiple exposures in the same group of children [43]. Ninety of the GAS trial participants who received multiple general anesthetics after the original exposure and prior to age 5 years were compared with 141 children who had only one or zero more general anesthetics prior to age 5. In contrast with the CF-GAIN trial, children with multiple exposures had full scale IQ scores nearly 6 points lower (-5.8, 95% CI -10.2 to -1.4), lower scores in performance and verbal IQ, and more emotional, behavioral, and executive function difficulties, compared with children with one or no anesthetic exposures. These results should be interpreted cautiously due to the small sample size and potential confounding.
●In the studies in Minnesota that used learning disability as an outcome, deficits have only been found in multiple exposures, but not single exposures [50,66,67].
●In studies of academic achievement, the association between exposure number and decreased neurodevelopmental function has been inconsistent [45,46,48].
Doses of anesthetics — Some studies have evaluated types of drugs and exposure doses. Since higher doses of and more exposures to anesthetic medications may be associated with higher levels of comorbid illness, conclusions from these studies may be limited by potential confounders, as discussed above. (See 'Limitations of the literature' above.)
●In one retrospective observational study that evaluated the effects of administration of specific anesthetic or sedative drugs to very preterm infants on subsequent full scale intelligence quotient (FSIQ), administration of all drugs other than opioids was associated with 3 to 4 point average reduction in FSIQ at 36 months corrected age, but no association between opioid administration and any changes in FSIQ [68].
●Another retrospective study evaluated the cumulative sedative and anesthetic doses and subsequent neurodevelopmental outcomes in a cohort of 110 infants with surgery for congenital heart disease [69]. There was no association between the dose of volatile anesthetic agents, opioids, benzodiazepines and dexmedetomidine on motor, cognitive, and language scores at 18 months of age. However, higher ketamine doses were associated with poorer motor performance.
THE FDA WARNING —
Based on the published animal and clinical studies, as well as input from neurotoxicity investigators and stakeholders, in December 2016, the US Food and Drug Administration (FDA) announced warnings about potential risks of negative effects on the developing brain from administration of anesthetics and sedative drugs to pregnant women and children under age three, as well as pregnant women during their third trimester, especially for repeated exposures or procedures lasting more than three hours [70]. The FDA recommends that health care providers discuss with pregnant patients and parents of young children the benefits, risks, and appropriate timing of surgery requiring anesthesia that will take longer than three hours. However, the degree of risk remains unclear. The statements in the FDA warning that the children at risk are those who are under three years or in utero, and are exposed for more than three hours, are largely based on animal data and need to be further verified in clinical studies. Due to the lack of clinical evidence, in April 2017, the FDA modified their warning, stating that medically necessary procedures in pregnant women and children under age three should not be delayed and that practitioners should follow their usual practice paradigms [71].
In response to the FDA warning both the Society for Pediatric Anesthesia (SPA) [72] and the American College of Obstetrics and Gynecology (ACOG) [73] responded with statements of concern that necessary surgery and procedures requiring anesthesia might be delayed or avoided because of uncertain risks of deleterious effects of anesthesia.
WHAT WE TELL PARENTS AND PATIENTS —
When asked about the risks of anesthesia by parents, patients, and care providers, clinicians should state that while children who need surgery and anesthesia have worse scores in some neurodevelopmental outcomes compared with children who do not need surgery and anesthesia, these differences cannot yet be attributed to anesthesia exposure, and there is still uncertainty that exposure to anesthesia has any lasting effects in children or fetuses [74].
Some institutions and clinicians have elected to routinely include discussion of the potential risks of neurotoxicity of anesthetics as part of informed consent [75] and there are resources available for those interested in doing so, including information on the website for the Society for Pediatric Anesthesia.
SOCIETY GUIDELINE LINKS —
Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Pediatric anesthesia".)
SUMMARY AND RECOMMENDATIONS
●State of the literature
•The risk of adverse neurodevelopmental effects from anesthetic agents in children remains uncertain. Many clinical studies including meta-analyses have reported that compared to unexposed children, children with exposure to general anesthesia have more adverse neurodevelopmental outcomes, particularly in specific domains such as behavior and executive function. However, the differences are small, particularly after a single exposure. Anesthetic exposure appears to have limited to no impact on the cognition domain as measured by intelligence quotient (IQ) testing. At this time there is inadequate evidence to directly attribute adverse effects to the anesthesia. (See 'Summary of the evidence in human studies' above.)
•Results from the only published randomized clinical trial to date suggest that a single anesthetic exposure has no effect on intelligence or on a range of other neurodevelopmental outcomes at age five years. (See 'Randomized trial' above.)
•A study based on randomization for repeated treatment requiring anesthesia prior to age five years versus standard care found similar neurodevelopmental testing and MRI studies in patients who had more treatments versus those who had less. This study suggests that repeated exposure to anesthesia in young children has no effect on neurodevelopmental outcomes. (See 'Repeated exposure to anesthesia' above.)
•A number of observational studies have reported higher levels of learning disability and neuropsychiatric diagnoses in children with multiple exposures relative to a single exposure, but this may be confounded by the children’s underlying disease. (See 'Limitations of the literature' above.)
•Some observational studies have reported an association between prenatal general anesthetic exposure and subsequent abnormal behavioral scores or increased risk of neuropsychiatric diagnoses. However, potential confounders may be present. Given the risk of complications to mothers and children, necessary surgery during pregnancy should not be avoided or delayed. (See 'Prenatal exposure to anesthesia' above and "Anesthesia for nonobstetric surgery during pregnancy", section on 'Effects of anesthetics on the fetus and the pregnancy'.)
●What we tell parents – When asked about the risks of anesthesia, clinicians should state that while children who need surgery and anesthesia have worse scores in some neurodevelopmental testing compared with children who do not need surgery and anesthesia, these differences cannot yet be attributed to anesthesia exposure, and there is still uncertainty that exposure to anesthesia has any lasting effects in children or fetus. (See 'What we tell parents and patients' above.)
●Practice implications – The potential risks of exposure to anesthesia must be weighed against the benefits of the surgical or diagnostic procedure. Given the uncertainty that any long-term effects in children are due to the anesthetic medications, and the potential for complications if necessary surgery is avoided or delayed, changes in clinical management should not be made based on the potential risk of neurotoxicity of anesthesia. (See 'The FDA warning' above.)
