The measurement and interpretation of blood pressure (BP) in the perioperative period is a basic and fundamental practice for quality perioperative care. Indeed, the measurement of BP is required at least every 5 min, as a basic standard for monitoring during anesthesia [1]. For a historical perspective, the first measurements of BP during anesthesia were pioneered by Harvey Cushing in the early 1900s and gained widespread use in the 1920s. Yet, it wasn’t until the 1970’s when BP could be more reliably measured in young children, particularly neonates and infants [2].

Children present unique considerations and challenges in the measurement and interpretation of BP intraoperatively due to their varying anatomical size and developing physiology. While studies of complications of intraoperative BP perturbations are numerous in adults, much less is known about how to measure, interpret and treat BP in children. This narrative review will investigate how hypotension is defined in anesthetized children, describe the challenges in noninvasive automated blood pressure (NIBP) monitoring, identify complications of intraoperative hypotension, and discuss how to individualize BP management strategies in children under anesthesia.

One of the most challenging aspects in pediatric anesthesia practice is how to define hypotension in anesthetized children. The Pediatric Advanced Life Support Guidelines from the American Heart Association serve as a useful reference for accepted blood pressure ranges in awake children [3]. Table 1 shows accepted ranges of age-based blood pressure measurements in children. However, these definitions are based on awake patients as a means to recognize shock and consequently, may not have broad applicability to children under anesthesia.

In 2009, a survey study by Nafiu and colleagues [4] specifically examined how pediatric anesthesiologists define intraoperative hypotension and revealed substantial variability in the thresholds utilized for hypotension in pediatric patients. Participants from the United States and Great Britain were surveyed. Among 483 respondents (response rate of 56%), the majority reported they utilized mean BP (86.7%) or systolic blood pressure (72%) as the parameters to define hypotension. Additionally, the vast majority of respondents indicated that a reduction of systolic blood pressure (SBP) of 20% to 30% below the patient’s baseline blood pressure defines hypotension. Interestingly, when queried about SBP thresholds to define hypotension, there were regional differences. Anesthesiologists in the United States utilized SBP cutoffs to define hypotension that were approximately 5% to 7% lower across all age groups compared to practitioners in Great Britain.

As stated, many anesthesia practitioners will utilize a percent decrease from the “baseline” BP to define hypotension in anesthetized children. However, obtaining an accurate baseline BP in fasted, tired, hungry, and anxious children in the preoperative area is often difficult, especially without a reliable preoperative measurement. This was highlighted by a 2016 study which examined observed BPs in children 0-5 years of age over a 10-year period [5]. The study included 30,008 anesthetics with intraoperative BP measurements. Yet only 898 anesthetics (3%) had any BP data prior to induction, making the definition of hypotension as a reduction from baseline very limited. This was further supported by a prospective observational study in 2017, examining the ability to measure NIBP preoperatively in awake children just prior to a diagnostic or surgical procedure [6]. In 15% of patients, measurement of an awake BP was not possible, leading the authors to conclude that defining hypotension under anesthesia should be based on absolute values and not relative to awake baseline values.

Recently, retrospective data sets have been analyzed to report reference values for NIBP in children under anesthesia as well as neonates in the first week of life. In 2016, De Graaff and colleagues examined data from 9 centers in the United States and 1 center in the Netherlands via the Multicenter Perioperative Outcomes Group [7]. Healthy children, defined as American Society of Anesthesiologists physical status 1-2, from 0-18 years of age undergoing non-cardiac surgery were included. Two phases of surgery were analyzed; (1) the preparation phase, which occurred after induction and within 20 min before procedure start and (2) the initial surgical phase, where data was collected 15-35 min after procedure start. At least 2 artifact-free measurements in each phase were analyzed. The study reported sex-specific percentiles of NIBP values for age and weight for 116,000 cases analyzed, serving as an extremely useful reference of current intraoperative NIBP values among a large cohort of healthy children.

In 2023, a single-center study in the Netherlands reported NIBP reference values during the first week of life for premature and term neonates [8]. Median non-invasive SBP, mean BP, and diastolic blood pressure (DBP) values and reference ranges for gestational age and postnatal age were reported, with exclusion of neonates with severe comorbidities. Surprisingly, mean BP values in extreme preterm neonates increased abruptly during the first day after birth but quickly plateaued during the first week of life. Uniquely, neonates born closer to term had more gradual increases in mean BP during the first week of life. These data and references ranges provide valuable knowledge into baseline NIBP values in neonates and offer enhanced guidance in neonatal anesthetic management.

While both studies [7,8] employ sound methodology and provide useful reference ranges, they also have notable limitations. They do not provide any correlation or assessment of outcomes. Specifically, there is no evaluation of adequate end-organ perfusion, or association with any adverse outcomes, for a given reference range of blood pressure. As a result, one must be cautious about using the references ranges as a guide for clinical management. However, these references ranges provide a valuable resource to further study adequate blood pressure management in anesthetized children.

The most common method by which BP is measured under anesthesia is via the NIBP cuff. However, there are limitations in its accuracy of the NIBP results when compared to the direct arterial pressure measured by an invasive arterial line (ART-line), which is considered the gold standard. These limitations may be magnified during times of blood pressure perturbations. Fig. 1 demonstrates the discrepancy in the measurements recorded by a functioning direct ART-line compared to a simultaneous measurement by a NIBP in an infant.

The agreement between NIBP and ART-line readings was studied by Beringer and colleagues utilizing a multi-center prospective observational study of children under 16 years undergoing cardiac catheterization with general anesthesia [9]. The authors analyzed 683 paired and simultaneously recorded invasive ART-line and NIBP readings in 254 children. The overall bias (standard deviation) for mean blood pressure was 7.2 mmHg (11.4). During periods of hypotension, the bias increased to 15 mmHg (11.0). The Bland-Altman plot showing difference and poor correlation in mean BP for all patients is shown in Fig. 2. There were also recurrent differences in reported BPs by age. In infants, the mean BP was higher when measured by NIBP compared to direct ART-line measurements. Conversely, these values decreased in older children.

Further, Holt and colleagues in 2011 studied 40 children in the pediatric intensive care unit and compared SBP from 3 different measurement methods: (1) invasive ART-line, (2) from an automated NIBP, and (3) manual BP cuff with Doppler [10]. They found that automated NIBP readings overestimated SBP during hypotension by an average of 14 mmHg. The direct ART-line and manual BP cuff with doppler measurements showed very close agreement, while the automated NIBP measurements were more variable. Similar results were seen in neonates in a 2014 study from the neonatal intensive care unit examining 1,492 simultaneous BP measurements from direct ART-line (radial or umbilical ART-line) compared to NIBP [11]. In this study, NIBP consistently overestimates the SBP when compared to simultaneous direct arterial measurement, by a mean of 5-8 mmHg for SBP and 5 mmHg for mean BP.

The anatomical site of automated NIBP measurement can make a difference in the readings. In adults, when the NIBP cuff is placed on the leg compared to the arm, a higher BP reading is observed [12]. The opposite is seen in children up to 8 years old, where the BPs measured in the leg are lower than the arm, by a mean of 10 mmHg [13,14].

The exact reasons for the discrepancies and inaccuracies observed in NIBP readings are not entirely known, but likely have equipment and physiologic considerations. Appropriate-sized equipment must be used to achieve accurate results. Importantly in children, the cuff bladder width should be at least 40% of the upper arm circumference and cover 80-100% of the arm's circumference [15]. In addition, physiologic considerations, such as age and height-specific interpretation of BP values, the dynamic nature of BP changes during growth and development and possible differences in the stiffness of arterial structures between children and adults almost certainly complicate precise automated measurements [13]. Further, the individual software algorithms utilized in the NIBP equipment may have limited applicability in children. Many oscillometric NIBP devices have not been validated for pediatric use, leading to potential inaccuracies. The American Heart Association notes that few devices have undergone formal validation in children, which can result in unreliable readings [16]. Table 2 provides an overview of recommendations for measuring NIBP in children under anesthesia.

Recognition and treatment of hypotension is important to assure adequate end-organ perfusion, especially to the brain. While the brain is not the only organ in the body, it is arguably the most important, as inadequate perfusion to the brain can result in devastating irreversible neurologic injury and functional impairment, particularly in the rapidly developing brain in younger children. Indeed, the basic equation for cerebral perfusion pressure equals mean BP minus intracranial pressure (or central venous pressure, whichever is higher).

The dangers of inadequate cerebral perfusion were described in a 2014 case series [17]. Six infants, 4 of whom were former premature infants, ranging in size from 2.5 to 4 kg underwent elective procedures, including inguinal hernia repair (3 infants), mandibular distraction (1 infant), Kasai procedure (1 infant), and placement of a ventriculoperitoneal shunt (1 infant). Following the procedures, the infants were observed to be encephalopathic with seizures and found to have supratentorial watershed infarcts via brain magnetic resonance imaging. The outcomes were varied, including 1 infant who died, 2 had developmental delays, including 1 who had severe global delays, and the other with isolated motor delays. Two infants recovered normally. The presumed cause of these infarcts was due to intraoperative cerebral hypoperfusion.

This case series [17] suggests that a single definition of intraoperative hypotension in children may not be appropriate. In 5 of the 6 cases, the intraoperative blood pressures reported were at or above the limits suggested by Nafiu and colleagues [4] in the aforementioned survey study to define intraoperative hypotension. Further, considering the lower reference range for SBP in infants reported by de Graaff and colleagues [8] at approximately 30 mmHg, 5 of the 6 infants who developed encephalopathy were at or above this threshold for the majority of their case. This further emphasizes the need for better outcome-based and individualized blood pressure management.

In order to better understand individual variations in cerebral perfusion in anesthetized children, a 2003 study examined cerebral artery flow velocities measured by transcranial doppler ultrasound in 55 children, under general anesthesia with sevoflurane, who ranged in age from 6 months-14 years [18]. Surprisingly, the investigators found no difference in the lower limit of cerebral autoregulation between older children, defined as 6-14 years old and younger children, who were less than 6 years old. In fact, in a 9-month-old infant, the lower limit of autoregulation was 53 mmHg (Fig. 3). The study concluded that the baseline mean BP in young children may rest very close to the lower limits of cerebral autoregulation. Further work demonstrated that the lower limits of cerebral autoregulation in children under anesthesia may be widely variable, ranging anywhere from 30 all the way up to 55 mm [19]. Thus, this would support that there may not be a single “safe” BP cutoff for hypotension in children.

While blood pressure is a key factor in cerebral perfusion, it’s not the only consideration. There are other variables that affect cerebral perfusion including partial pressure of carbon dioxide (PaCO₂), oxygenation, glucose, and temperature [20]. Arguably, PaCO₂ is the most important of these and when hypotension is combined with hypocarbia, critically low cerebral perfusion may result.

With regard to monitoring cerebral perfusion as it relates to BP, the use of near-infrared spectroscopy (NIRS) monitoring showed mixed results in 2 conflicting studies.

The first study in 2015, examined 60 infants less than 3 months of age, undergoing non cardiac surgery [21]. The authors clearly showed that a decrease in SBP more than 20% from baseline was associated with cerebral desaturation. In a contrasting study of cerebral oxygenation, 453 infants less than 6 months undergoing general anesthesia for at least 30 min were studied with NIRS [22]. This study concluded that hypotension was common but not well associated with low cerebral saturation. The discrepancy between these studies may be attributed to additional factors that influence cerebral perfusion. Importantly, the first study[21] reported and controlled end-tidal CO₂ while the second study [22] did not, which further supports a dangerous situation when hypotension and hypocarbia are allowed to persist.

In adults, intraoperative hypotension is clearly associated with kidney injury and is directly related to mean BP below 65 mmHg and a decrease of 20% or more from baseline BP. In addition, prolonged hypotension further worsened the probability of acute kidney injury [23]. Contrary to adults, similar associations were not observed in children. In a retrospective study of about 4,500 children who underwent general anesthesia for noncardiac procedures lasting greater than 1 h who had perioperative BP and postoperative creatinine results available, the estimated probability of acute kidney injury did not change. Specifically, kidney injury was not associated with the lowest mean blood pressure or the largest reduction in mean BP [24].

The potential for general anesthetics to induce long-term adverse neurocognitive and behavioral effects in infants and children became an acute concern in the field of pediatric anesthesiology in recent years. Consequently, this phenomenon of potential anesthetic neurotoxicity has been the subject of much study and debate. While there have been multiple retrospective and prospective studies addressing neurotoxicity, these are not without criticism. In particular, that prior studies fail to consider some of the intraoperative variables that may affect cerebral perfusion, including hypotension [25,26].

In response, 2 major studies of anesthetic neurotoxicity in children further examined the impacts of BP on neurodevelopmental outcomes and found no association of adverse neurodevelopmental outcomes with hypotension. The General Anesthesia Compared to Awake Regional Anesthesia trial (GAS study) studied infants randomized to regional anesthesia or sevoflurane general anesthesia for elective inguinal herniorrhaphy and found no difference in their primary outcome of adverse neurodevelopmental outcomes [27]. A secondary aim was to evaluate BP data and hypotension [28]. Among 722 infants randomized to regional or general anesthesia, the investigators looked for any hypotension, which was defined as mean BP less than 45 mmHg as well as all periods of moderate hypotension, defined as a mean BP less than 35 mmHg. The risk of hypotension was 4.5 times greater with general anesthesia compared to regional anesthesia. Further, the individual risk factors for hypotension included general anesthesia, lower patient weight, and presence of hypothermia. Despite the greater incidence of hypotension in infants receiving general anesthesia, no differences in neurodevelopmental outcomes were observed.

The second major study of anesthetic neurotoxicity in children to examine the effects of blood pressure was the Mayo Anesthesia Safety in Kids Study (MASK) study [29]. In this study, retrospective and prospective cohorts were analyzed to assess for a wide variety of neurodevelopmental outcomes in children receiving general anesthesia before their 3rd birthday. In the retrospective cohort, those who had multiple exposures to general anesthesia had an increased frequency of both learning disabilities and attention deficit hyperactivity disorder [30]. In the prospective cohort, children who had multiple exposures to general anesthesia had decreased processing speed, decreased fine motor abilities and were more likely to be in the group of lowest performers on a wide battery of neuropsychiatric testing [29,31]. As part of the MASK study, the lowest 2 consecutive SBPs were also recorded. In a separate analysis, the BPs were compared to recently published reference ranges [8] , and a z-score was calculated [32]. Among 116 subjects with multiple exposures to anesthesia in the retrospective and 206 subjects in the prospective cohorts, the lowest SBPs were similar to published reference ranges, at a mean z-score of -0.26 and -0.62, respectively. When the association between lowest SBP (continuous variable) and intraoperative hypotension (dichotomous variable), defined as 1.0 Standard Deviation below expected SBP, were analyzed with outcomes by exposure category, there was no association of BP or hypotension with any adverse neurodevelopmental outcome.

What should the future hold? In addition to the need for better outcome studies on intraoperative BP management in children, future technology should bring improved monitoring. Ideally, the development of non-invasive, accurate and affordable monitors is needed to correctly detect decreases in end-organ perfusion, especially decreases in cerebral perfusion, which would allow a personalized definition of hypotension and prompt treatment by the anesthesiologist. As an example of developing work on improved monitoring, an experimental monitor was designed to report correlation coefficients as a way to detect the lower limit of cerebral autoregulation using laser-doppler flux and commercially available NIRS monitors [33]. In practice, when mean BP was within the individual cerebral autoregulation range, cerebral oximetry index and laser Doppler index approach a value of 0. However, when mean BP is below the lower limit of autoregulation, these monitors approach a value of 1, indicating that cerebral blood flow is pressure-passive and are directly related to blood pressure. Hopefully this technology will continue to evolve and become widely available.

The evaluation and management of BP in anesthetized children remains a complex and poorly defined process. Additional work needs to be done in order to better define hypotension and identify safe BP limits. While we have references ranges and other definitions for hypotension in anesthetized children, these are not linked to clinical outcomes and consequently, need to be interpreted with caution, particularly as they relate to cerebral perfusion. Complicating these issues are the wide individual factors, which likely require individualized approaches, based on the age, comorbidities and physiologic state of the patient. The future is ripe for improved studies of BP management in pediatric anesthesiology, especially with the development of improved monitoring capability.