Anaesthetic Conserving Device (Anaconda): Description and clinical results / Anaconda. Descripción y resultados clínicos.

Dr. FJ Belda, M Soro, R Badenes Valencia, Spain.

|a case study¶a case study}{La sedación es un componente esencial en el cuidado de los pacientes críticos, debido a su grave estado general, alto nivel de invasividad del tratamiento y monitorización, cuidados de enfermería y a muchos otros factores como el aislamiento, la deprivación del sueño, la incapacidad de comunicarse etc (1,2).|Sedation is an essential component in Intensive Care Units (ICU). It hinders the reaction to stress, prevents anxiety, increases comfort, and improves mechanical ventilation tolerance, facilitating nursing work (1-2).}{Además, los sedantes producen un efecto amnesiante que puede reducir los trastornos psicológicos graves a largo plazo observados en un 15% de pacientes que han estado ingresados en estas unidades (6-8)| Clinical studies shown that the most commonly used sedatives in ICU are those administered intravenously (3-4). }{Sin embargo, los anestésicos inhalatorios tanto el desflurano como el isoflurano han sido utilizados para sedaciones en pacientes críticos y han demostrado ventajas frente a la sedación intravenosa (17-22).|However, inhaled anaesthetics like desflurane and isoflurane have proved advantageous in critical patient sedation versus intravenous sedation (5-6). }{La eliminación pulmonar junto a un mínimo metabolismo de los agentes halogenados proporciona una gran precisión en el control de la sedación con una rápida y predecible recuperación.|Pulmonary elimination in addition to the very low metabolism of halogenated agents ensure sedation control accuracy and a predictable fast recovery. }{A nivel fisiológico, los agentes anestésicos inhalatorios son broncodilatadores y pueden prevenir el desarrollo de broncoespasmo (xx).|}{A las dosis utilizadas habitualmente para sedación son fármacos que permiten un buen control de la ventilación con gran estabilidad hemodinámica.|At usual sedation doses, these drugs provide good ventilation control and haemodynamic stability. }{Todo ello los aproxima a los sedantes ideales (16).|All this makes them an almost ideal sedative (7).}

{Sin embargo su uso no se ha generalizado porque los respiradores utilizados en Cuidados Críticos no permiten el acoplamiento fácil de vaporizadores y a la vez, en las Unidades no hay suficiente espacio para acoplar una máquina de anestesia.|However, their use has not yet spread, since critical care ventilators do not allow an easy fitting of vaporizers and because a gas scavenging system should be required. {No obstante, actualmente disponemos de un nuevo dispositivo para la administración de los agentes anestésicos inhalatorios isoflurano y sevoflurano denominado AnaConDa (Anesthetic Conserving Device, Hudson RCI, Uppsland Väsby, Sweden) (23-26).|Nevertheless, the new device AnaConDa (Anaesthetic Conserving Device, Hudson RCI, Sweden) (8) can be used for the administration of inhaled agents like isoflurane and sevoflurane with standard critical care ventilators.

AnaConDa

}{Briefly, the AnaConDa (ACD) is a modified heat and moisture exchanger (HME) and bacterial filter, which incorporates an extra layer of activated carbon.|AnaConDa (ACD) is a modified heat and moisture exchanger (HME) and a bacterial filter incorporating an extra layer of activated carbon. }{The anesthetic is supplied in liquid status via a syringe pump to a porous rod (evaporator), which diffuses the anesthetic over a large surface.|The anaesthetic is supplied in liquid status via a syringe pump to a porous rod (evaporator) which diffuses the anaesthetic over a large surface. }{Anesthetic is instantaneously dragged and vaporized inside the ACD by the inspiratory gas flow and delivered to the lungs.|The anaesthetic is instantaneously dragged and vaporized inside the ACD by the inspiratory gas flow and delivered to the lungs. }{The activated carbon layer absorbs some of the expired anesthetic vapour and desorbs some of it in the next inspiration (Enlund Anaesthesia 2001, 56: 429-32).|The activated carbon layer absorbs some of the expired anaesthetic vapour and desorbs some of it in the next inspiration. }{ This way, it can be used as a vaporizer device with a standard critical care ventilator, saving anaesthetic loss like a low flow circular anesthetic system.| Figure shows the main components of ACD.

 

 

 

 

 

 

 

 

 

This way, it can be used as a vaporizer device with a standard critical care ventilator, saving anaesthetic loss like a low flow circle anaesthetic system. }{De hecho, en anestesia, se ha visto que reduce el consumo de anestésico a un nivel equivalente al que se produce con el uso de 1.5 L de FGF ().|In fact, it has proved to reduce anaesthetic consumption to a level equivalent to that produced in a circle system using a fresh gas flow of 1.5 L (9).}

However, an infusion scheme ensuring the desired alveolar concentration of the inhaled anaesthetic has not yet been described.

Hand-driven infusion scheme of liquid sevofluorane for use with AnaConDa

Modelling pharmacokinetics of inhaled anesthetics has been used as an educational tool but also can be applied for developing systems of automated controlled infusion as already exist for intravenous drugs.

We developed a simple pharmacokinetic model to obtain an infusion layout for the clinical use of the ACD filter with sevoflurane.  Since infusion was hand-driven, a key objective was to make a single infusion rate change per hour, to facilitate its clinical use. The pharmacokinetic model is based on Lowe's classical model (10) for inhaled anaesthetics uptake and distribution. Losses produced via the ACD filter have been added to the consumption estimated by the model. Our model establishes the initial adjustment of the liquid sevoflurane infusion rate so that the target end-tidal concentration is attained in a preset induction time, as well as the adjustments to be made each hour for maintaining such concentration in the next hours. The model is a basic instrument for ACD filter clinical application, as it allows us to predict sevoflurane end-tidal concentration based on the infusion rate adjustment of the liquid anaesthetic. Its clinical importance stems from an easier use of sevoflurane as a sedative agent in any clinical area because it could be used with any ventilator, not requiring an anaesthesia machine.

Clinical study to determine the predictive performance of the model {Nosotros hemos desarrollado un modelo farmacocinético simple para obtener un esquema de infusión para el uso clínico del ACD filter con sevoflurano.|}{El modelo farmacocinético se fundamenta en el modelo clásico de Lowe (Lowe HJ. The quantitative practice of anesthesia. Baltimore, Williams and Wilkins, 1981) de captación y distribución de los anestésicos inhalatorios que utiliza multiples compartimentos y que ha sido reajustado posteriormente (Heffernan Anaesthesia 1982).|}

We studied }{Fifty-seven adult patients entered the study in two groups.|6030 Critical Care Unit patients who received sevoflurane for 6 hours via the ACD filter (Critical Care patients). }{Aleatoriamente fueron distribuidos en 2 subgrupos:|Infusion rate was randomly adjusted following the specific pharmacokinetic model so that a 1% (n=15) and 1.5% (n=15) alveolar target concentration of sevoflurane was reached

All patients were sedated with continuous iv perfusion propofol (starting from 2 mg/Kg/hour) and analgesia with morphine boluses or remifentanil in iv continuous infusion.} They were mechanically ventilated with different }{El patrón respiratorio estaba ajustado según las necesidades de cada paciente, con un volumen corriente máximo de 10 ml/Kg de peso y una FR para obtener una PaCO2 entre 35 y 45 mmHg.|respiratory patterns according to their needs. }{ La relación PaO2/FiO2 era en todos los casos superior a 300 y la PEEP ajustada entre 5 y 10 cmH2O.|PaO2/FiO2 ratio was over 300 for all cases and PEEP was adjusted between 5 and 10 cmH2O.

}{Todos ellos estaban recibiendo sedación con propofol en perfusión iv continua (starting from 2 mg/Kg/hour) y analgesia con bolus de morfina o remifentanilo en pc. iv.|}A BIS 2000 monitor was used to evaluate the sedation level. }{Para el estudio se añadio un monitor sidestream de capnografía y gases anestésicos (Vamos, Drager, Lubeck, Germany) que se conectó al ACD filter.|A sidestream capnograph and anaesthetic gas monitor (Vamos, Drager, Germany) connected to the ACD filter was also used. }The sampling flow (150 mL/min) was redirected to the breathing system after analysis through a port located between the ACD filter and the ET tube.

{En el momento de iniciar el estudio todos los estaban normotérmicos y hemodinámicamente estables.|Liquid sevoflurane infusion to the ACD filter started using a standard syringe infusion pump (Alaris, mod 2006, USA). }{La velocidad de infusión de la bomba se ajustó según los valores obtenidos del modelo farmacocinético.|The pump’s infusion rate was adjusted following the values obtained from the pharmacokinetic model. }{El primer ajuste, se realizó para obtener la target alveolar concentration en 10 minutos.|The first adjustment was intended to reach the alveolar target concentration in 10 minutes. }{En ese momento se redujo la velocidad de infusión e iniciar la infusion rate de mantenimiento.|At that point infusion rate was reduced to the first-hour maintenance rate. }{La velocidad de infusión solo se reajustó una vez cada hora hasta cumplirse las 6 horas del estudio, no realizando ninguna modificación de la velocidad entre los ajustes horarios.|Infusion rate was only readjusted once each hour until the study’s 6-hour period elapsed, no rate changes being made between the hourly adjustments. }{Se tomaron los valores de las constantes hemodinámicas, BIS y End-tidal Sevoflurane concentration, cada 5 minutos durante la primera hora y cada 30 minutos posteriormente.|Haemodynamic levels, BIS, and end-tidal sevoflurane concentration values were recorded every 2 minutes for the first hour and every 15 minutes afterwards.}

Computation of performance parameters

{Performance parameters were determined following the methods described by Varvel et al (Varvel JR et al. Measuring the predictive performance of computer-controlled infusion pumps. J Pharmacokinetics Biopharmaceutics 1992; 20: 63-94) as follows:|Performance parameters were determined following the methods described by Varvel et al (11) as follows:

The performance error (PE) for each data point sampled is the basic estimation of accuracy of each measurement (against the target) and is calculated as:

                        PE =  ( C_m - C_p ) / C_p ´ 100

{where C_m is the measured concentration and C_p is the predicted concentration at each time-point.|where C_m is the measured concentration and C_p is the predicted concentration at each time-point.} {From this measurement at each time-point, four derived parameters are calculated:|From this measurement at each time-point, four derived parameters are calculated:}

MDAPE: Median absolute performance error (%): {Is the median of the absolute values of PE.|median of the absolute values of PE. }{This value reflects Inaccuracy of the model and is the single best predictor of clinical acceptability of the performance of the model.|This value reflects the inaccuracy of the model and is the single best predictor of clinical acceptability of the performance of the model.}

 MDPE: Median performance error (%): {Is the median of the positive or negative values of PE.|median of the positive or negative values of PE. }{Is a measurement of bias (above-below the target) better than accuracy.|It measures bias (above-below the target) better than accuracy.}

Wobble (%): {Is the median of the absolute variability of PE, calculated as PE-MDPE.|median of the absolute variability of PE, calculated as PE-MDPE.} {Less wobble, more stability of the obtained concentration.|Less wobble means more stability in the obtained concentration.}

 

Divergence (% per hour): {Is a time-related parameter that indicates how the inaccuracy of the model changes as time increases.|time-related parameter that indicates how the inaccuracy of the model changes as time increases. }{Is obtained from the slope of the PE versus time in each patient.|It is obtained from the slope of PE versus time in each patient. }{Zero indicates that the accuracy is maintained over time.|Zero indicates accuracy is maintained over time. }{When positive, it means that the PE increases over time en porcentaje de aumento por hora.|If positive, it means that PE increases over time showing the percentage of increment per hour. }{When negative, indicates that PE is reduced over time.|If negative, then PE decreases over time.}

Clinical results{Thirty critical care patients consented to participate.|}{Los datos demográficos y diagnósticos de los pacientes de Cuidados críticos se muestran en la tabla I.|

The performance accuracies of the model with the “two-stage approach” for the first hour analysis are showed in the table III and for the 6 hours analysis in the table IV.  Performance parameters of the model are shown in the Table. The results of the “pooled data approach” are illustrated in the figures 3 and 4 for the first hour analysis and for the six hour analysis respectively.  Divergence is close to 0 for both first hour and 6 hour analysis and their confidence intervals (-2.4;9.9  for the first hour and –1.1;0.2 for the 6 hour period) both include zero, which implies no overall deviation with time.  Divergence is close to 0 and its confidence intervals (–1.1;0.2) include zero, which implies no overall deviation of the error with time.

Table : Mean Performance parameters

(Values are expressed as Mean ± standard deviation)

 

Co}Comments

{En el presente estudio se evaluó la capacidad predictiva de un modelo farmacocinético simple, para la administración de sevoflurano con el ACD filter en clínica.|WeWe }{El objetivo del modelo fue obtener una guía sencilla para el ajuste manual de una velocidad de infusión de sevoflurano líquido con una capacidad predictiva de la Et-concentration dentro de unos límites aceptables.|The model was intended as a simple clinical guide to manually adjust liquid sevoflurane infusion rate and as a way to predict Et-concentration within acceptable limits. }{Según hemos visto, la capacidad predictiva del modelo, tiene un error medio del 3.9% cuando se analiza durante 6 horas aunque es algo mayor en la primera hora (7%) no existiendo diferencias significativas en la media del error de predicción absoluto (mdape)  para las distintas target concentraciones de sevoflurano estudiadas.|As seen, the model’s predictive performance has a 3.9% average error and no significant differences are found in the median absolute prediction error (mdape) for the different sevoflurane target concentrations studied.}

{La capacidad predictiva del modelo es superior a la publicada para modelos predictivos de anesthetic uptake mucho mas compejos, que incorporan la variabilidad del gasto cardíaco, del coeficiente de partición B/T con la edad, la temperatura etc.|The model’s predictive performance is greater than that published for far more complex anesthetic uptake predictive models which include the variability in cardiac output, B/T partition coefficient with age, temperature, etc. (12). }{Estos modelos mas sofisticados se han aplicado tanto en la administración de bolus de anestésico líquido en circuito cerrado como con el ajuste del vaporizador en anestesia con flujos mínimos.|}{ Los mejores valores predictivos publicados son los obtenidos con el modelo mas simple de Kenedy y Hefferman, que es el que nos sirvió de base para construir el modelo utilizado por nosotros.|The better predictive performance of our model can be attributed to its use with AnaConDa and an open circuit, this meaning all elements not related to uptake by the organs over time are ruled out. {También cabe resaltar que la capacidad predictiva del modelo (para cualquier concentración y tiempo de infusión) no solo está dentro de los márgenes aceptables para los sistemas de infusión de anestésicos y sedantes intravenosos (Swinhoe 20-40% Anaesthesia 1998) sino que está muy por debajo de los valores obtenidos en diferentes estudios (Veselis Anesth Analg 1997, xx) que utilizan los modelos ampliamente aceptados y utilizados en clínica (Gepts, Marsh,XX).|Likewise, the model’s predictive performance is not only within acceptable limits for intravenous sedative and anaesthetic infusion systems but also below values reached in different studies using widely accepted clinically used models (13-14). }{Las causas de las grandes oscilaciones de la capacidad predictiva de los modelos de infusión intravenosa continua entre pacientes se deben a la distinta metodología de evaluación (veselis) así como a variaciones interindividuales fisiológicas, genéticas, ambientales etc que han sido analizadas por Gepts el al (Gepts Anaesthesia 1998).|}{Los mejores resultados obtenidos con nuestro modelo, mucho mas simple y con una secuencia de variación de infusión mucho menor, se deben sin duda al escaso metabolismo de los anestésicos inhalatorios.|The better results obtained by our model -simpler and with a much lower infusion variation sequence- are explained by the scarce metabolism of inhaled anaesthetics.

}{De hecho, el aproximadamente 3% de sevoflurano biodegradated of total body uptake (Shiraishi Y, Ikeda K. Uptake and biotransformation of sevoflurane in humans: a comparative study of sevoflurane with halotane, enflurane and isoflurane. J Clin Anesthesia 1990; 2: 381-6), no fue tenido en cuenta en nuestro modelo de 6 horas.|{Para sacar conclusiones sobre el modelo, vale la pena analizar el significado clínico de los resultados obtenidos.|When drawing conclusions on the system, it is worth analysing the clinical meaning of the results. }{El error medio de predicción (promedio de las medianas de cada paciente, mdape) del 3.9% en las 6 horas de estudio, significa que cuando la velocidad de infusión se ajusta para obtener una target concentración por ejemplo del 1.5% de sevoflurano, el 50% de los valores Et obtenidos se encontraron entre el 1.44 y 1.56%.|The 3.9% mean predictive error (average of each patient' medians, mdape) shows that by adjusting the infusion rate to reach a 1.5% target sevoflurane concentration, for instance, 50% of the obtained Et values ranged between 1.44 and 1.56%. }{A la vez, un bias promedio negativo, con intervalo de confianza negativo en sus dos márgenes superior e inferior indica que los errores de predicción producen en el 95% de los casos, valores Et-sevoflurano por debajo de los valores target, no produciéndose valores superiores a los deseados.|At the same time, the negative mean bias, with a negative confidence interval both in the top and bottom limits, shows that prediction errors produce, in 95% of the cases, Et-sevoflurane values under the target values. }{Esta es una característica importante en primer lugar por motivos de seguridad en la administración ya que es prácticamente imposible la sobredosificación.|First, this is important for safety reasons, as overdosing is almost impossible. }{Si a ello le añadimos la escasa variabilidad del error (mean wobble of 1.3%) y su estabilidad en el tiempo (mean divergence of 0.4 %h^-1), el modelo garantiza que los valores Et-concentration tras la primera hora, no van a sufrir modificaciones clínicamente relevantes en las horas sucesivas.|Together with the low error variability (mean wobble of 1.3%) and the stability over time (mean divergence of -0.4 %h^-1), the model ensures Et-concentration values after the first hour will not undergo clinically relevant changes in the next hours. }{De este modo, un bias negativo y constante, no solo facilita el reajuste de la et-concentration con el simple aumento de la infusion rate, sino que garantiza que los nuevos valores Et se mantendrían también estable en las siguientes horas.|Thus, the negative and constant bias does not only facilitates Et-concentration readjustment by simply increasing the infusion rate but also ensures that the new Et-values will also remain stable in the following hours.}

{Otra característica importante del modelo es que la rate de infusión, se reajusta después del “tiempo de inducción” y solo se modifica 1 vez por hora, lo que facilita su aplicación práctica at the bedside, ya que puede hacerse coincidir con los controles de las constantes clínicas realizados por la enfermera cada hora en los pacientes de cuidados críticos.|Another important feature is the fact that the infusion rate is readjusted once "induction time" has elapsed and is only changed once every hour. This makes bedside application easier, as one can make it coincide with hourly clinical sign check-ups by the nurse in critical care patients. }{Mayor número de ajustes cada hora, como se precisa en los modelos de circuito cerrado de diversa complejidad (p.ej. 7 inyecciones de enflurano líquido por hora. Vermeulen el at Br. J 2002) no serían clínicamente factibles en sedaciones prolongadas en cuidados críticos.|}{La comodidad de manejo manual del modelo descrito, hace prácticamente prescindible un sistema controlado por ordenador.|With the easy-to-handle model described, a computer-controlled system is no longer essential. }{De hecho, para un mismo paciente, la reducción de la tasa de infusión en las 6 horas oscila entre un 5 y un 15%  (según target % and weight) y se hace mas pequeña a medida de avanza el tiempo, por lo que un olvido en la modificación de la tasa de infusión de 2-3 horas, no es previsible que llegue a producir variaciones clínicamente importantes en la et-concentration de Sevoflurano.|In fact, for the same patient, the infusion rate reduction in the 6 hours fluctuates from 5 to 15% (depending on target % and weight) and decreases over time. Then, forgetting to modify the infusion rate in 2-3 hours is not likely to cause clinically relevant variations in the sevoflurane Et-concentration. }{Esta característica añade otro factor de seguridad en el uso del sistema.|This feature also adds to system safety.}

{En conclusión, este estudio demuestra la excelente capacidad predictiva durante 6 horas de un simplificado pharmacokinetic model hand-driven infusion of liquid sevoflurane for use with the ACD filter, en pacientes postoperados sin patología respiratoria en cuidados críticos.|In conclusion, the study evidences the excellent 6-hour predictive performance of a simplified pharmacokinetic model for hand-driven infusion of liquid sevoflurane for use with the ACD filter in post-operative critical care patients with no respiratory pathologies. }{Las características predictivas garantizan la seguridad del sistema y el ajuste horario, facilita su aplicación clínica.|Predictive features ensure system safety and hourly adjustment, this facilitating the clinical application. }{El significado clínico de este estudio está por establecer.|}{Su aplicación mas generalizada y con otros agentes, dictará el futuro de este nuevo device.|A more generalised application and with other agents will dictate the future of this new device.

References.

1. Mazzeo AJ. Sedation for the mechanically ventilated patient. Crit Care Clin 1995;11:937-55.

2. Ostermann ME et al. Sedation in the ICU. JAMA 2000 ;283 :1451-9

3. Izurieta R et al. Sedation during mechanical ventilation: a systematic review. Crit Care Med 2002;30:2644-8

4. Jacobi J et al. Task Force of the American College of Critical Care Medicine (ACCM) of the Society of Critical Care Medicine (SCCM). Crit Care Med 2002; 30: 119-41

5. Ibrahim AE et al. Speed  of recovery and  side–effect profile of sevoflurane sedation compared with midazolam. Anesthesiology 2001; 94: 87-94

6. Meiser A et al. Desflurane compared with propofol for postoperative sedation in the  intensive care unit. Br J Anaesth 2003; 90: 273-80.

7. Kong KL, Bion JF. Sedating patients undergoing mechanical ventilation in the intensive care unit. Winds of change? Br J Anaesth 2003; 90: 267-9

8. Enlund M et al. A new device to reduce the consumption of a halogenated anaesthetic agent. Anaesthesia 2001; 56: 429-32

9. Tempia A et al.The anesthetic conserving device compared with conventional circle system used under different flow conditions for inhaled anesthesia. Anesth Analg. 2003; 96: 1056-61.

10. Lowe HJ. The quantitative practice of anesthesia. Baltimore, Williams and Wilkins, 1981

11.  Varvel JR, Donoho DL, Shafer SL. Measuring the predictive performance of computer-controlled infusion pumps. J Pharmacokinet Biopharm. 1992; 20: 63-94.

12. Kennedy RR, French RA, Spencer C. Predictive accuracy of a model of volatile anesthetic uptake. Anesth Analg 2002; 95: 1616-21,  

13. Veselis RA, Glass P, Dnistrian A, Reinsel R. Performance of computer-assisted continuous infusion at low concentrations of intravenous sedatives. Anesth Analg. 1997; 84: 1049-57. 

14. Swinhoe CF, Peacock JE, Glen JB, Reilly CS.  Evaluation of the predictive performance of a 'Diprifusor' TCI system. Anaesthesia. 1998; 53 (Suppl 1): 61-7.