Comparison of the Efficiency of Anesthetic Depth Control Using Closed Loop Anesthesia Delivery System vs. Manual Control Using Target Controlled Infusion in Patients With Moderate to Severe Left Ventricular Systolic Dysfunction
Comparison of the Efficiency of Anesthetic Depth Control Using Closed Loop Anesthesia Delivery System vs. Manual Control Using Target Controlled Infusion in Patients With Moderate to Severe Left Ventricular Systolic Dysfunction
Advancement in techniques for anaesthetic drug delivery and real time monitoring has facilitated safe induction and maintenance of anaesthesia in severely compromised patients. Cardiac diseases are the commonest causes of morbidity and left ventricular failure is the commonest clinical presentation at the end stage. LV systolic dysfunction is defined as reduction in LVEF ≤55%. Patients with LVEF 55%-46% have mild, 45%-36% moderate and ≤35% severe LV systolic dysfunction. Patients with heart failure have a diminished cardiac reserve capacity that may be further compromised by anaesthesia. In addition to depression of sympathetic activity, most anaesthetics interfere with cardiovascular performance, either by a direct myocardial depression or by modifying cardiovascular control mechanisms. Propofol with fentanyl is advocated as the best anaesthetic combination for induction of anaesthesia in patients undergoing CABG. Propofol is a drug with narrow therapeutic index and may cause severe hypotension and hemodynamic instability during induction of anaesthesia, especially if it is given in too large doses.
Automated drug delivery systems are popular for delivery of propofol. They can be of two types, depending on whether they are based on pharmacokinetic or pharmacodynamic principles. Closed Loop Anaesthesia Delivery system has been used world-wide and in our institute in patients of various age groups and in patients undergoing cardiac surgery. But still the studies are lacking in patients with moderate to severe left ventricular systolic dysfunction. Moreover none of the studies have compared the efficacy of anaesthetic drug delivery using these two devices in this group of patients.
Thus there is paucity of literature regarding PK and PD of propofol in patients with cardiac failure. The investigators hypothesized that as the Closed Loop Anaesthesia Delivery System is based on pharmacodyanamic principles, it should perform better than the Target Control Infusion system, which works on pharmacokinetic principles. The investigators planned to conduct this study to determine the anaesthetic depth control using Closed Loop Anaesthesia Delivery system vs. manual control using Target Controlled Infusion in patients with moderate to severe left ventricular systolic dysfunction.
MATERIALS AND METHODS:
Patient will be pre-medicated with oral tablet alprazolam 0.25 mg the night before and on the morning of surgery. In the anesthetic room, 16 G i.v. cannula will be inserted and following monitors will be attached: continuous pulse-oximetry (SpO2), electrocardiogram (ECG), periodic non-invasive blood pressure (S/Anesthesia monitor, Datex Ohmeda Inc., Madison, WI) , and continuous BIS (BIS XP, Aspect Medical Systems, Newton, MA in the S/5 Anesthesia monitor). Pre-induction arterial line and central line will be inserted in all patients and continuous arterial blood pressure and central venous pressure measurements will be recorded. Pulmonary artery catheter will also be inserted in all patients for monitoring of cardiac output. All invasive lines will be inserted under local anesthesia. NIBP measurements will be stopped after transducing arterial line, because it may interfere with the assessment of MOAA/S score.
All the patients will receive i.v. fentanyl at a dose of 3µg.kg-1 over a period 3 minutes, followed by propofol administration either by CLADS or TCI.
Group 1:
Propofol will be administered with the Diprifusor (Marsh Pharmacokinetic model, Master TCI pump, Fresenius Kabi; Bad Homburg; Germany) starting at a predetermined target plasma propofol concentration (Cp) of 1.8µg.ml-1 based on a previous study. Further target concentrations were set according to the Dixon up and down method. A failure will be followed by an increase in the target plasma concentration in the next patient by 0.2µg.ml-1 and a success will be followed by a decrease in the target plasma concentration in the next patient by 0.2µg.ml-1. TCI pump will be controlled manually and all the data including BIS and vitals will be recorded through CLADS into a laptop. MOAA/S will be assessed every 15 second for assessing time to loss of consciousness; BIS and HR will be recorded every 5 seconds and MAP and PAP every 10 seconds. Thus, the anesthetic depth control will be assessed based on time to achieve loss of consciousness and by comparing the dose of propofol required for induction and maintenance. BIS overshoots and hemodynamic stability will be noted. In case the patient achieves loss of consciousness within 5 minutes, a 2 minutes observation period will be allowed for achieving target BIS. If the patient does not achieve target BIS after this period, the case will be considered a failure. On the other hand, if the patient achieves target BIS within this period it will be considered a success. In cases where neither LOC nor target BIS will be achieved within 5 minutes will also be considered a failure Vecuronium will be administered at a dose of 0.1 mg.kg-1 after the LOC and the patient will be intubated after 4 minutes. Patients will be ventilated through face mask to maintain normocapnia administering air-oxygen mixture with FiO2 of 0.6 during the period of induction till intubation. Fentanyl infusion 1µg.kg.h-1 will provide analgesia with additional boluses of 1µg.kg-1 before skin incision, sternotomy and at the commencement of CPB. Additional bolus of analgesic supplements (fentanyl 1µg.kg-1) will be given when the mean arterial pressure (MAP) or heart rate exceeds 25% of the baseline <50. If hypertension or tachycardia persists with a BIS≤50, either nitroglycerine infusion or esmolol will be used. In conditions of hypotension, inotropic support and/or vasopressor will be initiated after ensuring normovolaemia. Atropine sulphate will be used to treat bradycardia (heart rate <45b.p.m.) after excluding other treatable causes. The MAP on CPB will be maintained between 50 and 80 mmHg using phenylephrine / nitroglycerine as required. Systemic hypothermia up to 28ºC will be practiced during CPB and patients will be actively re-warmed to 36ºC before separation from CPB.
The number of patients exhibiting episodes of hypotension or hypertension and the number of episodes of hypotension or hypertension requiring a change in the propofol target concentration or the administration of any vasoactive drugs will be recorded.
Based on response of the previous patient to propofol administration using Diprifusor at a predetermined target plasma propofol concentration (Cp), further increments or decrements of Cp in the subsequent patients will be based on Dixon's up and down method. The target Cp will be increased by 0.2 µg.ml-1 in patients who do not achieve BIS of 50 and Cp will be is decreased in subsequent cases by 0.2µg.ml-1 in patients who do. Thus, six to eight pairs of failure to success transition will be needed for calculation of the EC50 (i.e. the target concentration at which 50% of the patients achieve BIS of 50). The mean of the mid value of these transitions will be used to derive the EC50.
At the end of the study the BIS data will be analyzed for offline determination of induction time. It is defined as the time at which the BIS reaches ≤50 for 2 consecutive readings from the start of propofol infusion.
Group 2:
In the CLADS group induction and maintenance will be carried out automatically according to the BIS. Hemodynamics will be controlled automatically through CLADS as well as through manual ionotropic support as described above.
CLADS is a pharmacodynamic-pharmacokinetic model-based adaptive infusion system, which uses BIS as the controlled variable and standard infusion pump as the 'actuator'. The 'control algorithm' is based on the relation between various rates of propofol infusion (producing different plasma concentrations) and BIS, taking into consideration the pharmacokinetic variables (distribution, clearance). This was established in the developmental stage of CLADS. The algorithm alters the rate of propofol infusion to steer and maintain BIS to the set target taking into account existing BIS, time-elapsed since the initiation of infusion, pharmacokinetics, time-delay factor between sensing and averaging of BIS data, time-delay factor between change in infusion rate and actual change in the plasma concentration of propofol, as well as the peak effect of propofol.
An IBM-compatible PC with PENTIUM 4 or higher processor is used to implement the control algorithm, provide a user interface, and control communication through serial ports (RS 232) with infusion system (Pilot-C; Fresenius, Paris, France) and vital sign monitor (AS5; Datex Ohmeda Division, GE Healthcare, Singapore). The drug delivery can operate in two modes -manual and automatic. In manual mode, rate of propofol infusion is controlled manually through the keyboard. In 'automatic' mode, the system automatically controls propofol / isoflurane administration as per the control algorithm. The automatic mode further has three options: (i) induction, (ii) maintenance, and (iii) induction combined with maintenance. The user must enter a target BIS value, age, weight, height along with the status of patient-low risk (ASA I-III), high risk (ASA III-IV, NYHA class III), very high risk (ASA IV-V, NYHA III-IV), or pediatric. The system can also function in 'monitor' mode, where it only updates BIS and other patient data and provides a graphic display of current and trend values.
The system updates the electroencephalographic data every 5 s and calculates the BIS error (target BIS-actual BIS). It uses PID (proportional integral differential) algorithm based on this error to make the changes in propofol infusion rate to achieve target BIS. The algorithm fine tunes the rate and duration of propofol delivery differently during induction and maintenance phases of anesthesia delivery. During induction, controller tries to achieve the target concentration in a stepwise fashion (while continuously receiving feedback of BIS every 5 s) on the basis of the relationship between plasma concentrations and BIS. During maintenance, 30 s is deemed as one epoch. The initial three as well as the last three BIS values of each epoch are averaged and compared to assess the trend. When the trends indicate an increasing BIS, higher target concentrations and so higher propofol rates are set and vice versa if the trends indicate a decreasing BIS. These trends are also cross-checked with larger epoch trends before making drug alterations.
A safety feature has been added wherein propofol and other anesthetics are stopped automatically whenever hemodynamics drops below the safety limits set by the operator. This would restart automatically when hemodynamics improves to values above the predefined lower limit. The time delay for this automatic cut-off is at the most 10 s, which is the interval at which the vitals are updated in the controller.
Time to achieve loss of consciousness, dose of propofol required for induction and maintenance, BIS overshoots and hemodynamics will be recorded in patients administered propofol using CLADS.
At the end of the surgery, CLADS/TCI will be discontinued and patient will be shifted to the post-surgical intensive care unit for elective mechanical ventilation.
STATISTICAL ANALYSIS- Physiologic data are presented as mean (SD) and time intervals are presented as median (range). The performance of the system was assessed by calculating median performance error (MDPE), median absolute performance error (MDAPE), wobble and divergence (time related trends) using methods of Varvel et al.
The performance error is given by the formula:
PE = [(BISmeasured-BIStarget)÷BIStarget]×100 The median performance error (MDPE) which reflects the bias of CLADS in the ith subject is determined as MDPE=median {PEij j=1........Ni } where Ni is the number of (PE) values obtained in the ith sub Median Performance Error (MDPE) Median Absolute Performance Error (MDAPE)= median {I PEij j=1........Ni I} Wobble is determined from the equation Wobble=median {I PEij - MDPEiI , j=1......Ni }
Global Score (GS) - The Global Score is used to score EEG-guided depth of anesthesia control systems with one scalar:
GS=(MDAPE+WOBBLE)÷fraction of time BIS∈(40,60) Divergence is the slope of the linear regression equation of absolute performance error against time and is expressed in units of percentage divergence per minute. A positive value indicates progressive widening of the gap between targeted and measured values, whereas a negative value indicates that the measured values are converging on target. MDPE and MDAPE are measures of bias and precision, respectively, wobble measures the intra- individual variability in the performance errors. The percent of time when BIS remains within ±10 of target BIS during closed-loop (CLADS group) or manual control (TCI group) will also be calculated. Adequacy of hemodynamic control will be adjudged by the percentage of anesthesia time the mean arterial blood pressure and heart rate (HR) are within ±25% of baseline.
Differences between the groups will be analyzed using the unpaired t-test for parametric data and the Mann- Whitney U test for non-parametric data. Count data will be analysed using the chi-square test. All analyses will be performed using SPSS v21.0 for Windows (SPSS Inc., Chicago IL, U.S.A.) and a P value <0.05 will be taken to be significant.
For calculation of the EC50 (plasma concentration at which 50% patients achieve a BIS) of propofol required for induction will be calculated using Dixon and Massey method. Six to eight pairs are necessary to calculate EC50 from Dixon up and down method. Probit analysis of the data for TCI group will be then performed to determine EC50 and EC95.
From a pilot study, the investigators estimated that manual control using TCI maintained BIS in the range of ±10 of the target 70% of the time. In order to assess the 20% improvement from this control, the investigators calculated that they would need to recruit 36 patients to achieve 80% power at 5% Type I error. A total of 40 patients were recruited, taking into account possible inadvertent patient attrition.
Inclusion Criteria:
Exclusion Criteria: