Continuous and accurate assessment of intravascular volume status in patients undergoing surgical procedures with the potential for massive hemorrhage is vital for facilitating the timely replacement of blood loss and maintaining stable hemodynamics. The utility of central venous pressure (CVP) monitoring (CVP) for this purpose in infants and young children is controversial and not well studied. Children undergoing major craniofacial reconstruction surgery commonly have intraoperative blood loss exceeding the circulating blood volume,1,2 and therefore comprise an excellent population in which to evaluate different modalities for monitoring intravascular volume status. We recently conducted a review of the perioperative course of children undergoing craniofacial procedures at our institution, where we found hypotension to be relatively common, occurring in 17% of patients.3 During the time period of this review, intravascular volume status was assessed through arterial blood pressure measurement combined with the qualitative assessment of the invasive arterial blood pressure waveform. In response to our findings, we began routinely monitoring CVP in these children in an effort to improve our assessment of intravascular volume status and to prevent hypotension. This practice change was implemented in March 2008 by our pediatric craniofacial anesthesia team; subsequent central line insertion was not influenced by which surgeon was operating.
We conducted the present study to assess the effect of this practice change on the incidence and duration of hypotension in infants and children undergoing craniofacial reconstruction surgery. The secondary aim of this study was to assess the relationship between CVP and hypotension in this population.
Effect of CVP Monitoring on the Incidence and Duration of Hypotension
To assess the effect of CVP monitoring on the incidence and duration of hypotension, perioperative data from a historical cohort were compared with matching data from a cohort of subjects with CVP monitoring. Our primary hypothesis was that the use of CVP monitoring would be associated with a reduction in the total duration of hypotension in subjects who experienced hypotension. Our secondary hypothesis was that the incidence of subjects having at least 1 episode of hypotension would be reduced with the use of CVP monitoring.
The historical cohort subjects in the present study are a subset of subjects on whom we reported in a previous publication.3 Of the 159 subjects from this previous study, the historical cohort included subjects who were 6 to 24 months of age who underwent fronto-orbital advancement or posterior cranial vault reconstruction between December 1, 2001, and January 1, 2006. This age group was chosen for study because the majority of cranial vault reconstructions are performed at these ages. Data for the CVP cohort were obtained through a query of our prospective craniofacial surgery perioperative registry between April 14, 2008, and July 1, 2011, for subjects who had CVP monitoring and who met the same age and surgical procedure criteria as in the historical cohort. A previous publication from our group examined heart rate behavior during hypotension in this same group of subjects from our prospective registry.4 IRB approval was obtained for data collection and research use of data for the historical cohort3 and the CVP cohort.4
Data were collected through review of the electronic anesthesia record (Philips Healthcare, Bothell, WA), patient chart, intensive care unit (ICU) flow sheet, progress notes, and discharge summary. Collected data included age, weight, gender, diagnosis, surgical procedure, history of previous craniofacial surgery, presence of a craniosynostosis syndrome, and all fluids and blood products administered intraoperatively. Preoperative, intraoperative, and postoperative laboratory values were also recorded. For each hypotensive episode in the CVP cohort, the expired volatile anesthetic agent concentration at the onset of hypotension was recorded from the electronic anesthetic record. In both study groups, intraoperative blood loss was calculated in the manner described by Kearney et al.3,5,6
Episodes of hypotension were identified through examination of intraoperative invasive arterial blood pressure data captured in 15-second intervals from stored electronic anesthesia records. For the purposes of this study, an episode of hypotension was defined as occurring when the mean arterial blood pressure (MAP) was <40 mm Hg for at least 3 consecutive entries (captured at 15-second intervals). The end of an episode was defined as the last time point at which the MAP was <40 mm Hg before increasing to ≥40 mm Hg for at least 2 consecutive entries.
Demographic data, including age, weight, gender, diagnosis, and surgical procedure, were compared between the 2 study cohorts. The calculated blood loss for all subjects and for subjects with hypotension in the 2 study groups was compared. The total duration of hypotension in the hypotensive subjects was compared in the 2 study groups. The incidence of subjects who had at least 1 episode of hypotension was compared in the 2 cohorts.
Assessment of the Relationship Between CVP and Hypotension in the CVP Cohort
To assess the relationship of CVP to hypotension in the CVP cohort, CVP at the onset of hypotension (when the MAP first decreased below 40 mm Hg) was compared with CVP 5 minutes before and with CVP 5 minutes after this time point. We also compared CVP at the onset of hypotension to the baseline CVP. We hypothesized that CVP values at the onset of hypotension would be significantly decreased relative to baseline CVP and CVP 5 minutes before and after the onset of hypotension. Lastly, we assessed the amount of time spent at various CVP levels relative to the baseline CVP and determined the percentage of time that was associated with hypotension. A description of how these analyses were performed follows.
For each subject in the CVP cohort, intraoperative vital signs were extracted from the stored electronic anesthesia record. Extracted data included the invasive systolic blood pressure, MAP, and CVP. These data were extracted over the time period starting 10 minutes after scalp infiltration with epinephrine until 20 minutes before the recorded end of surgery, as was done in our previous study of this group of subjects (Fig. 1).4 The time points delimiting this studied intraoperative interval were selected to capture the period during which the craniofacial operation was being performed.4 The extracted vital signs for each subject were graphed. As previously described, visual inspection of the graphs was performed by 2 investigators (PS, EL) to identify and remove artifact using strict criteria.4 After removal of artifact data, the graphs were inspected to identify all episodes of hypotension.
For the analysis of the relationship between CVP and hypotension, the baseline CVP was defined as the median CVP over the first 2 minutes of the studied intraoperative interval (Fig. 1). For each hypotensive episode, the median CVP over 1-minute epochs was calculated at the onset of hypotension (T0, starting when the MAP first decreased below 40 mm Hg) and 5 minutes before (T – 5) and 5 minutes after (T + 5) the start of the episode (Fig. 2). One-minute median values were used to account for natural variability present in the raw data. A median value was used to represent the central tendency of 4 CVP integer data points each minute rather than a mean value, because median values are more clinically meaningful, as medians of 4 CVP integers occur in increments of 0.5 mm Hg, rather than continuous real numbers that would result from the use of means. The 1-minute median CVP values were compared at T0, T – 5, and T + 5. We also compared CVP at T0 with the baseline CVP. Comparisons between CVP at T0 and CVP at T – 5 or T + 5 were excluded if there was overlap with another episode of hypotension (i.e., the subject was hypotensive at this time point). For the analysis of CVP at T0 compared with the baseline CVP, all episodes of hypotension were included.
We then assessed the amount of time spent at various CVP levels relative to the baseline CVP over the entire studied interval for each subject. For this analysis, 1-minute median CVP data over the entire studied interval in each subject were analyzed to determine how many minutes were spent at different CVP levels below the baseline CVP. The total number of studied intraoperative minutes was tabulated for all subjects together with the total time spent at various CVP levels below baseline, as well as the total time that there was hypotension at each of these CVP levels. Cases were excluded from this analysis if the baseline CVP could not be determined or if the baseline CVP was >15 mm Hg.
Conduct of Anesthesia and Surgery
All subjects had general endotracheal anesthesia with standard American Society of Anesthesiologists monitoring. Mask induction of anesthesia was performed with sevoflurane, nitrous oxide, and oxygen. Each subject had 2 peripheral IV catheters and a radial arterial catheter placed. In the CVP cohort, central venous lines were placed in the internal jugular vein under ultrasound guidance followed by radiographic confirmation of tip position in the superior vena cava or at the cavoatrial junction. The CVP transducer was secured to the operating table and was positioned with the zero reference at the level of the anterior midaxillary line. Normothermia was facilitated by a circulating warm water blanket, a forced air warmer, and an IV fluid warmer. Before skin incision, the scalp was infiltrated with epinephrine 1:400,000 up to a maximum of 4 mL/kg. Anesthetic maintenance was with isoflurane, sevoflurane, or desflurane in the air and oxygen. Intraoperatively, fentanyl or morphine was administered for anesthetic supplementation and postoperative analgesia. Anesthetic management, fluid management, and blood loss replacement were at the discretion of anesthesia providers without a fixed protocol. In the historical cohort, blood loss was replaced with crystalloid, colloid (albumin), and packed red blood cells (PRBCs). Hemostatic blood products (e.g., fresh frozen plasma [FFP], platelets) were administered as judged necessary by anesthesia providers. In the CVP cohort, blood loss was replaced with crystalloid and reconstituted blood (made by combining FFP and PRBCs). Additional hemostatic blood products were administered as judged necessary by anesthesia providers. The antifibrinolytic epsilon-aminocaproic acid was administered to 11 subjects in the CVP cohort. Additional techniques to limit blood loss and transfusion such as preoperative erythropoietin administration, controlled hypotension, cell salvage, or acute normovolemic hemodilution were not used in either study group.
Craniofacial Surgery Perioperative Registry
Enrollment in the registry began on April 14, 2008. Children of English-speaking parents having craniofacial reconstruction surgery involving a craniotomy were systematically approached for enrollment at the preoperative anesthesia visit. After we obtained written informed consent, enrolled subjects had demographic data and data pertaining to their perioperative course and management entered into a deidentified computer database (Microsoft SQL Server 2000). A research assistant collected data through intraoperative questioning of anesthesia providers and by inspection of the electronic anesthesia record, the ICU flow sheet, and progress notes. Data collected prospectively through intraoperative questioning of providers included an assessment of intraoperative complications such as whether a venous air embolism was suspected to have occurred.
Data were stored using Excel (Microsoft Corp, Redmond, WA) and were analyzed using JMP 9 (SAS Institute Inc., Chesterbrook, PA) and SAS version 9.2 (SAS Institute Inc., Cary, NC). Histogram plots were used to check the normality of age, weight, preoperative hematocrit and surgery duration. The values of weight, preoperative hematocrit, and surgery duration were normally distributed; age was not normally distributed. χ2 or Fisher exact test was used for categorical data and the Wilcoxon rank sum test or t test was used for comparisons with continuous data. To account for multiple episodes of hypotension that occurred in some subjects, a linear mixed effects model was used to calculate the 95% prediction intervals for the change in CVP between the study time points. A separate analysis using the linear mixed effects model was also fitted to assess the relationship between the severity of hypotension and the magnitude of the change in CVP from T – 5 to T0. The severity of hypotension was defined as the magnitude of the difference between the nadir MAP and 40 mm Hg. The nadir MAP was defined as the lowest MAP recorded during a hypotensive episode. A P value of <0.05 was considered to be significant.
For the comparisons between the historical cohort and the CVP cohort, we treated our data as if patients were randomized to study group. A propensity score analysis was performed to assess the balance of demographic covariates between the 2 study cohorts. We defined the propensity score as the probability of a subject being assigned to the historical cohort. We estimated the propensity score for each individual using multivariable logistic regression with the group assignment (CVP versus historical) as the outcome of interest. Age, weight, gender, surgery duration, preoperative hematocrit, diagnosis, surgical procedure, and primary versus secondary operation were included as independent variables. The propensity score distribution was then used to group subjects into 5 quartiles according to the probability of being assigned to the historical cohort. We used t test and χ2 test to determine if the covariates were balanced within the quartile and among the entire sample.
The application of study inclusion criteria to our historical dataset yielded a historical cohort with data from 115 procedures performed on 105 subjects. The registry query yielded a CVP cohort with data from 57 procedures performed on 53 subjects. Demographic data from these 2 cohorts are presented and compared in Table 1. The mean (SD) propensity score was 0.67 (0.06) for the entire sample [CVP: 0.658 (0.063); historical: 0.674 (0.058)]. Two quartiles were identified among 172 subjects. Nineteen subjects (8 CVP and 11 historical) had a propensity score ranging from 0.4 to 0.6 in quartile 1. One hundred fifty-three subjects (49 CVP and 104 historical) had a propensity score ranging from 0.6 to 0.8 in quartile 2. All covariates were not statistically different between CVP and historical subjects within each quartile (all P > 0.34; all standardized difference ≤ 0.56). In addition, the mean propensity score was not different for CVP and historical subjects in each quartile and among the entire sample. This suggests that the covariates were balanced within the quartile and among the entire sample between the CVP and historical cohorts.
In the assessment of the effect of CVP monitoring on hypotension, the incidence of hypotension was 18% in the CVP cohort versus 21% in the historical cohort; the difference in the incidence of hypotension was –3% (95% confidence interval, –10% to 15%, Table 2). The 95% confidence intervals for the differences in calculated blood loss between the 2 study cohorts comparing all subjects and subjects experiencing hypotension are shown in Table 2.
There were 29 episodes of hypotension occurring in 10 subjects in the CVP cohort. The median baseline CVP in these 10 subjects was 7 mm Hg (range: 3–9 mm Hg). The median end-tidal expired concentration of volatile anesthetic at the onset of hypotension was 1.9 vol% sevoflurane (range: 1.6–2.9 vol%), 5.9 vol% desflurane (range: 3.6–7.8 vol%), and 1.0 vol% isoflurane (range: 0.9–1.3 vol%). Using age-specific minimal alveolar concentration (MAC) values of 9.9 vol% desflurane in 6 to 12 month olds,7 8.7 vol% desflurane in 12 to 24 month olds,7 2.5 vol% sevoflurane,8 and 1.8 vol% isoflurane,9 the median expired anesthetic dose at the time of hypotension was 0.7 MAC (range: 0.4–1.2 MAC). None of the episodes of hypotension in the CVP cohort was preceded by an IV opioid bolus within 10 minutes of the onset of hypotension. The 1-minute median CVP at the onset of hypotension was less than the baseline CVP for 28 of 29 episodes (97% of the episodes). Raw hypotensive episode data are presented in Table 3.
For the analysis comparing CVP at the onset of hypotension (T0) with CVP 5 minutes before (T – 5) and 5 minutes after (T + 5), 5 of the T – 5 data points and 5 of the T + 5 data points were excluded because of overlap with another episode of hypotension. CVP data were not recorded in the electronic anesthesia record at T + 5 for 1 episode (episode 28, Table 3), leaving 24 and 23 data comparisons in each group, respectively, for analysis. Analysis using a linear mixed effects model showed a significant decrease in CVP from T – 5 to T0, a significant increase in CVP from T0 to T + 5, no significant difference in CVP between T – 5 and T + 5, and a significant decrease in CVP from baseline to T0. The estimated change in CVP between these study time points together with the 95% prediction intervals are shown in Table 4. The change in CVP from T – 5 to T0 was significantly associated with the severity of hypotension (P = 0.002). The model estimated that CVP from T – 5 to T0 decreased 0.25 mm Hg (SE = 0.066) for every 1 mm Hg decrease in MAP below 40 mm Hg.
We then evaluated the total time spent by all evaluable subjects at or below CVP levels ranging from 1 to ≥5 mm Hg below the baseline CVP and calculated the percentage of the total studied intraoperative period for all subjects. The associated incidence of hypotension when CVP was at or below each of the levels from 1 to ≥5 was calculated (Table 5). We also determined the total amount of time spent at specific CVP levels below the baseline together with the associated incidence of hypotension at each specific CVP level (Table 6). Lastly, we examined CVP relative to the baseline during the studied interval for all cases and determined the median and the 2.5th, 25th, 75th, and 97.5th quantiles (Table 7). For each of the above assessments, 1 case was excluded because the baseline CVP could not be determined because CVP was steadily increasing at the start of the studied interval, and 3 cases were excluded because baseline CVP was >15 mm Hg. There were no episodes of hypotension that occurred in any of the excluded cases.
Massive hemorrhage during craniofacial surgery is common and often results in hypovolemia.5,10–12 Monitoring of CVP has been advocated as a tool for intravascular volume status assessment in this population.13 We began monitoring CVP in these children routinely, after identifying hypotension to be a common problem in our practice. Before this, we assessed volume status in these children through a combination of invasive arterial blood pressure measurement and qualitative assessment of the invasive arterial waveform. In our evaluation of the effect of this practice change, we found that neither the incidence nor the duration of hypotension was decreased. Although there was no specific protocol in this study for how CVP was used to direct fluid management, our data represent the effect of this monitoring tool that was specifically placed to aid intravascular volume assessment as it was used for this purpose by subspecialty trained pediatric anesthesiologists.
In the infants and children in this study experiencing massive blood loss, hypotension was associated with a decrease in CVP, whereas resolution of hypotension was associated with an increase in CVP to prehypotensive levels. In the majority of cases (97%), CVP at the onset of hypotension was less than the baseline CVP. However, although hypotension was consistently associated with a decrease in CVP, the reverse was not true. As an illustration of this, the median decrease in CVP at the onset of hypotension was 3 mm Hg lower than the baseline CVP. Of the 9,773 minutes that were studied, CVP was 3 mm Hg or more below the baseline for 1,583 minutes (16% of the total time), with an associated hypotension incidence of only 2% (Table 5). These findings suggest that CVP monitoring (as it was used by the practitioners in cases evaluated in this study) was not a useful monitor of intravascular volume status, because CVP was often below the baseline without associated hypotension. It should be noted that anesthesiologists may have used the trend in CVP rather than absolute values to guide fluid therapy. Nonetheless, our finding that neither the incidence nor the duration of hypotension was improved calls into question the effectiveness of this hemodynamic monitor.
The use of CVP monitoring in craniofacial surgery is not a routine practice at most institutions. A recent practice survey revealed that among centers where complex cranial vault reconstructions are performed, the majority of centers monitor CVP in <10% of patients undergoing these procedures, whereas only 15% of centers monitor CVP in nearly all of their patients.14 Some authors have commented subjectively on the value of the information from CVP monitoring in this population, suggesting that its use may help avoid hypotension and hypovolemia.13 There is currently little if any published evidence that CVP monitoring in children is an effective monitor of intravascular volume status that is helpful in preventing hypovolemia. The results of the present study show that routine use of this monitor was not associated with a decreased incidence or duration of hypotension in children undergoing craniofacial surgery. However, because our study was retrospective and fluid management was not standardized, a controlled prospective study would be needed to assess whether the incidence and severity of hypotension could be meaningfully decreased by systematically directing fluid therapy on the basis of CVP.
Perioperative cardiac arrest in children is most commonly attributed to cardiovascular causes, and in these cases, hypovolemia from blood loss is the most common etiology of arrest.15 In the most recent report from the Pediatric Perioperative Cardiac Arrest registry, the absence of CVP monitoring was identified as a contributing factor in 22% of children who suffered intraoperative cardiac arrest as a result of hemorrhagic hypovolemia.15 However, the conclusion of 1 review recommends against the use of CVP monitoring for guiding fluid therapy decision making.16 All of the 24 studies included in this review were conducted on adults, and nearly all of the studies involved ICU patients, coronary artery bypass patients, vascular surgical patients, or septic patients.16 Much attention has been focused on the use of dynamic variables to direct fluid therapy, such as plethysmographic waveform amplitude variability17,18 and changes in variables derived from the arterial waveform.19–22 Nearly all of the available evidence in adults suggests that these dynamic hemodynamic variables provide more predictive information for fluid responsiveness of hypotension than static hemodynamic variables such as CVP.22,23
Venous capacitance is large, and significant changes in total intravascular blood volume may only result in small changes in CVP.24 In our study, the magnitude of CVP change associated with hypotension was often small, both when compared with the baseline and when compared with CVP 5 minutes before hypotension. Most studies evaluating the performance of hemodynamic variables as monitors of intravascular fluid status have assessed whether a response to a fluid challenge can be predicted. These studies have generally involved hemodynamically stable patients, critically ill patients, or patients after cardiac surgery.19,25–28 Studies evaluating the performance of these variables in humans in the setting of actual hemorrhage and replacement are generally lacking. Most experiments of this nature have been done in animal models29–31; we are aware of 1 study in adult subjects that evaluated hemodynamic variables during controlled induction of hypovolemia with a diuretic,32 an experiment that will likely never be done on children.
It is interesting that some studies have shown that many of the hemodynamic variables that predict fluid responsiveness in adults do not predict fluid responsiveness in children,33,34 whereas other studies have shown conflicting results.35,36 Common to most of these studies, however, is the poor ability of CVP to predict fluid responsiveness. The results of our study seem to be consistent with this, supporting the notion that central line insertion should be reserved for children in whom adequate venous access is difficult or in those in whom central venous access is needed for medication administration.
Insertion of central venous lines in children is not without risk.37,38 The Pediatric Perioperative Cardiac Arrest registry report found that injury (vascular and lung) occurring during placement of central venous catheters was the most common cause of equipment-related cardiac arrest.15 The development of monitors that provide continuous automated quantitative dynamic hemodynamic variables and their validation as monitoring tools for intravascular volume status assessment in children may provide alternative monitoring modalities with less associated risks. Further study may enable evidence-based decision making regarding the assessment of intravascular volume status in this population.
Although there were numerous instances of hypotension observed in this study, there were no apparent sequelae of these episodes. The long-term effects of relatively brief periods of hypotension in infants are unclear. Nevertheless, few would argue against directing efforts at avoiding hypovolemia and hypotension.
Although this study was retrospective, the fidelity and high frequency of data capture with our electronic anesthesia record-keeping system provide a rigorous dataset that is objective and without reporting bias.4 Another possible limitation is that our definition of baseline CVP may not have been appropriate. We chose CVP over the initial 2 minutes of the recording period because, at this point, fasting fluid deficits have been replaced and significant blood loss has not yet occurred; therefore this CVP likely represented euvolemia. As mentioned above, an important limitation is that there was no specific protocol for how CVP was used to direct fluid management. A prospective intervention in which fluid management is specifically targeted to maintain CVP at or above a certain value may yield different results.
There were differences in what fluids were used to replace blood loss in the 2 groups. In the historical cohort, blood loss was replaced primarily with crystalloid and PRBCs, whereas in the CVP cohort, blood loss was replaced using reconstituted blood made by combining FFP and PRBCs.2 Because of this, we were unable to assess the effect of CVP monitoring on intraoperative fluid administration. We are unaware of any reason to believe that this difference in blood loss replacement would have had an effect on the incidence or duration of hypotension. The calculated blood loss variable considers these differences in how blood loss was replaced, and reveals that there were no statistically significant differences in blood loss between the 2 groups.
The definition of hypotension used in this study warrants discussion. There is no standardized definition of hypotension in children.39 Although safe limits for hypotension in pediatric patients are not well defined, it is doubtful that clinicians would argue that a MAP <40 mm Hg need not trigger a response. Others have specifically described anesthetic management in this population to maintain MAP above 45 mm Hg.6 Also, we have previously reported on the incidence of hypotension in this surgical population using this definition.3
It is important to note that hypotension may have been from causes other than hypovolemia, such as venous air embolism,40 anesthetic overdose, or monitoring artifact. During prospective registry data collection, whether a suspected venous air embolism occurred was systematically assessed, and in no case was this identified. Anesthetic overdose seems unlikely based on the anesthetic concentrations that were present at the onset of hypotension and the absence of bolus opioid administration before any hypotensive episode. Monitoring artifact was limited by careful examination of the extracted data. One-minute medians were calculated to account for the natural variability in CVP present in the raw data.
Finally, some occurrences of hypotension in this population may be unavoidable, such as what might occur during massive precipitous hemorrhage associated with a dural venous sinus tear. Hypotension in these situations will occur regardless of what monitoring modality is used when blood loss replacement cannot keep pace with losses.
The implementation of routine CVP monitoring was not associated with a decreased incidence and likely was not associated with a decreased duration of hypotension in this population experiencing massive hemorrhage. Hypotension was associated with a decrease in CVP, and resolution of hypotension was associated with an increase in CVP to prehypotensive levels. In 97% of hypotensive episodes, CVP at the onset of hypotension was less than the baseline CVP. One interpretation of these results is that maintaining CVP at or above baseline would prevent hypotension. However, decreases in CVP below baseline were common and infrequently associated with hypotension, suggesting that fluid management directed on the basis of CVP in this way might result in excessive fluid administration. Taken together, these results suggest that the routine use of CVP monitoring is of questionable utility as a monitor of intravascular volume status in this population.
Name: Paul A. Stricker, MD.
Contribution: This author helped design the study, conduct the study, collect the data, analyze the data, and prepare the manuscript.
Name: Elaina E. Lin, MD.
Contribution: This author helped conduct the study, analyze the data, and prepare the manuscript.
Name: John E. Fiadjoe, MD.
Contribution: This author helped design the study, conduct the study, and prepare the manuscript.
Name: Emily M. Sussman, BA.
Contribution: This author helped conduct the study.
Name: Eric Y. Pruitt, BA.
Contribution: This author helped conduct the study.
Name: Huaqing Zhao, PhD.
Contribution: This author helped analyze the data and prepare the manuscript.
Name: David R. Jobes, MD.
Contribution: This author helped design the study, analyze the data, and prepare the manuscript.
This manuscript was handled by: Peter J. Davis, MD.
1. White N, Marcus R, Dover S, Solanki G, Nishikawa H, Millar C, Carver ED. Predictors of blood loss in fronto-orbital advancement and remodeling. J Craniofac Surg. 2009;20:378–81
2. Stricker PA, Fiadjoe JE, Davis AR, Sussman E, Burgess BJ, Ciampa B, Mendelsohn J, Bartlett SP, Sesok-Pizzini DA, Jobes DR. Reconstituted blood reduces blood donor exposures in children undergoing craniofacial reconstruction surgery. Paediatr Anaesth. 2011;21:54–61
3. Stricker PA, Shaw TL, Desouza DG, Hernandez SV, Bartlett SP, Friedman DF, Sesok-Pizzini DA, Jobes DR. Blood loss, replacement, and associated morbidity in infants and children undergoing craniofacial surgery. Paediatr Anaesth. 2010;20:150–9
4. Stricker PA, Lin EE, Fiadjoe JE, Sussman EM, Jobes DR. Absence of tachycardia during hypotension in children undergoing craniofacial reconstruction surgery. Anesth Analg. 2012;115:139–46
5. Kearney RA, Rosales JK, Howes WJ. Craniosynostosis: an assessment of blood loss and transfusion practices. Can J Anaesth. 1989;36:473–7
6. Goobie SM, Meier PM, Pereira LM, McGowan FX, Prescilla RP, Scharp LA, Rogers GF, Proctor MR, Meara JG, Soriano SG, Zurakowski D, Sethna NF. Efficacy of tranexamic acid in pediatric craniosynostosis surgery: a double-blind, placebo-controlled trial. Anesthesiology. 2011;114:862–71
7. Taylor RH, Lerman J. Minimum alveolar concentration of desflurane and hemodynamic responses in neonates, infants, and children. Anesthesiology. 1991;75:975–9
8. Lerman J, Sikich N, Kleinman S, Yentis S. The pharmacology of sevoflurane in infants and children. Anesthesiology. 1994;80:814–24
9. Cameron CB, Robinson S, Gregory GA. The minimum anesthetic concentration of isoflurane in children. Anesth Analg. 1984;63:418–20
10. Tunçbilek G, Vargel I, Erdem A, Mavili ME, Benli K, Erk Y. Blood loss and transfusion rates during repair of craniofacial deformities. J Craniofac Surg. 2005;16:59–62
11. Meyer P, Renier D, Arnaud E, Jarreau MM, Charron B, Buy E, Buisson C, Barrier G. Blood loss during repair of craniosynostosis. Br J Anaesth. 1993;71:854–7
12. Stricker PA, Shaw TL, Desouza DG, Hernandez SV, Bartlett SP, Friedman DF, Sesok-Pizzini DA, Jobes DR. Blood loss, replacement, and associated morbidity in infants and children undergoing craniofacial surgery. Paediatr Anaesth. 2010;20:150–9
13. Carver E, Marcus R, Tatman A. FFP use in paediatric craniofacial surgery. Paediatr Anaesth. 2010;20:471
14. Stricker PA, Cladis FP, Fiadjoe JE, McCloskey JJ, Maxwell LG. Perioperative management of children undergoing craniofacial reconstruction surgery: a practice survey. Paediatr Anaesth. 2011;21:1026–35
15. Bhananker SM, Ramamoorthy C, Geiduschek JM, Posner KL, Domino KB, Haberkern CM, Campos JS, Morray JP. Anesthesia-related cardiac arrest in children: update from the Pediatric Perioperative Cardiac Arrest Registry. Anesth Analg. 2007;105:344–50
16. Marik PE, Baram M, Vahid B. Does central venous pressure predict fluid responsiveness? A systematic review of the literature and the tale of seven mares. Chest. 2008;134:172–8
17. Cannesson M, Attof Y, Rosamel P, Desebbe O, Joseph P, Metton O, Bastien O, Lehot JJ. Respiratory variations in pulse oximetry plethysmographic waveform amplitude to predict fluid responsiveness in the operating room. Anesthesiology. 2007;106:1105–11
18. Desebbe O, Cannesson M. Using ventilation-induced plethysmographic variations to optimize patient fluid status. Curr Opin Anaesthesiol. 2008;21:772–8
19. Perel A, Minkovich L, Preisman S, Abiad M, Segal E, Coriat P. Assessing fluid-responsiveness by a standardized ventilatory maneuver: the respiratory systolic variation test. Anesth Analg. 2005;100:942–5
20. Preisman S, Kogan S, Berkenstadt H, Perel A. Predicting fluid responsiveness in patients undergoing cardiac surgery: functional haemodynamic parameters including the Respiratory Systolic Variation Test and static preload indicators. Br J Anaesth. 2005;95:746–55
21. Michard F, Lopes MR, Auler JO Jr. Pulse pressure variation: beyond the fluid management of patients with shock. Crit Care. 2007;11:131
22. Marik PE, Cavallazzi R, Vasu T, Hirani A. Dynamic changes in arterial waveform derived variables and fluid responsiveness in mechanically ventilated patients: a systematic review of the literature. Crit Care Med. 2009;37:2642–7
23. Michard F. Changes in arterial pressure during mechanical ventilation. Anesthesiology. 2005;103:419–28 quiz 449–5
24. Gelman S. Venous function and central venous pressure: a physiologic story. Anesthesiology. 2008;108:735–48
25. Marik PE. The systolic blood pressure variation as an indicator of pulmonary capillary wedge pressure in ventilated patients. Anaesth Intensive Care. 1993;21:405–8
26. Zimmermann M, Feibicke T, Keyl C, Prasser C, Moritz S, Graf BM, Wiesenack C. Accuracy of stroke volume variation compared with pleth variability index to predict fluid responsiveness in mechanically ventilated patients undergoing major surgery. Eur J Anaesthesiol. 2010;27:555–61
27. Tavernier B, Makhotine O, Lebuffe G, Dupont J, Scherpereel P. Systolic pressure variation as a guide to fluid therapy in patients with sepsis-induced hypotension. Anesthesiology. 1998;89:1313–21
28. Kramer A, Zygun D, Hawes H, Easton P, Ferland A. Pulse pressure variation predicts fluid responsiveness following coronary artery bypass surgery. Chest. 2004;126:1563–8
29. Perel A, Pizov R, Cotev S. Systolic blood pressure variation is a sensitive indicator of hypovolemia in ventilated dogs subjected to graded hemorrhage. Anesthesiology. 1987;67:498–502
30. Berkenstadt H, Friedman Z, Preisman S, Keidan I, Livingstone D, Perel A. Pulse pressure and stroke volume variations during severe haemorrhage in ventilated dogs. Br J Anaesth. 2005;94:721–6
31. Fujita Y, Yamamoto T, Sano I, Yoshioka N, Hinenoya H. A comparison of changes in cardiac preload variables during graded hypovolemia and hypervolemia in mechanically ventilated dogs. Anesth Analg. 2004;99:1780–6, table of contents
32. Durga P, Jonnavittula N, Muthuchellappan R, Ramachandran G. Measurement of systolic pressure variation during graded volume loss using simple tools on Datex Ohmeda S/5 monitor. J Neurosurg Anesthesiol. 2009;21:161–4
33. Durand P, Chevret L, Essouri S, Haas V, Devictor D. Respiratory variations in aortic blood flow predict fluid responsiveness in ventilated children. Intensive Care Med. 2008;34:888–94
34. Pereira de Souza Neto E, Grousson S, Duflo F, Ducreux C, Joly H, Convert J, Mottolese C, Dailler F, Cannesson M. Predicting fluid responsiveness in mechanically ventilated children under general anaesthesia using dynamic parameters and transthoracic echocardiography. Br J Anaesth. 2011;106:856–64
35. Renner J, Broch O, Gruenewald M, Scheewe J, Francksen H, Jung O, Steinfath M, Bein B. Non-invasive prediction of fluid responsiveness in infants using pleth variability index. Anaesthesia. 2011;66:582–9
36. Renner J, Broch O, Duetschke P, Scheewe J, Höcker J, Moseby M, Jung O, Bein B. Prediction of fluid responsiveness in infants and neonates undergoing congenital heart surgery. Br J Anaesth. 2012;108:108–15
37. Casado-Flores J, Barja J, Martino R, Serrano A, Valdivielso A. Complications of central venous catheterization in critically ill children. Pediatr Crit Care Med. 2001;2:57–62
38. Karapinar B, Cura A. Complications of central venous catheterization in critically ill children. Pediatr Int. 2007;49:593–9
39. Nafiu OO, Voepel-Lewis T, Morris M, Chimbira WT, Malviya S, Reynolds PI, Tremper KK. How do pediatric anesthesiologists define intraoperative hypotension? Paediatr Anaesth. 2009;19:1048–53
40. Faberowski LW, Black S, Mickle JP. Incidence of venous air embolism during craniectomy for craniosynostosis repair. Anesthesiology. 2000;92:20–3