Prediction of expiratory desflurane and sevoflurane concentrations in lung-healthy patients utilizing cardiac output and alveolar ventilation matched pharmacokinetic models: A comparative observational study.

ABSTRACT: The Gas Man simulation software provides an opportunity to teach, understand and examine the pharmacokinetics of volatile anesthetics. The primary aim of this study was to investigate the accuracy of a cardiac output and alveolar ventilation matched Gas Man model and to compare its predictive performance with the standard pharmacokinetic model using patient data.Therefore, patient data from volatile anesthesia were successively compared to simulated administration of desflurane and sevoflurane for the standard and a parameter-matched simulation model with modified alveolar ventilation and cardiac output. We calculated the root-mean-square deviation (RMSD) between measured and calculated induction, maintenance and elimination and the expiratory decrement times during emergence and recovery for the standard and the parameter-matched model.During induction, RMSDs for the standard Gas Man simulation model were higher than for the parameter-matched Gas Man simulation model [induction (desflurane), standard: 1.8 (0.4) % Atm, parameter-matched: 0.9 (0.5) % Atm., P = .001; induction (sevoflurane), standard: 1.2 (0.9) % Atm, parameter-matched: 0.4 (0.4) % Atm, P = .029]. During elimination, RMSDs for the standard Gas Man simulation model were higher than for the parameter-matched Gas Man simulation model [elimination (desflurane), standard: 0.7 (0.6) % Atm, parameter-matched: 0.2 (0.2) % Atm, P = .001; elimination (sevoflurane), standard: 0.7 (0.5) % Atm, parameter-matched: 0.2 (0.2) % Atm, P = .008]. The RMSDs during the maintenance of anesthesia and the expiratory decrement times during emergence and recovery showed no significant differences between the patient and simulated data for both simulation models.Gas Man simulation software predicts expiratory concentrations of desflurane and sevoflurane in humans with good accuracy, especially when compared to models for intravenous anesthetics. Enhancing the standard model by ventilation and hemodynamic input variables increases the predictive performance of the simulation model. In most patients and clinical scenarios, the predictive performance of the standard Gas Man simulation model will be high enough to estimate pharmacokinetics of desflurane and sevoflurane with appropriate accuracy.

Context-sensitive decrement times for inhaled anesthetics in obese patients explored with Gas Man®.

Anesthesia care providers and anesthesia decision support tools use mathematical pharmacokinetic models to control delivery and especially removal of anesthetics from the patient's body. However, these models are not able to reflect alterations in pharmacokinetics of volatile anesthetics caused by obesity. The primary aim of this study was to refine those models for obese patients. To investigate the effects of obesity on the elimination of desflurane, isoflurane and sevoflurane for various anesthesia durations, the Gas Man® computer simulation software was used. Four different models simulating patients with weights of 70 kg, 100 kg, 125 kg and 150 kg were constructed by increasing fat weight to the standard 70 kg model. For each modelled patient condition, the vaporizer was set to reach quickly and then maintain an alveolar concentration of 1.0 minimum alveolar concentration (MAC). Subsequently, the circuit was switched to an open (non-rebreathing) circuit model, the inspiratory anesthetic concentration was set to 0 and the time to the anesthetic decrements by 67% (awakening times), 90% (recovery times) and 95% (resolution times) in the vessel-rich tissue compartment including highly perfused tissue of the central nervous system were determined. Awakening times did not differ greatly between the simulation models. After volatile anesthesia with sevoflurane and isoflurane, awakening times were lower in the more obese simulation models. With increasing obesity, recovery and resolution times were higher. The additional adipose tissue in obese simulation models did not prolong awakening times and thus may act more like a sink for volatile anesthetics. The results of these simulations should be validated by comparing the elimination of volatile anesthetics in obese patients with data from our simulation models.

Linshom thermodynamic sensor is a reliable alternative to capnography for monitoring respiratory rate.

Monitoring ventilation accurately is a technically challenging, yet indispensable aspect of patient care in the intra- and post-procedural settings. A new prototypical device known as the Linshom Respiratory Monitoring Device (LRMD) has been recently designed to non-invasively, inexpensively, and portably measure respiratory rate. The purpose of this study was to measure the accuracy and variability of LRMD measurements of respiratory rate relative to the measurement of capnography. In this prospective study, participants were enrolled and individually fitted with a face mask monitored by the LRMD and capnography. With a baseline oxygen flow rate and digital metronome to pace their respiratory rate, the participants were instructed to breathe at 10 breaths per minute (bpm) for 3 min, 20 bpm for 3 min, 30 bpm for 3 min, 0 bpm for 30 s, and resume regular breathing for 30 s. Both sensors were connected to a computer for continuous temperature and carbon dioxide waveform recordings. The data were then retrospectively analyzed. Twenty-six healthy volunteers, mean (range) age 27.8 (23-37) and mean (range) BMI 23.1 (18.8-29.2) kg/m2 were recruited. There were 15 males (57.7%) and 11 females (42.3%). After excluding 3 subjects for technical reasons, 13,800 s of breathing and 4,140 expiratory breaths were recorded. Throughout the protocol, the average standard deviation (SD) for the LRMD and capnography was 1.11 and 1.81 bpm, respectively. The overall mean bias (±2SD) between LRMD and capnography was -0.33 (±0.1.56) bpm. At the lowest and intermediate breathing rates reflective of hypoventilation and normal ventilation, the LRMD variance was 0.55 and 1.23 respectively, compared to capnography with 5.54 and 7.47, respectively. At higher breathing rates indicative of hyperventilation, the variance of the test device was 4.52, still less than that of capnography at 5.73. This study demonstrated a promising correlation between the LRMD and capnography for use as a respiratory rate monitor. The LRMD technology may be a significant addition to monitoring vital signs because it offers a minimally intrusive opportunity to detect respiratory rate and apnea, without expensive or complex anesthetic equipment, before the need for life-saving resuscitation arises.

Incomplete Spontaneous Recovery from Airway Obstruction During Inhaled Anesthesia Induction: A Computational Simulation.

BACKGROUND: Inhaled induction with spontaneous respiration is a technique used for difficult airways. One of the proposed advantages is if airway patency is lost, the anesthetic agent will spontaneously redistribute until anesthetic depth is reduced and airway patency can be recovered. There are little and conflicting clinical or experimental data regarding the kinetics of this anesthetic technique. We used computer simulation to investigate this situation. METHODS: We used GasMan, a computer simulation of inhaled anesthetic kinetics. For each simulation, alveolar ventilation was initiated with a set anesthetic induction concentration. When the vessel-rich group level reached the simulation specified airway obstruction threshold, alveolar ventilation was set at 0 to simulate complete airway obstruction. The time until the vessel-rich group anesthetic level decreased below the airway obstruction threshold was designated time to spontaneous recovery. We varied the parameters for each simulation, exploring the use of sevoflurane and halothane, airway obstruction threshold from 0.5 to 2 minimum alveolar concentration (MAC), anesthetic induction concentration 2 to 4 MAC sevoflurane and 4 to 6 MAC halothane, cardiac output 2.5 to 10 L/min, functional residual capacity 1.5 to 3.5 L, and relative vessel-rich group perfusion 67% to 85%. RESULTS: In each simulation, there were 3 general phases: anesthetic wash-in, obstruction and overshoot, and then slow redistribution. During the first 2 phases, there was a large gradient between the alveolar and vessel-rich group. Alveolar do not reflect vessel-rich group anesthetic levels until the late third phase. Time to spontaneous recovery varied between 35 and 749 seconds for sevoflurane and 13 and 222 seconds for halothane depending on the simulation parameters. Halothane had a faster time to spontaneous recovery because of the lower alveolar gradient and less overshoot of the vessel-rich group, not faster redistribution. Higher airway obstruction thresholds, decreased anesthetic induction, and higher cardiac output reduced time to spontaneous recovery. To a lesser effect, decreased functional residual capacity and the decreased relative vessel-rich groups' perfusion also reduced the time to spontaneous recovery. CONCLUSIONS: Spontaneous recovery after complete airway obstruction during inhaled induction is plausible, but the recovery time is highly variable and depends on the clinical and physiologic situation. These results emphasize that induction is a non-steady-state situation, thus effect-site anesthetic levels should be modeled in future research, not alveolar concentration. Finally, this study provides an example of using computer simulation to explore situations that are difficult to investigate clinically.

Hypoventilation after inhaled anesthesia results in reanesthetization.

BACKGROUND: During emergence from volatile anesthesia, hypoventilation may result from many causes. In this study, we examined the effect of hypoventilation after initial emergence from volatile anesthesia and the potential for reanesthetization. METHODS: The uptake and excretion of desflurane (Des), sevoflurane, and isoflurane were studied using the Gas Man® computer simulation program for a 70-kg simulated patient. The vaporizer setting was adjusted so that a VRG (vessel-rich tissue group, including brain) level of 0.75 minimum alveolar concentration (MAC), 1.0 MAC, and 1.5 MAC was rapidly achieved and maintained within tight limits for a 1-, 2-, 4-, and 6-hour period of anesthesia.At the end of the simulated period of anesthesia, the vaporizer was set to 0 and fresh gas flow was set to 8 L/min. Ventilation (VA) was continued at 4 L/min until the anesthetic level in the VRG reached MAC awake, equal to 0.33 MAC for each drug. Then, the VA was adjusted to 0.1 L/min to simulate near-apnea and 0.0 L/min to simulate true apnea. Severe reanesthetization was said to occur if the VRG level increased to or above 0.5 MAC. Mild reanesthetization was said to occur if VRG increased from its value of 0.33 MAC but did not reach 0.5 MAC. The minimum VA required to avoid severe reanesthetization was studied by trials of decreased VA beginning at the time the VRG reached 0.33 MAC. RESULTS: After emergence from 1 hour of anesthesia, all simulated patients were protected against mild and severe reanesthetization if anesthesia was at 0.75 or 1.0 MAC. After 4 or 6 hours of anesthesia, severe reanesthetization occurred with all drugs with near or true apnea if anesthesia was at 1.0 or 1.5 MAC. The minimum alveolar VA to protect against severe reanesthetization after 6 hours of anesthesia was no more than 0.5 L/min for all drugs at 0.75 MAC, no more than 0.5 L/min at 1.0 MAC, and no more than 1.2 L/min at 1.5 MAC. In all simulated cases, the source of anesthetic drug that allowed reanesthetization was muscle (MUS), which reached a value of 0.8 MAC within 4 hours with all drugs and reached a value of 0.75 MAC with desflurane after 2 hours. Fat levels of anesthetic remained less than 0.15 MAC for all drugs up to the 6 hours tested. CONCLUSIONS: Reanesthetization from hypoventilation after inhaled anesthesia is possible. After initial emergence, muscle is a source of anesthetic and predisposes to reanesthetization while fat is a sink for anesthetic and fosters continued emergence. Severe hypoventilation will cause some degree of reanesthetization from anesthetic released from muscle after 4 hours of 1 MAC inhaled anesthesia with desflurane, sevoflurane, or isoflurane.

Airway ventilation pressures during bronchoscopy, bronchial blocker, and double-lumen endotracheal tube use: an in vitro study.

OBJECTIVE: To quantify inspiratory flow resistance of instrumented single-lumen and double-lumen endotracheal tubes. DESIGN: Bench-top in vitro experiments. SETTING: Laboratory of a university hospital. PARTICIPANTS: In vitro lung simulator. INTERVENTIONS: A lung simulator was ventilated mechanically via several single- and double-lumen endotracheal tubes (ETT) that were instrumented with adult and pediatric bronchoscopes as well as bronchial blockers. While ventilating with a square-flow wave and increasing peak inspiratory flow from 10-100 L/min, the pressures proximal and distal to the instrumented ETT were measured. Flow (Q) and the pressure loss (∆P) were related with regression of the quadratic equation: ∆P=k1Q+k2Q2. MEASUREMENTS AND MAIN RESULTS: With all combinations of single-lumen endotracheal tubes, double-lumen endotracheal tubes, bronchial blockers, and adult and pediatric bronchoscopes, ∆P was accurately related to Q using the quadratic equation with excellent fit, R2>0.99 for all combinations. The regression parameters k1 and k2 were statistically significant for all combinations except k1 with a bronchoscope through 37-Fr double-lumen endotracheal tube. Parameter k2 was dominant at flows above 10 L/min for uninstrumented airways and 20 L/min for instrumented airways. ∆P increased dramatically with flow, and increased with decreasing endotracheal tube size or addition of instrumentation in a quantitatively predictable manner. CONCLUSIONS: Pressure loss across instrumented endotracheal tubes follows a predictable flow-dependant quadratic pattern. Using the quantitative in vitro results of this study, a clinician can maximize inspiratory ventilation pressures during these situations without delivering excessive airway pressures to the patient.

Critical parameters in drug delivery by intravenous infusion.

INTRODUCTION: Intravenous infusion is commonly used to deliver medications and fluids to patients. The duration of an infusion is short (hours) in the operating room where intravenous agents are infused to anesthetize patients and to manage circulation. Critically ill patients often receive infusions for days. Infusion technology has become increasingly sophisticated and complex. The technical advances broaden the clinical application of intravenous infusion methodology and provide safety features. AREAS COVERED: This article provides an historical overview of intravenous infusion and discusses components of infusion systems. A section describes configuration of components to meet clinical needs. The article describes physical properties of infusion systems, emphasizing how critical parameters of resistance to flow, infusion pump performance and interactions between fluid flows and the dead volume influence medication and fluid delivery. The authors emphasize the use of infusions in the intensive care and operating room environments, although the general principles apply to other clinical settings. EXPERT OPINION: Intravenous infusion systems contribute significantly to clinical care, but in a deceptively simple way. Several critical parameters combine to influence the performance of an infusion system, with a number of pitfalls potentially confounding utility of the technology. Safe and effective clinical application of intravenous infusion technology depends on an appreciation of this complexity which impacts the performance of infusion systems.

Closed-form solutions for the optimum equivalence of first-order compartmental models and their implications for classical models of closed-circuit anesthesia.

Given a function that describes the uptake of a substance into the body with time, an analytical technique is described which transforms that function into a model of parallel first-order compartments that converges to the same uptake profile as the number of compartments is increased. The fitting of the compartmental model to the given uptake function is optimized to minimize the squared error. A necessary condition of the analytical method is that the uptake function be capable of being successively integrated at least as many times as the number of desired compartments. The uptake function should also be monotonically decreasing as all parallel first-order compartment models predict monotonically decreasing uptake. We applied this technique and ascertained the compartmental structure of the Severinghaus relationship, a longstanding observation in the field of clinical anesthesia that the uptake of nitrous oxide follows an inverse-square-root of time profile. The Severinghaus relationship is numerically poorly behaved at a time of zero elapsed minutes, predicting an instantaneously infinite uptake. Nevertheless, modeling of the first minute of anesthesia is necessary for characterizing the initial induction of anesthesia and methods of maintaining closed-circuit anesthesia such as the unit dose method. Using solely analytical methods, solutions for the compartmental properties of a mammillary model that matches the Severinghaus relationship for any expressed time interval are produced. These properties are compared to currently accepted values for the uptake of nitrous oxide. When matched to the Severinghaus relationship in the range of 0-100 min with a three-compartment model, we identified time constants of 0.28, 4.69 and 33.49 min with associated apparent volumes of 1.44, 2.14 and 7.97 l, respectively. The time constants in particular contrast to our earlier findings for the range of 1-100 min (1.46, 7.41 and 42.0 min). Our earlier findings were well matched to published time constants for tissues in classical pharmacokinetic models for volatile uptake. Consequently, we conclude that rigid adherence to the Severinghaus relationship from a time of zero minutes may lead to the over-administration of anesthetic agent due to an implicit mischaracterization of the relevant compartmental properties.

The Severinghaus square root of time relationship for anesthetic uptake and its implications for the stability of compartmental pharmacokinetics.

For nitrous oxide, the first anesthetic for which uptake was measured in humans, Severinghaus noted empirically that a plot of the log of uptake against the log of elapsed time produced a straight line with slope -0.5, suggesting that uptake is proportional to the inverse square root of time. This result is something of a black box model, based on empirical curve fitting without regard to physiology. Some authors (e.g., Lowe) repeatedly returned to this inverse square root of time model as a benchmark while others (e.g., Hendrickx) questioned its validity and demanded the relationship be expressed with a physiologic model whose structure matches the known physiology being modeled. Nevertheless, the fact that authors have repeatedly come back to this inverse square root of time model as a benchmark suggests that it might have some underlying validity which has not previously been recognized. We re-explored this mathematically in an attempt to reveal hitherto undiscovered insights or limitations. In this study, we examined the square root of time model (viewed as a power function) and compared it with multi-compartment models. Further, we explored the stability of this relationship to systematic variation in the power value and also to the superimposition of noise-like perturbations, seeking conditions under which it might not work. Based upon this theoretical analysis, we also speculate on the existence of a physiological compartment with a time constant between that of the vessel-rich group (VRG) and muscle, and what the identity of such a compartment might be.

Temporal patient state characterization using Iterative Order and Noise (ION) estimation: applications to anesthesia patient monitoring.

OBJECTIVE: As more sensors are added to increasingly technology-dependent operating rooms (OR), physicians such as anesthesiologists must sift through an ever-increasing number of patient parameters every few seconds as part of their OR duties. To the extent these many parameters are correlated and redundant, manually monitoring all of them may not be an optimal physician strategy for assessing patient state or predicting future changes to guide their actions. METHODS: The method is illustrated by application to seventy-six anesthetized patients for which thirty-two fundamental and derived variables were recorded at 20-second intervals. The Iterative Order and Noise estimation algorithm (ION) estimated the noise on each parameter. The performance of principal components analysis (PCA) was improved by normalizing the noise estimated by ION to unity. A linear regression of the resulting seven high signal-to-noise ratio principal components (PC's) predicted tachycardia 140 seconds in advance. RESULTS: ION estimated the noise on each parameter with sufficient accuracy to increase the number of significant PC's from two to seven, all of which had identifiable physiological correlates. The resulting receiver operating characteristic (ROC) suggested that a 70 percent prediction rate with 5 percent false alarms could be achieved. CONCLUSIONS: This paper illustrates the use of ION to improve significantly the performance of PCA in the efficient representation of patient state and in improving the performance of linear predictors of clinically significant parameters.