Single shot spinal anaesthesia for caesarean section is currently accepted as the favoured method in the absence of contraindications, for reasons of safety and comfort. Firstly, there is an increased risk of failed intubation associated with general anaesthesia. Secondly, spinal anaesthesia, if practiced correctly, allows for a superior experience of the delivery and improved bonding with the infant. Maternal haemodynamic stability is desirable both for maternal and neonatal safety, and to diminish maternal side-effects such as nausea and vomiting. Therefore, after an extensive literature review, clinically relevant aspects of spinal anaesthesia were studied, with a view to contributing to knowledge which could improve safety and outcome. The central themes explored in this thesis were fluid management during spinal anaesthesia for caesarean section in healthy parturients, the haemodynamic effects of the vasoactive agents ephedrine, phenylephrine and oxytocin during spinal anaesthesia for caesarean section in healthy patients and in patients with preeclampsia, and short term neonatal outcome after spinal anaesthesia in patients with severe preeclampsia. Research methodology included non-invasive measures as well as the use of a pulse wave form analysis monitor to measure maternal cardiac output. A validation study was performed comparing this method with thermodilution in patients with postpartum complications of preeclampsia. Abstract viii The results of these studies showed that: The pulse wave form monitor employed showed acceptable limits of agreement with the thermodilution method. Crystalloid coload was associated with lower vasopressor requirements than conventional preload. Spinal anaesthesia was associated with afterload reduction, which was more pronounced in healthy patients than in preeclamptics. Ephedrine maintained or increased, and phenylephrine reduced maternal cardiac output in healthy patients. Oxytocin was associated with transient haemodynamic instability in healthy and preeclamptic patients, which was obtunded by phenylephrine in the healthy population. Spinal anaesthesia for caesarean section was associated with a greater umbilical arterial base deficit than general anaesthesia in patients with preeclampsia. Overall, these studies should contribute to improved knowledge of haemodynamic responses during spinal anaesthesia for caesarean section, and ultimately to improved maternal morbidity and mortality.

Includes abstract. / Includes bibliographical references. / The main purpose of the study was to help establish the best dose regimen of hyperbaric bupivacaine, when combined with intrathecal fentanyl, for spinal anaesthesia for brachytherapy for carcinoma of the cervix. This procedure is performed as a day case at Groote Schuur Hospital.

Bibliography: leaves 307-312.

Background: Magnetic resonance imaging (MRI) scanning places paediatric patients at risk of both hypothermia and hyperthermia. The aim of this study is to determine primarily if paediatric patients gain or lose heat during MRI scanning, and secondarily to examine potential risk factors for any such change. Methods: A prospective audit was conducted from February 2015 until April 2015 involving 200 children aged five days to 12 years. Tympanic temperatures were recorded pre- and post- MRI scan. Variables including age, height, weight, head circumference, area scanned, length of time in the scanning room and duration of scan were recorded. The type of anaesthetic management was decided by the anaesthetist and recorded. Results: Tympanic temperature decreased in 111 patients, with a loss of greater than 0.5°C in 29% of patients (n=58) and a range of decrease from 0.1°C to 1.9°C. Hypothermia, defined as a core temperature of less than 36°C for this study, occurred in 13.5% (n=27) patients. A total of 23 patients had no change in pre- and post-scan temperature, and 66 recorded a higher temperature post-scan. The range of gain in temperature was 0.1°C to 1.5°C, with 14.5% (n=29) of patients' temperatures increasing by 0.5°C or more. The mean pre-scan temperature was 36.603 °C ± 0.512°C (range: 35.5 - 38.70 °C) and the mean post-scan temperature was 36.442°C ± 0.615 (34.80-40.0 °C). Overall, the mean post scan temperature was 0.16°C (P<0.001) less than the pre-scan temperature. Linear regression analysis identified sedation and general anaesthesia as signficant risk factors for heat loss. Conclusion: Overall paediatric patients tend to have a minor decrease in temperature during MRI scanning. Individually, each paediatric patient may have an increase or decrease in temperature, or no change. A significant proportion of paediatric patients are at risk of hypothermia post MRI scan, and almost half are at risk of an increase or decrease in temperature of a minimum of 0.5°C. These factors are clinically significant and may be associated with adverse outcomes. For these reasons, temperature monitoring and active temperature management should be implemented during MRI scanning in paediatric patients.

Sickle cell disease (SCD) is the most common inherited haematological disorder, producing a mutation of the haemoglobin molecule known as haemoglobin S (HbS). The presence of HbS in the erythrocyte makes it prone to sickling - a process which may lead to vaso-occlusive injury, haemolysis and a hypercoagulable state. Sickling is precipitated by dehydration, hypoxia, hypothermia, acidosis and low flow states. Over time, multi-organ damage develops with significant morbidity and mortality. Paediatric patients with SCD and congenital heart defects may require anaesthesia for corrective cardiac surgery on cardiopulmonary bypass (CPB). During the perioperative period these high-risk patients may suffer significant complications when exposed to the conditions that favour erythrocyte sickling. This case series details our experience of four paediatric patients with SCD patients who underwent corrective cardiac surgery at Red Cross War Memorial Children’s Hospital. The pathophysiology is discussed and the perioperative management of transfusion, cardiopulmonary bypass and temperature regulation is highlighted.

Introduction: In South Africa, up to 30% of pregnant women are human immunodeficiency virus (HIV) positive and morbidity and mortality is high in this group. HIV positive men and women may have multiple cardiovascular comorbidities, which include systolic dysfunction that may progress to heart failure secondary to dilated cardiomyopathy. However the concurrent effect of pregnancy and HIV infection on haemodynamics has not been extensively researched. The aims of this study were to quantify haemodynamics using transthoracic echocardiography (TTE) in term pregnant women with HIV on antiretroviral (ARV) treatment and compare the data with term healthy women in the same population. Method: After ethics approval and written consent, 30 consecutive term HIV positive women and 40 healthy term pregnant women were recruited. HIV positive women had a CD4 count greater than 200 and were either on Highly Active Anti-Retroviral Therapy (HAART) or single drug management. Results: Haemodynamic assessment was possible in all patients and women in the two groups were similar in age, and body mass index. Mean CD4 count was 452 ± 187.8 and duration of therapy was 15.9 ± 22.4 months. Compared with healthy pregnant women, women with HIV have systolic changes exhibited by reductions in left ventricle (LV) septal and right ventricle (RV) systolic myocardial velocities as well as increased LV end-diastolic (ED) areas and diastolic changes of increased RV isovolumetric (IV) relaxation and reduced RV e′ diastolic myocardial velocities. These changes occur in the presence of a reduced LV mass. Pericardial effusions occurred more frequently and are of a larger size in women with HIV. These findings suggest subclinical impairment of systolic function in the LV as well as subclinical impairment of both systolic and diastolic function in the RV. Discussion: Transthoracic echocardiography can quantify cardiac function in healthy pregnant women and in pregnant women with HIV and is acceptable to the patients. HIV positive pregnant women at term on anti-retroviral therapy have hearts that have subclinical systolic dysfunction in the presence of decreased LV mass and increased end-diastolic areas. This may represent a failure to compensate for the increased haemodynamic demands of pregnancy and may be as a result of the direct effects of HIV itself or due to anti-retroviral drugs.

On reviewing the medical literature, there is a clear resurgence of interest in the use of TEA (thoracic epidural analgesic) in cardiac anaesthesia. This resurgence was brought about by laboratory-based evidence that TEA-induced sympatholysis may be cardioprotective through the promotion of myocardial blood flow to areas at-risk and subsequent early, small clinical studies suggesting that TEA was feasible, and possibly also beneficial in CABG surgery [Joachimsson et. al, 1989; Liem (1-3) et. al, 1992; Stenseth et. al, 1994]. Despite the positive results of these early studies and suggestions that TEA may be the preferred anaesthetic/analgesic technique in select groups of patients (promoting early extubation and fast-tracking) undergoing cardiac surgery, many anaesthetists are still reluctant, however, to use this technique because of the theoretical increased risk of the patient suffering a spinal haematoma and subsequent paraplegia. In order to outweigh this theoretical risk it is important that we show that added benefit, in addition to the provision of analgesia and expedited postoperative convalescence, can be obtained by using TEA. It is therefore our duty as anaesthetists and perioperative physicians to determine whether TEA may also affect the pathophysiology of the disease process, especially in the perioperative period - and thereby influencing the subsequent long term outcome and quality of life of the patient. An example of this latter point would be the potential role of TEA in; • reducing the incidence of perioperative myocardial infarction (P-MI), through the suggested cardioprotective effects of TEA, • reducing the incidence of early postoperative graft failure, through either; * reduction of native coronary artery and/or graft (conduit) spasm, or * reduction of postoperative hypercoagulability.

In this study, the administration of remifentanil to cover the tunnelling phase of shunt insertion in children caused good attenuation of haemodynamic and endocrine markers of stress, no delay in recovery and no additional post operative respiratory depression in all age groups, including xpremies and neonates.

Background: Complex regional pain syndrome (CRPS) is characterised by neuropathic-type pain in a regional distribution with associated signs and symptoms including: swelling, change in temperature, change in skin colour and motor dysfunction. Although CRPS is not a common condition, the disease burden is high and can lead to long-term impairment and disability. Current guidelines for the management of paediatric CRPS are mostly extrapolated from adult data. Literature regarding treatment and long-term prognosis of paediatric CRPS is sparse and often lacks well-defined and reproducible outcome measures. Aim: This study aims to assess the efficacy of treatment of CRPS Type 1 in children admitted to Red Cross War Memorial Children's Hospital (RCWMCH) in eliminating pain and improving function. Methods A retrospective folder review and follow-up telephonic survey of all children admitted to RCWMCH over a 5 year period was performed. Long-term follow-up was defined as a minimum of 6 months after discharge from hospital. Follow-up questionnaires included the Faces Pain Scale, and the Paediatric Quality of Life Inventory Version 4 generic core scales (PedsQL 4.0) Results: Nine children with confirmed CRPS Type1 were included in the study, all of whom participated in long-term follow-up. Children received in-hospital treatment from a multidisciplinary team, including intensive physiotherapy and psychological therapy. Pharmacological treatment consisted of an intravenous ketamine infusion and varying combinations of gabapentin, clonidine and amytriptilline. One child received epidural local anaesthetic infusion. On average, pain scores improved by 67.8% by the time of discharge. At an average long-term follow-up interval of 10.3 months (ranging between 6 to 21 months), the pain score had worsened by 10.5% compared to discharge. Two children (22.2%) experienced complete recovery while the remainder experienced partial recovery. Three children (33.3%) suffered a relapse, of which one recovered completely. At long-term follow-up, children for whom PedsQL 4.0 data was available from admission all demonstrated clinically meaningful improvements in all four functional domains (physical, social, emotional and school). Children scored on average similar to healthy peers in terms of social, emotional and school function, but worse in terms of overall quality of life. The area most affected was that of physical function, in which children with CRPS scored worse than other children with chronic health conditions. Conclusion: Children with CRPS experience significant improvement in pain and function with an in-patient, multidisciplinary, non-invasive treatment approach. Most children return to a level of school functioning comparable to healthy peers. However, the rate of complete recovery is low, relapse is common, and most children have significant persistent impairment of physical function. We recommend a long-term support program for children with CRPS and their families, and standardisation of follow-up intervals and outcome measures to enable comparison between future studies.

The reasons for this study were multi-factorial, but mostly due to some interesting data obtained from a pilot study conducted at University College London (UCL). In that study, the coagulation effects in vitro of two hetastarch solutions were compared with two crystalloids by means of thrombelastography (TEG®). The fluids compared were: 1. Hespan® (HES), a high molecular weight hetastarch (450kDa/O.7 substitution ratio) in a 0.9% saline solution - Laevosan, Austria. 2. Hextend® (HEX), also a high molecular weight hetastarch (670/0.75 substitution ratio) in a balanced electrolyte, lactate and glucose solution - BioTime Inc, Berkeley, California, USA. 3. Saline 0.9% 4. Hartmann's Solution (Ringer's Lactate) The crystalloids revealed no surprising differences known from previous published data, but data obtained from the hetastarch solutions revealed contradictory results to known in vivo results found in a phase III trial. This previous Phase III in vivo trial showed that HEX haemodilution produced a superior coagulation profile to HES, along with a significantly shorter r-time than HES. There was also a significantly smaller transfused volume of blood than HES in the HEX-treated patients. This Phase III study prompted the initial UCL in vitro haemodilution study mentioned above. In the UCL study, there were significantly impaired TEG® results, indicating severe hypocoagulability with HEX, when compared with HES. This included prolonged r-and k-times, as well as reduced a-angles and maximum amplitudes in the HEX group, compared with HES and crystalloid groups. Many theories were discussed for these controversial UCL results, but the thought was that a container-effect could have been responsible, as the in vitro UCL study methodology included the use of a polycarbonate container for initial storage, as well as for haemodilution of the blood in vitro. In view of the known wettable surface, as well as a strong negative surface charge of polycarbonate, it was suggested that the container surface itself could have affected coagulation. When different ionic compositions of the various fluids and starches were taken into account, it seemed possible that some interaction between the fluids and the material of the containers could have induced or inhibited coagulation at the container surface. The suspicion was that the observed change in TEG® variables was likely due to a methodologic idiosyncrasy. Previous track record of haemodilution and TEG research at the University of Cape Town made it an obvious setting for exploration of this problem. Preparations were thus made to test container effects with haemodilution in vitro at Prof MFM James' anaesthesia laboratory at the University of Cape Town. The hypothesis was that the use of polypropylene and polycarbonate containers, with their different chemical and surface properties, would lead to a variability in TEG® results obtained from fresh whole blood, as well as blood diluted with various fluid solutions. Choosing TEG® as a monitor of coagulation was essential, as it has a well-established track record in monitoring coagulation effects in trials of haemodilution (in vitro and in vivo). TEG® produces reliable and quick results, giving a reflection of global coagulation function. It, along with the Sonoclot®, are the only two devices which can reliably diagnose a hypercoagulable state. More will be mentioned on the TEG® later.