High thoracic epidural and remifentanil for anaesthesia for cardiac surgery in morbidly obese patients with obese hypoventilation syndrome

Patients with morbid obesity and obese hypoventilation syndrome are at a high risk for respiratory complications after surgery. This risk is increased substantially if sedative opioid analgesia is necessary. This group of patients present particular challenges when undergoing cardiac surgery, in which opioid anaesthesia is a mainstay of perioperative management. We present a series of such patients anaesthetized using a combination of high thoracic epidural and remifentanil infusion.

No safety problems were associated with the use of the catheter, but the catheter could not be wedged in one patient. The radiograph taken as a control showed a kinked catheter in the PA, but no stiffness in the catheter was noted after withdrawal. One of the catheters was rejected after use owing to a thermistor defect. Of the remaining 18 patients, two showed a sudden change in the correlation between CO obtained continuously and that obtained by thermodilution. Closer inspection of the obtained data suggested that the heat transfer device of the catheter was located in a branch of the PA, rather than within the PA itself, and may have resulted from patients being moved or extubated.

Conclusion:
The present study showed that this catheter is a safe and effective device for continuous real-time monitoring of CO in critically ill patients.
Method: Eighty-six patients aged 60 ± 10 years (mean ± standard deviation) were studied. A structured neurological examination and a battery of cognitive tests were completed the day before and 3 months after surgery by 68 of these patients. A catheter was positioned in the jugular bulb and blood samples were drawn immediately after CPB (post bypass [PBP]) and 6 h after surgery for estimation of S100β and malondialdehyde levels. Lactated Ringer's was used to prime the CPB circuit in the first 21 patients, and subsequent patients received nonlactated prime. Administration of aprotonin was noted. Intervariable relationships was assessed by two-tail Pearson correlation, and stepwise linear regression analysis was used to model malondialdehyde and S100β production.

Conclusion:
There is a positive relationship between lipid peroxidation and S100β immediately after CPB. S100β 6 h after surgery is related to neurological score, but not to cognitive score. However, neurological score accounts for only 3% of the variance in S100β, and on the basis of this evidence it can not be recommended as a surrogate marker of subtle neurological outcomes after CABG surgery. Multiple regression models for S100β release.

Introduction:
The hypothesis behind the present study was that remifentanil produces greater haemodynamic stability during rigid bronchoscopy than does the conventional propofol injection technique.

Method:
This was a double-blind, parallel-group, randomized study. All patients received propofol (2-3 mg/kg) and rocuronium (0.6 mg/kg iv) at induction of anaesthesia. Patients then received either remifentanil (n = 11; 1 µg/kg bolus over 1 min followed by infusion of 0.5 µg/kg) or fentanyl (n = 11; 2 µg/kg bolus followed by saline placebo infusion). Escape medications of propofol, atropine and ephedrine were given as required. Their lungs were ventilated with 100% oxygen via a Sanders' injector. Haemodynamic instability was defined as one of the following: lacrimation, sweating; systolic blood pressure = 20 mmHg above or below preoperative baseline for 1 min; and heart rate >100 or <50 beats/min for 1 min.

Results and conclusion:
There were no significant differences in patient characteristics nor duration of bron-choscopy between the two groups. Remifentanil was found to attenuate significantly the haemodynamic response to insertion of a rigid bronchoscope (P < 0.05 for increase in arterial pressure; P < 0.01 for increase in heart rate). During bronchoscopy, five patients showed somatic or autonomic responses in the fentanyl group, compared with none in the remifentanil group (P < 0.01). ST-segment changes occurred in eight patients in the fentanyl group compared with four patients in the remifentanil group (P < 0.05). The total doses of escape medications used and the total episodes of haemodynamic instability during bronchoscopy were significantly reduced in remifentanil infusion group.

Shackleton Department of Anaesthesia, Southampton General Hospital, Southampton, UK
Introduction: The reduction of autologous blood transfusion is a national priority. The use of ANH as a blood conservation strategy remains controversial and underinvestigated [1]. A recent paper [2] concluded that ANH should be avoided in patients with critical coronary stenosis or moderate-to-poor left ventricular function. There are no randomized controlled studies of the safety of ANH.

Objectives:
To investigate the safety of ANH using continuous Holter electrocardiography monitoring, serial analysis of daily postoperative electrocardiography, and by the use of troponin I levels.
Methods: Patients presenting for elective coronary artery bypass surgery were randomized into a control or an ANH group. All patients had a four-lead Holter electrocardiography monitor attached 1 h before surgery. After a standard anaesthetic induction, patients in the ANH group had 10 ml/kg blood removed while being maintained within 20% of their baseline blood pressure. This blood was reinfused after cardiopulmonary bypass. Troponin I levels were taken preoperatively, prebypass and at 24 h. All patients had daily postoperative electrocardiography analysis.

Results:
The results are shown in Table 1.
Conclusion: ANH is safe. There is no evidence of an additional ischaemic burden after haemodilution. of patients present particular challenges when undergoing cardiac surgery, in which opioid anaesthesia is a mainstay of perioperative management. We present a series of such patients anaesthetized using a combination of high thoracic epidural and remifentanil infusion.

Methods and results:
Over the past 2 years, nine patients presented for anaesthesia to one of the authors (JCB) for cardiac surgery, with both morbid obesity (body mass index ranged from 40.3 to 45.7) and obese hypoventilation syndrome requiring either nocturnal continuous positive airways pressure or, in two cases, nocturnal biphasic positive airways pressure. There were eight men and one woman (age range 43-69 years); seven required coronary revascularization and two required aortic valve replacement.
No sedative premedication was administered. All patients consented to a high thoracic epidural, which was placed by one of the authors (JCB) immediately before induction of anaesthesia after arterial and venous cannulae were in place. The epidural was sited using loss of resistance to saline through a 16g Tuohy needle with the patient in the sitting position in either the C7/T1 (seven cases) or the T1/T2 (two cases) intervertebral space. Anaesthesia was induced with propofol 0.5 mg/kg and remifentanil 1 µg/kg ideal body weight, muscle relaxation with atracurium 0.5 mg/kg, and the patient was intubated.
In the patients who required coronary revascularization 15 ml 0.5% bupivacaine was given as a bolus before sternotomy, and the remifentanil was titrated back to 0.1-0.15 µg/kg per min. In the patients who required aortic valve replacement the bupivacaine was given after the institution of cardiopulmonary bypass, and the remifen-tanil was maintained at 0.5-1 µg/kg per min until that point. Propofol 3 mg/kg per h was administered throughout.
Propofol and remifentanil were continued after surgery in the intensive care unit, where an epidural infusion of 0.15% bupivacaine was commenced at 8 ml/h. All patients were weaned from mechanical ventilation when usual criteria had been met. The time to extubation ranged from 2 to 18 h (median 7.5 ) after the end of surgery. No patient required reintubation. One patient spent 3 days in intensive care because of a post-drain removal pneumothorax; no other patient spent more than the first postoperative night in intensive care. Only one patient remained an inpatient longer than 7 days, and he went home 10 days after surgery. The epidural infusions were continued for between 48 and 72 h after surgery and supplemental analgesia was with diclofenac 50-75 mg three time a day or tramadol 50-100 mg three times a day. The epidural infusion required to produce satisfactory analgesia ranged from 6 to 18 ml/h (median 8 ml/h).

Discussion:
The use of opioids in patients with obese hypoventilation syndrome is known to carry a significant risk of respiratory depression and cardiorespiratory arrest. Providing postoperative analgesia with epidural local anaesthesia avoids the use of opioids postoperatively. Remifentanil administered during anaesthesia allows the benefits of opioid analgesia in such patients during surgery without any residual effects after surgery due to its rapid metabolism by tissue esterases. By using this combination of ultrashort-acting opioid and high thoracic epidural, it has been possible to anaesthetize this high-risk group of patients safely with no increase in duration of intensive care or hospital stay.

Department of Anaesthesia, Western Infirmary, Glasgow, Scotland
Introduction: Cardiac surgical patients use 10-20% of the nation's blood products [1]. Recent work [2] suggests a lower acceptable haematocrit for cardiac surgical patients during the perioperative period. This, and a change in our unit's policy to increase the availability of platelets and clotting factors ('crash packs') for these patients, prompted us to conduct this audit in order to determine whether our therapy was appropriately targeted.
Method: During a 3-month period we issued forms to the attending anaesthetist for all cardiac surgical procedures requesting information regarding demographics, preoperative drug history and baseline haematology. Intraoperative use of and indications for packed cells, fresh frozen plasma and platelets were noted. Postoperatively the attending nurse recorded the haematocrit, drain losses, and the use of and indications for red cells and clotting factors. Coagulation studies, and haemoglobin on CICU discharge and hospital discharge were also documented. Fifteen per cent of patients received a 'crash pack' in CICU. Only 68% of these had formal coagulation studies done. Of these only 20% had abnormal screens, defined as prothrombin time and partial thromboplastin time greater than 3 and 5 s over control, respectively. Of those who received a crash pack, only 48% had a full blood count done first. Of these only 20% had platelet counts below 100 × 10/l. Only one patient of the six who had fibrinogen measured had levels below 1; five had 'crash packs'. The mean total drainage was 1725 ml where a crash pack was given, compared with 797 ml where no clotting factors were given. The mean drainage in the 4.4% of patients done 'off-pump' was 1120 ml, compared with 901 ml in the 95.6% whose procedures were performed on cardiopulmonary bypass (CPB). The mean CICU admission haematocrit was 32% (off-pump) versus 26% (post-CPB). Those done 'off-pump' were given a mean of 2 units PRC versus 1.7 units PRC in those done on CPB. Of cases done off-pump, 33% had a crash pack. All of the 4.6% who were re-explored had crash packs.

Results
Preoperative medication made no difference to mediastinal drainage or to crash pack use. The drain losses and crash pack use in smokers was no different from that in nonsmokers.

Conclusion:
The present results suggest that our use of red cells could be targeted at lower haematocrits both intraoperatively and in CICU. There is a subgroup of patients who are given relatively small transfusions and are discharged with high haemoglobin concentrations. It would be prudent to evaluate means to avoid this, and this may particularly benefit those weighing below 60 kg. The drain losses and transfusion requirements in those patients undergoing coronary artery bypass grafting 'offpump' were significantly higher than expected, although this sample was small and may reflect a 'learning curve'. Those patients receiving clotting factors and platelets had significantly higher mediastinal drainage than those who did not. Only a small percentage of postoperative coagulation studies/platelet counts were abnormal. From our current data it is difficult to assess the contribution to haemostasis of the blood products given. A more accessible/rapid assessment of the coagulation system and platelet numbers and function may be helpful in directing our use of clotting factors and platelet concentrate.