Anaesthesia, 2007, 62, pages 1251–1256 doi:10.1111/j.1365-2044.2007.05257.x .....................................................................................................................................................................................................................
Closed-loop feedback computer-controlled infusion of phenylephrine for maintaining blood pressure during spinal anaesthesia for caesarean section: a preliminary descriptive study* W. D. Ngan Kee,1 Y. H. Tam,2 K. S. Khaw,3 F. F. Ng,4 L. A. Critchley1 and M. K. Karmakar3 1 Professor, 2 Scientific Officer, 3 Associate Professor and 4 Research Nurse, Department of Anaesthesia and Intensive Care, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, Hong Kong, China Summary
We describe the novel use of a closed-loop feedback computer-controlled infusion of phenylephrine for maintaining blood pressure in 53 patients having spinal anaesthesia for elective caesarean section. A simple on–off algorithm was used that activated an intravenous phenylephrine infusion at 100 lg.min)1 when systolic blood pressure was less than or equal to baseline and stopped the infusion when systolic blood pressure exceeded baseline. Up to uterine incision, 94.6% of all systolic blood pressure measurements were within the range (baseline ± 20%). Seven patients (13.2%) had one or more episodes of hypotension (systolic blood pressure < 80% of baseline) and 23 patients (37.7%) had one or more episodes of hypertension (systolic blood pressure > 120% of baseline). No patient had nausea or vomiting and in no case was umbilical arterial blood pH < 7.2. Calculated system performance parameters were comparable with those of previously published closed-loop systems and provide a reference for the potential development and comparison of more advanced algorithms. . ......................................................................................................
Correspondence to: Warwick D. Ngan Kee E-mail:
[email protected] *Presented in part at the 66th National Scientific Congress of the Australian Society of Anaesthetists, Perth, Australia, September 2007. Accepted: 21 July 2007
Recent studies have demonstrated the efficacy and clinical utility of phenylephrine infusions for maintaining maternal blood pressure during spinal anaesthesia for caesarean section [1–4]. Previously, this has been performed using a syringe or infusion pump with titration by manual rate adjustments according to the maternal blood pressure. Although this method is effective for maintaining maternal blood pressure and preventing adverse effects of hypotension on mother and fetus, manual titration can be labour-intensive and distracting. Thus, automation of the process has potential benefits. Delivery of drugs by closed-loop feedback computercontrolled infusion is well described in anaesthesia and other specialties and techniques for assessing the performance of such systems have been established [5, 6]. However, this technique has not been applied to the 2007 The Authors Journal compilation 2007 The Association of Anaesthetists of Great Britain and Ireland
administration of vasopressors for maintaining blood pressure during regional anaesthesia. The purpose of this investigation was to describe the use of closed-loop feedback computer-controlled infusion of phenylephrine for maintaining maternal blood pressure during spinal anaesthesia for elective caesarean section. In a preliminary study, we developed a simple computer-controlled on–off algorithm based on our previously described manual algorithm [2–4]. We aimed to evaluate its feasibility and performance and to establish a basis for the possible development of more advanced algorithms. Methods
Institutional approval was obtained from the Joint Chinese University of Hong Kong–New Territories East 1251
Æ
W. D. Ngan Kee et al. Computer-controlled infusion of phenylephrine Anaesthesia, 2007, 62, pages 1251–1256 . ....................................................................................................................................................................................................................
Cluster Clinical Research Ethics Committee, Shatin, Hong Kong, China, and the study was registered with the Centre for Clinical Trials Clinical Registry of the Chinese University of Hong Kong. The computer-controlled infusion program used was developed and modified for this study by one of the authors (YHT) using Microsoft VISUAL STUDIO 6.0 using Visual C++ running under the Windows XP operating system. By user selection, the program can be switched between fixed rate mode or closed-loop feedback mode. A schematic diagram showing the basic operation of the system running under closed-loop feedback mode is shown in Fig. 1. Automatic pump error correction was incorporated into the program so that at each 2-s data sampling interval, the predicted delivered infusion volume was compared with the actual volume delivered and a corrective adjustment was made to the infusion rate. The program was run on a laptop computer that was connected by serial RS232 connections to the anaesthesia machine (Narkomed 4, North American Dra¨ger, Telford, PA, USA) for downloading and processing of monitoring User inputs: Set point (target SBP) Alarm limits (SBP, HR) Infusion rate limits
Graphical user interface
Set-point & limits
Display data Input signal (measured SBP & HR)
Control algorithm
Monitor
Control signal
Syringe pump (phenylephrine infusion)
Drug effect
Patient
Figure 1 Schematic diagram of closed-loop controller for infu-
sion of phenylephrine during caesarean section under spinal anaesthesia. The target systolic blood pressure was entered as the set-point and user-defined upper and lower values for systolic pressure and heart rate were entered for alarm limits. Limits for infusion rate were kept fixed at 0–60 ml.h)1. Monitoring data (input signal) were received from the anaesthesia machine and analysed at 2-s intervals. A simple on–off control algorithm activated the syringe pump (control signal) to deliver phenylephrine at 100 lg.min)1 (60 ml.h)1) when the measured value for systolic pressure was less than or equal to the set-point, or stop the infusion when the systolic pressure exceeded the setpoint. When the systolic pressure or heart rate exceeded the alarm limits, a visual alert was activated on the display but these data were not used by the control algorithm. SBP, systolic blood pressure; HR, heart rate. 1252
data and to a syringe pump (Graseby 3500 Anaesthesia Pump, Graseby Medical Ltd, Watford, Herts, UK) for controlling the phenylephrine infusion. The phenylephrine infusion was prepared at a concentration of 100 lg.ml)1 in a 50-ml syringe that was connected via fine-bore extension tubing to the patient’s intravenous cannula by a three-way stopcock without using a oneway valve. We recruited 60 women of ASA physical status 1–2, with term singleton pregnancies scheduled for elective caesarean section under spinal anaesthesia. All patients gave written, informed consent. We did not include patients with pre-existing or pregnancy-induced hypertension, cardiovascular or cerebrovascular disease, known fetal abnormality, or any signs of onset of labour. Patients were given oral premedication of famotidine 20 mg the night before and on the morning of surgery and 30 ml sodium citrate 0.3 M on arrival in the operating theatre. Standard monitoring was attached, including non-invasive blood pressure measurement, electrocardiography, and pulse oximetry. Fetal heart rate was monitored by external cardiotocography until the time of surgical preparation. We allowed patients to rest undisturbed in the left tilted supine position for several minutes, during which blood pressure was measured every 1–2 min. Measurements of blood pressure were continued until they became consistent (three successive measurements of systolic pressure that had a difference of no more than 10%). Baseline systolic pressure was calculated as the mean of the three recordings and this value was used as the target blood pressure or set-point for the closed-loop system. We then inserted a 16-G cannula into a forearm vein under local anaesthesia and connected this using a widebore infusion administration set to a 1-l bag of warmed Hartmann’s solution that was suspended at a height of 1.5 m above the operating table. The infusion was initially adjusted to provide a minimal rate to maintain the patency of the vein. No intravenous prehydration was given. We then turned the patient to the right lateral position and induced spinal anaesthesia. After skin infiltration with lidocaine, a 25-G pencil-point needle was inserted at what was estimated to be the L2-3 or L3-4 vertebral interspace and 2.0 ml hyperbaric bupivacaine 0.5% (10 mg) with fentanyl 15 lg was injected intrathecally. We then immediately returned the patient to the tilted supine position. Directly after spinal injection, we started rapid intravenous fluid infusion (cohydration) and activated the computer program to commence phenylephrine infusion at a fixed rate of 100 lg.min)1 (60 ml.h)1). Cohydration was continued up to a total of 2 l, after which the rate was slowed to a slow rate sufficient to allow continuous flushing of the phenylephrine solution through the intravenous cannula. 2007 The Authors Journal compilation 2007 The Association of Anaesthetists of Great Britain and Ireland
Æ
Anaesthesia, 2007, 62, pages 1251–1256 W. D. Ngan Kee et al. Computer-controlled infusion of phenylephrine . ....................................................................................................................................................................................................................
One minute after spinal injection, the automated noninvasive blood pressure monitor was restarted and cycled every 1 min. Immediately after the first measurement, the computer was switched to closed-loop feedback control. From this time until delivery, the computer activated the infusion if systolic pressure was £ baseline and stopped the infusion if it was > baseline. Hypotension was defined as a decrease in systolic pressure by > 20% below baseline and hypertension was defined as an increase by > 20% above baseline. Whenever hypotension or hypertension occurred, this triggered a visual alarm display. The operation of the closed-loop controller was monitored continuously by one of the investigators who, according to his clinical discretion, could manually override the system in the event of perceived error or gross fluctuations of blood pressure. In particular, this included problems with the accuracy of the non-invasive monitor such as those caused by movement of the patient or shivering. As an additional safety measure, the program was set to default to an infusion rate of zero and to display a visual alarm in the event that the blood pressure monitor did not produce a reading after any measurement cycle. Five minutes after intrathecal injection, we measured the upper sensory level of anaesthesia by assessing loss of pinprick discrimination and then invited the surgeon to scrub. Further checks of the height of block were made as required before the start of surgery but these levels were not recorded as part of the study. We recorded the times of skin incision, uterine incision and delivery with a stopwatch. We continued the computer-controlled infusion until the time of uterine incision. After this, the study was terminated and further management was at the discretion of the attending anaesthetist. We recorded the total dose of phenylephrine given up to the time of uterine incision. We did not routinely give oxygen unless the arterial oxyhaemoglobin saturation decreased to less than 95%, when we gave oxygen 5 l.min)1 by clear facemask. We recorded any incidences of nausea (reported by patients) or vomiting (observed by investigators) and the total amount of intravenous fluid given up to the time of uterine incision. After delivery, we gave oxytocin 5 IU by slow intravenous injection. The attending paediatrician assessed Apgar scores at 1 and 5 min after delivery. We took arterial and venous blood samples from a double-clamped segment of umbilical cord for immediate measurement of blood gases using a Rapid Point 400 analyser (Bayer Diagnostics Mfg (Sudbury) Ltd, Sudbury, UK). Statistical analysis Performance of the closed-loop controller was assessed using previously described methods [5], and parameters were calculated as follows [6]. 2007 The Authors Journal compilation 2007 The Association of Anaesthetists of Great Britain and Ireland
Percentage performance error (PE) Percentage performance error was defined as the difference between each measured value of systolic pressure and the baseline value, expressed as a percentage of the baseline value. For each patient until the time of uterine incision, it was calculated as follows: PEij ¼
ðmeaSBPij tarSBPiÞ 100 tarSBPi
ð1Þ
where PEij is the percentage performance error for the ith patient at the jth minute, meaSBPij is the measured systolic pressure for the ith patient at the jth minute, and tarSBPi is the target systolic pressure (set-point for the closed-loop system) for the ith patient. Median performance error (MDPE) Median performance error is a measure of bias and describes whether the measured values for systolic pressure are systematically either above or below the baseline value. For each patient, it was defined as the median of all values of PE and was calculated as follows: MDPEi ¼ medianfPEij; j ¼ 1; . . . ; Nig
ð2Þ
where MDPEi is the median performance error for the ith patient and Ni is the number of values for PE obtained for the ith patient. Median absolute performance error (MDAPE) Median absolute performance error is a measure of inaccuracy and represents an average of the magnitudes of the differences of measured values for systolic pressure above or below the baseline value. For each patient, it was defined as the median of the absolute values of PE (|PE|) and was calculated as follows: MDAPEi ¼ medianfjPEijj; j ¼ 1; . . . ; Nig
ð3Þ
where MDPEi is the median absolute performance error for the ith patient. Wobble Wobble is a measure of the intrasubject variability of PE about MDPE. It was calculated as follows: WOBBLEi ¼ medianfjPEij MDPEij; j ¼ 1; . . . ; Nig ð4Þ where WOBBLEi is the wobble for the ith patient. Divergence Divergence describes the trend of changes in |PE| with time and is a measure of whether the magnitudes of the differences between measured and target values for systolic pressure increase (positive value for divergence) or decrease (negative value for divergence) with time. It was defined for each patient as the slope of the linear 1253
Æ
W. D. Ngan Kee et al. Computer-controlled infusion of phenylephrine Anaesthesia, 2007, 62, pages 1251–1256 . ....................................................................................................................................................................................................................
regression equation of the values of |PEij| for that patient against time. It was calculated as follows: DIVERGENCEi P PNi PNi Ni Ni j¼1 ðTij jPEijjÞ j¼1 Tij j¼1 jPEijj ¼ P PNi Ni 2 Ni j¼1 Tij2 ð j¼1 ðTijÞ ð5Þ where DIVERGENCEi is the divergence for the ith patient and T is time in minutes. All calculations were performed using Microsoft Office EXCEL 2003 (Microsoft Corporation, Redmond, WA, USA). Results
A total of 53 patients completed the study. In seven patients, the closed-controller could not be used or was stopped: three because shivering prevented accurate measurement of blood pressure, two because of technical problems with the intravenous cannula, one because spinal anaesthesia failed, and one because the patient was asked by husband to withdraw consent. Patients’ characteristics are shown in Table 1. Overall, 94.6% of all systolic pressure measurements were within the range baseline (baseline ± 20%). Seven patients (13.2%) had one or more episodes of hypotension. This was transient in all except one patient who was given a single manual bolus of phenylephrine 100 lg after three consecutive recordings of hypotension. Twenty patients (37.7%) had one or more episodes of hypertension. Of these, six patients had more than two consecutive episodes of hypertension (three patients with three episodes and three patients with four episodes). In all cases of hypertension, blood pressure decreased spontaneously back towards baseline without intervention. No patient complained of nausea or vomited and no patient required treatment for bradycardia. No patient required supplementary oxygen. Other clinical outcomes are summarised in Table 2. One neonate had a 1-min Apgar score of 5; all other 1-min Apgar scores were = 7 and all 5-min Apgar scores were = 9. In no case was umbilical arterial blood pH < 7.2. Table 1 Characteristics and surgical times of patients under-
going caesarean section with computer-controlled infusion of phenylephrine. Values are mean (SD) or median [IQR (range)]. Age; year Weight; kg Height; cm Baseline systolic blood pressure; mmHg Baseline heart rate; beats.min)1
1254
32 (4.5) 68.5 (9.5) 159 (5.8) 112 [104–120 (96–139)] 85 [75–93 (62–112)]
Table 2 Clinical outcomes of patients undergoing caesarean
section with computer-controlled infusion of phenylephrine. Values are median [interquartile range (range)]. Block height at 5 min; dermatome Induction-to-uterine incision time; min Uterine incision-to-delivery time; s Total dose of phenylephrine; lg Total volume of intravenous fluid; ml Maximum systolic pressure; mmHg Minimum systolic pressure; mmHg Maximum diastolic pressure; mmHg Minimum diastolic pressure; mmHg Maximum heart rate; beats.min)1 Minimum heart rate; beats.min)1 Apgar score at 1 min Apgar score at 5 min Birth weight; kg Umbilical arterial pH Umbilical arterial PCO2; kPa Umbilical arterial PO2; kPa Umbilical arterial base excess; mmol.l)1 Umbilical venous pH Umbilical venous PCO2; kPa Umbilical venous PO2; kPa Umbilical venous base excess; mmol.l)1
T5 (T4–T6 [T2–T8]) 26.1 (23.7–30.0 [18.8–52.0]) 111 (76–181 [28–276]) 1130 (930–1450 [270–2870]) 1800 (1500–2020 [450–2100]) 130 (123–146 [106–171]) 99 (92–108 [60–127]) 75 (70–81 [57–89]) 45 (40–51 [16–62]) 102 (92–113 [70–162]) 60 (53–65 [40–84]) 9 (9–10 [5–10]) 10 (10–10 [9–10]) 3.10 (2.94–3.30 [2.42–4.43]) 7.30 (7.28–7.32 [7.24–7.36]) 7.0 (6.4–7.5 [5.1–9.1]) 2.1 (1.9–2.5 [1.3–3.4]) )1.6 () 3.5–0.6 [)7.7–1.0]) 7.35 (7.33–7.37 [7.30–7.40]) 6.1 (5.5–6.3 [4.2–7.6]) 3.6 (3.1–4.1 [2.3–7.0]) )1.3 () 2.6–0.5 [)5.4–1.7])
Figure 2 Systolic blood pressure plotted against time for all
patients during caesarean section under spinal anaesthesia with computer-controlled infusion of phenylephrine.
Raw data for systolic pressure are shown in Fig. 2. A graphical presentation of the performance of the system is provided by plotting percentage performance error (PE) against time (Fig. 3). A summary of performance characteristics is given in Table 3. Discussion
This study is the first description of the use of closed-loop feedback computer-controlled infusion of a vasopressor 2007 The Authors Journal compilation 2007 The Association of Anaesthetists of Great Britain and Ireland
Æ
Anaesthesia, 2007, 62, pages 1251–1256 W. D. Ngan Kee et al. Computer-controlled infusion of phenylephrine . ....................................................................................................................................................................................................................
Figure 3 Percentage performance error (PE) plotted against
time for all patients during caesarean section under spinal anaesthesia with computer-controlled infusion of phenylephrine. Table 3 Performance characteristics of closed-loop feedback
computer controller for infusion of phenylephrine during caesarean section (see text for details). Values are mean (SD). Median performance error (MDPE);% Median absolute performance error (MDAPE); % Wobble; % Divergence; %.min)1
2.4 (4.5) 6.0 (3.1) 5.5 (2.3) )0.06 (0.30)
for maintaining maternal blood pressure during spinal anaesthesia for caesarean section. Once set up, the system required minimal intervention from the anaesthetist and resulted in good clinical outcomes for both the mothers, as evidenced by the absence of maternal symptoms, and the neonates, as evidenced by umbilical cord blood gases and Apgar scores. The incidence of hypotension in our patients was small and in only one case was a manual bolus of phenylephrine required. Several patients had one or more episodes of hypertension but in all cases systolic pressure decreased back towards baseline shortly after the phenylephrine infusion stopped and no interventions were required. These results are similar to those achieved during manual controlled phenylephrine infusion [4], but because the closed-loop controller is automated, we believe it has the advantage of being much less demanding of time and attention from the anaesthetist. However, our study was observational, and to prove this would require a direct comparison with a control group receiving phenylephrine manually. In this preliminary system, we utilised a basic on–off control algorithm that was modelled on our previously described manual algorithm [2–4]. This algorithm was originally devised to be simple and easy to use when the 2007 The Authors Journal compilation 2007 The Association of Anaesthetists of Great Britain and Ireland
syringe pump is adjusted manually by the anaesthetist; by not utilising variable rates of infusion, we aimed to reduce the margin for error and produce a system that could be easily used in clinical practice. We assessed the performance of the system using parameters that have previously been used to evaluate closed-loop feedback computer-controlled infusion of drugs [5, 6]. The positive value for MDPE showed that our system resulted in values for systolic pressure that on average were slightly above baseline values, with a median difference of 2.4%. The value for MDAPE showed that the median magnitude of the difference between each measured value of systolic pressure and the target systolic pressure was 6.0%. The value for wobble showed that the PE fluctuated around MDPE with a median magnitude of 5.5%. The negative value for divergence showed that the absolute performance error tended to decrease with time, indicating that the performance of the controller tended to improve with time, although there was great variability for this parameter. These results are comparable to those found in other clinical trials of closed-loop controllers [7–9] and provide reference data that may facilitate comparison of other algorithms that could be developed to improve performance. For example, possible modifications could include incorporation of proportional, integral and derivative components to the algorithm. It is possible that use of more advanced algorithms might result in smaller rates of hypotension and hypertension than found in the current study. We administered rapid intravenous cohydration simultaneously with the computer-controlled infusion of phenylephrine in a similar fashion to that used when we have manually infused phenylephrine [4]. Previously, we reported that use of cohydration with a phenylephrine infusion improved haemodynamic stability compared with a phenylephrine infusion alone [4]. Cohydration may be beneficial because it results in augmentation of intravascular volume that coincides the time that the block and consequent vasodilatation are evolving. Alternatively, it may simply facilitate rapid circulation of the vasopressor. The total dose of phenylephrine given to patients in our study was greater than that reported in studies by other investigators [1, 10, 11]. However, our technique was based on our previous studies [2–4] in which we have not observed adverse neonatal effects despite similarly large doses of phenylephrine. When we previously compared different phenylephrine infusion regimens we found that the incidence of nausea and vomiting was smallest and values for umbilical arterial pH were greatest when we maintained the maternal systolic pressure closest to baseline values, despite this group receiving the largest total dose of phenylephrine [3]. 1255
Æ
W. D. Ngan Kee et al. Computer-controlled infusion of phenylephrine Anaesthesia, 2007, 62, pages 1251–1256 . ....................................................................................................................................................................................................................
We relied on non-invasive measurement of blood pressure to provide feedback data to the closed-loop controller. Disadvantages of this included the fact that measurement was intermittent and also prone to artefacts from movement, including shivering. The latter resulted in the controller having to be stopped in three patients. It is likely that use of invasive arterial pressure measurement would have resulted in better control but this would have reduced the applicability of our system to normal clinical practice because the use of invasive monitoring is uncommon in routine uncomplicated obstetric cases. Our system was also imperfect in that it delivered active treatment to increase systolic pressure when the measured value was below the set-point, but there was no active treatment to decrease systolic pressure when the measured value was above the set-point. Instead, the system relied on passive decreases in systolic pressure after the controller stopped the phenylephrine infusion. In some cases this resulted in transient hypertension, but there was no evidence that this was harmful in any patient. Of note, we selected phenylephrine as the vasopressor to use in the system. We believe that use of ephedrine would have resulted in inferior performance because of its slower speed of onset and longer duration of effect compared with phenylephrine. We configured the closed-loop controller to maintain systolic pressure, consistent with the majority of previous studies using manual control. As an alternative, mean arterial pressure could be used and may be a more physiologically appropriate endpoint. A comparison between the use of systolic and the use of mean blood pressure as targets would be of interest. In summary, this study described the use of a closedloop feedback computer-controlled infusion system for delivering phenylephrine for maintaining blood pressure during spinal anaesthesia for caesarean section. A computer-controlled system has the advantage of freeing the anaesthetist from the tasks of regularly assessing changes in blood pressure and manually adjusting the vasopressor infusion rate, and eliminates the potential for human error. However, it is subject to the limitations imposed by non-invasive measurement of blood pressure. The performance characteristics of a simple on–off algorithm described in this study may be a useful reference for the future development and comparison of alternative algorithms that might potentially improve haemodynamic control.
for their assistance and co-operation. The work described in this paper was substantially supported by a grant from the Research Grants Council of the Hong Kong Special Administrative Region, China (Project No. CUHK4408 ⁄ 05 M). References
The authors thank the midwives of the Labour Ward, Prince of Wales Hospital, Shatin, Hong Kong, China,
1 Cooper DW, Carpenter M, Mowbray P, Desira WR, Ryall DM, Kokri MS. Fetal and maternal effects of phenylephrine and ephedrine during spinal anesthesia for cesarean delivery. Anesthesiology 2002; 97: 1582–90. 2 Ngan Kee WD, Khaw KS, Ng FF, Lee BB. Prophylactic phenylephrine infusion for the prevention of hypotension during spinal anesthesia for cesarean delivery. Anesthesia and Analgesia 2004; 98: 815–21. 3 Ngan Kee WD, Khaw KS, Ng FF. Comparison of phenylephrine infusion regimens for maintaining maternal blood pressure during spinal anaesthesia for Caesarean section. British Journal of Anaesthesia 2004; 92: 469–74. 4 Ngan Kee WD, Khaw KS, Ng FF. Prevention of hypotension during spinal anesthesia for cesarean delivery: an effective technique using combination phenylephrine infusion and crystalloid cohydration. Anesthesiology 2005; 103: 744–50. 5 Varvel JR, Donoho DL, Shafer SL. Measuring the predictive performance of computer-controlled infusion pumps. Journal of Pharmacokinetics and Biopharmaceutics 1992; 20: 63–94. 6 Struys MM, De Smet T, Greenwald S, Absalom AR, Binge S, Mortier EP. Performance evaluation of two published closed-loop control systems using bispectral index monitoring: a simulation study. Anesthesiology 2004; 100: 640–7. 7 Kansanaho M, Olkkola KT. Performance assessment of an adaptive model-based feedback controller: comparison between atracurium, mivacurium, rocuronium and vecuronium. International Journal of Clinical Monitoring and Computing 1996; 13: 217–24. 8 Absalom AR, Sutcliffe N, Kenny GN. Closed-loop control of anesthesia using Bispectral index: performance assessment in patients undergoing major orthopedic surgery under combined general and regional anesthesia. Anesthesiology 2002; 96: 67–73. 9 Luginbuhl M, Bieniok C, Leibundgut D, Wymann R, Gentilini A, Schnider TW. Closed-loop control of mean arterial blood pressure during surgery with alfentanil: clinical evaluation of a novel model-based predictive controller. Anesthesiology 2006; 105: 462–70. 10 Alahuhta S, Ra¨sa¨nen J, Jouppila P, Jouppila R, Hollme´n AI. Ephedrine and phenylephrine for avoiding maternal hypotension due to spinal anaesthesia for caesarean section. Effects on uteroplacental and fetal haemodynamics. International Journal of Obstetric Anesthesia 1992; 1: 129–34. 11 Hall PA, Bennett A, Wilkes MP, Lewis M. Spinal anaesthesia for Caesarean section: comparison of infusions of phenylephrine and ephedrine. British Journal of Anaesthesia 1994; 73: 471–4.
1256
2007 The Authors Journal compilation 2007 The Association of Anaesthetists of Great Britain and Ireland
Acknowledgements
