Background: It has been suggested that the dose requirement for spinal anesthesia (SA) is lower in obese patients for cesarean delivery. In this prospective, observational, non-inferiority study we tested the hypothesis that obesity would not have a clinically important effect on vasopressor requirements or block height. Methods: Two groups of 25 parturients, Group O (BMI >40 kg/m²) and Group N (BMI <32 kg/m²) requiring elective cesarean delivery were recruited. All patients received 10 mg intrathecal hyperbaric bupivacaine co-administered with 10 μg fentanyl. Dermatomal levels were assessed at 5 and 25 minutes after SA, and at completion of surgery, using light touch and cold sensation in response to ethyl chloride. The primary outcomes were phenylephrine requirement in the first thirty minutes following spinal anesthesia, and maximum block height, measured by the sensation of touch and cold. Secondary outcomes were total phenylephrine dose required, changes in hand grip strength, and peak flow rate. Results: There were no significant between-group differences in median block height as assessed by touch at 5 or 25 minutes, or by temperature at 5 minutes. At 25 minutes, there was a two-dermatome difference in median block height for loss of temperature sensation between Group O and Group N (T2 vs. T4, 95% confidence interval (CI) of the difference in medians 0-2 dermatomes). No blocks extended to cervical dermatomes. The median (range) phenylephrine dose for the first 30 minutes was 150 μg (0-900 μg), and 100 μg (0-1250 μg) in Group N and O respectively. The 95% CI for the difference between the two median doses was -150 μg to 100 μg. There were no differences in median percentage reductions in peak flow rate or median hand grip strength after SA. Mean surgical time was longer in Group O than in Group N (49.1 vs 39.4 minutes, 95% CI difference 1.7 to 17.7 minutes). The mean time for recovery of touch sensation to T10 was longer in Group O (152 vs 132 minutes, 95% CI difference 3.8 to 36.2 minutes). No analgesic supplementation was required. Conclusion: Only a minor increase block height as assessed by temperature occurred in Group O at 25 minutes. Vasopressor requirements during the first 30 min of SA were equivalent. Time for regression of SA block level was longer in the Group O, which may be beneficial considering the longer surgical time. A dose of spinal bupivacaine 10 mg for single-shot SA should not be reduced in morbidly obese parturients.

Includes bibliographical references / Background: A new appreciation of relevant risks, as well as the increased availability of technologies that facilitate the use of regional techniques, have increased the number of patients that are allowed to breathe spontaneously during their procedures. The ever-growing caseload of surgical patients in resource poor environments demands an anaesthetic service and equipment capable of meeting with these demands. Methods: Patients were recruited to receive their general anaesthesia by means of either the Universal Anaesthesia Machine (UAM) or the standard plenum system available. Anaesthesia was administered according to a protocol and the consumption of electricity, carrier gases, volatile hypnotic agent and carbon dioxide absorbent was measured. The cost per minute was then calculated for each device respectively. Results: Our study recruited 50 patients (25 into each group) across several surgical specialties. We found that when calculated as a total South African Rand (ZAR) per minute cost (for our centre) the UAM was statistically significantly more expensive (R 0.974/min vs. R 0.459/min, p < 0,00001). We were able to derive equations to predict the cost consumption of the respective devices, allowing the use of this data in a wide array of clinical settings. Conclusion: Whilst our finding is by no means surprising, it allowed us to produce formulae by which individual centres can calculate the implications of each option using the specific costs of the various consumables available to them.

Background: The increasing prevalence of obesity worldwide is a major threat to global health. Structural and functional changes in the heart are well documented for obesity as well as for pregnancy, but there is limited literature on the impact of obesity on cardiac function in pregnancy. We hypothesized that cardiac maladaptation to pregnancy occurs more frequently in otherwise healthy morbidly obese pregnant women than in pregnant women of normal body mass index (BMI). Methods: This prospective cohort study was performed in two referral maternity units in Cape Town, South Africa, over a 3-month period. Forty morbidly obese pregnant women (BMI ≥40 kg.m⁻²) (Group O) were compared to 40 pregnant women of BMI 30 kg.m⁻² (Group N). Cardiac structure and function were assessed by transthoracic echocardiography, according to the recommendations of the British Society of Echocardiography. Results: Acceptable echocardiographic images were obtained in all obese women. Statistical significance was defined as P< .005 after applying the Bonferroni correction for multiple comparisons. Mean [SD] mean arterial pressure was higher in Group O (91 [8.42] vs 84 [9.49] mmHg, P<0.001). There were no between-group differences in heart rate, cardiac output, or cardiac index (84 [12] vs 79 [13] beats.min⁻¹, P=0.103; 5447 [1048] vs 4740 [1183] mL.min⁻¹, P=0.006; 2551 [474] vs 2729 [623] mL.min⁻¹.m⁻², P=0.156, respectively). Stroke volume index was lower, and left ventricular mass higher in Group O (30.14 [4.51] vs 34.25 [7.00] mL.m⁻², P=0.003; 152 [24] vs 115 [29] g, P<0.001). Isovolumetric relaxation time was significantly prolonged in Group O (73 [15] vs 61 [15] milliseconds, P<0.001). The septal tissue Doppler index E' sept was lower in Group O (9.08 [1.69] vs 11.28 [3.18], P<0.001). There were no between- group differences in E' average (10.7 [2.3] vs 12.0 [2.7], P=0.018), or E/E' average (7.85 [1.77] vs 7.27 [1.68]). Conclusion: Obese pregnant women had a similar cardiac output and cardiac index to those with normal BMI. Their increased left ventricular mass and lower stroke volume index could indicate a limited adaptive reserve. Obese women had minor decreases in septal left ventricular tissue Doppler velocity, but the E/E' average values did not suggest clinically significant diastolic dysfunction.

Introduction The global rate of major noncardiac surgical procedures is increasing annually, and of those patients presenting for surgery increasing numbers are taking either an angiotensin-converting enzyme inhibitor (ACE-I) or an angiotensin receptor blocker(ARB). The current recommendations whether to continue or withhold ACE-I and ARB in the perioperative period are conflicting. Previous metaanalyses have linked preoperative ACE-I /ARB therapy to the increased incidence of post induction hypotension, however have failed to correlate this with adverse patient outcomes. The aim of this meta-analysis was to determine whether continuation or withholding ACE-I or ARB therapy in the perioperative period is associated with mortality and major morbidity. Methods This meta-analysis was prospectively registered on PROSPERO (CRD42017055291). A comprehensive search of MEDLINE (PubMed), CINAHL (EBSCO host), ProQuest, Cochrane database, Scopus and Web of Science was conducted on 06 December 2016. We included adult patients >18years, on chronic ACE-I or ARB therapy who underwent noncardiac surgery, where ACE-I or ARB was either withheld or continued on the morning of surgery. Primary outcomes included all-cause mortality and major cardiac events (MACE). Secondary outcomes included the risk of congestive heart failure (CHF), acute kidney injury, stroke, intra/postoperative hypotension and the length of hospital stay (LOS). Results Following abstract review, the full text of 25 studies were retrieved, of which nine fulfilled the inclusion criteria; five were randomized control trials (RCTs) and four cohort studies. These studies included a total of 6022 patients on chronic ACE-I/ARB therapy prior to noncardiac surgery. 1816 patients withheld treatment the morning of surgery and 4206 continued their ACE-I/ARB. Preoperative demographics were similar between the two groups. Withholding ACE-I/ARB therapy was not associated with a difference in mortality (odds ratio [OR], 0.97; 95% confidence interval [CI], 0.62-1.52; I2 =0%) or MACE (OR 1.12; 95% CI 0.82-1.52; I2 =0%). Withholding therapy was however associated with significantly less intra-operative hypotension (OR 0.63 95% CI 0.47;0.85, I 2 =71%). No effect estimate could be pooled concerning length of hospital stay and CHF. Conclusions This meta-analysis did not demonstrate an association between perioperative administration of ACEI/ARB, and mortality or MACE. It did however confirm the current observation that perioperative continuation of ACE-I/ARBs is associated with an increased incidence of intra-operative hypotension. A large randomized control trial is necessary to determine the appropriate perioperative management of ACE-I and ARBs.

Background: In high-income countries, preoperative anaemia has been associated with poor postoperative outcomes. To date, no large study has investigated this association in South Africa. The demographics of South African surgical patients differ from those of the European and Northern American surgical patients where the preoperative anaemia data are derived. These associations between preoperative anaemia and postoperative outcomes are therefore not necessarily transferable to South African surgical patients. Objectives: The primary objective was to determine the association between preoperative anaemia and in-hospital mortality in South African adult noncardiac, non-obstetric patients. The secondary objectives were to describe the association between preoperative anaemia and i) critical care admission, and ii) length of hospital stay, and to describe the prevalence of preoperative anaemia in adult South African surgical patients. Methods: We performed a secondary analysis of the South African Surgical Outcomes Study (SASOS) – a large, prospective, observational study of patients undergoing in-patient noncardiac, non-obstetric surgery at 50 hospitals across South Africa over a one-week period. To determine whether preoperative anaemia is independently associated with mortality or admission to critical care following surgery, we conducted a multivariate logistic regression analysis, which included all the independent predictors of mortality and admission to critical care identified in the original SASOS model. Results: The prevalence of preoperative anaemia was 1727/3610 (47.8%). Preoperative anaemia was independently associated with in-hospital mortality (odds ratio (OR) 1.66, 95% confidence interval (CI) 1.06-2.60, p=0.028) and admission to critical care (OR 1.49, 95% CI 1.08-2.05, p=0.015). Conclusion: Almost 50% of patients undergoing surgery at government-funded hospitals in South Africa had preoperative anaemia, which was independently associated with postoperative mortality and critical care admission. These numbers indicate a significant perioperative risk, with a clear opportunity for quality improvement programmes which may improve surgical outcomes. Long waiting lists for elective surgery allow time for assessment and correction of anaemia preoperatively. With a high proportion of patients presenting for urgent or emergency surgery, it behoves perioperative clinicians in all specialities to educate themselves in the principles of patient blood management.

Background: We compared the accuracy and precision of estimation of the systolic blood pressure (SBP) by disappearance of the pulse oximeter trace (DOT), with noninvasive blood pressure (NIBP) measurement, across the range of body mass index (BMI), during spinal anaesthesia (SA) for caesarean section (CS). Methods: Three groups of 25 parturients were recruited, with BMI of < 30- (Group 1), 30-40- (Group 2), and > 40 kg/m2 (Group 3) respectively. SBP was measured using the DASH® 3000 monitor (GE Health Care, UK) NIBP monitor, placed on the same arm as the pulse oximeter probe. Estimations of SBP were done before- and 5 minutes after induction of SA, during cuff inflation and deflation. The times taken for the estimations and the actual NIBP measurements were noted. Bland and Altman analysis was performed and the correlation coefficient estimated. Results: Concerning the most clinically relevant estimation, namely SBP during inflation post-SA: For Groups 1, 2 and 3, r = 0.56, 0.74 and 0.91; bias = -0.4, -2.9 and 0.8 mmHg, and limits of agreement = -27.7 to 26.9, -27.7 to 21.9 and -15.9 to 17.5 mmHg respectively. The mean (SD) time required for estimation of the SBP during inflation post- SA was 7.5 (1.1) s, 11.8 (3.8) s and 16.8 (4.2) s in the Groups 1, 2 and 3 respectively. The mean (SD) time required for measurement of post-SA SBP during inflations in Groups 1, 2 and 3, was 30.3 (13.1) s, 41.3 (10.2) s and 49.8 (14.6) respectively. In the post-SA period, mean time saved by estimating SBP during inflation was approximately 28.5 seconds, compared with 9 seconds during deflation. The time saved in Groups 1, 2 and 3 was 22.8 (13.2) s, 30.0 (11.6) s and 33.0 (15.6) s respectively. In the BMI > 40 kg/m2 group, the percentage error is ±13% of the mean systolic blood pressure observed, and the absolute error is ±16 mmHg, compared to ±27 mmHg in the normal BMI group. Conclusion: Post-SA estimation of SBP during cuff inflation in morbidly obese patients is more accurate and precise than in the other BMI Groups. Time to estimation is shorter than measurement by a clinically relevant period (33 s). This should improve patient safety in morbidly obese parturients.

This study was undertaken to assess the performance of the peri-operative pain management and describe the patient satisfaction following elective total abdominal hysterectomy cases at Rahima Moosa Mother and Child Hospital in Johannesburg, South Africa. The data obtained were collected utilising the questionnaire of the international pain registry PAIN-OUT, to allow for a standardised means of assessment and follow-up. All patients who presented for this procedure during a three-month period from September to November 2015 were approached on the first post-operative day. Patient demographics and details regarding the peri-operative interventions were collected from patient charts and a self-reported patient questionnaire was completed. The data obtained examined the personal pain experience as well as resultant functional limitations and emotional factors involved. Medication side-effects, patient opinions regarding appropriateness of their pain management, their degree of participation, whether information was offered to them and their overall satisfaction levels with care was sought. The study included 76 patients and highlighted unacceptable acute pain levels. It revealed deficiencies in administration of analgesics on the ward and a lack of standardised protocols or collaboration between Anaesthesiology and Gynaecology in managing acute postoperative pain. Patients were not uniformly advised regarding the post-operative expectations of their pain and non-pharmacological measures were not emphasised by the treating team. Despite these deficiencies, patients reported high levels of satisfaction with the care received. The information provided allows for a more focused quality improvement strategy to manage acute post-operative pain in this group more effectively and move toward better collaboration between departments involved in the care of patients peri-operatively.

Rationale: Major burn surgery can be associated with significant blood loss, often requiring transfusion of blood products. In an effort to decrease aforementioned blood loss, various blood conservation strategies have been developed, rendering older formulae to predict intraoperative blood loss ineffective and outdated. Currently there are no clear guidelines on when to transfuse burn victims but, the trend is towards employing a more conservative transfusion practice in an attempt to reduce transfusion related complications. The predicament has become one of containing cost by not ordering blood unnecessarily and/or excessively, versus putting a patient at risk by not having blood available when he or she needs it. A guideline, based on haemoglobin and extent of surgery, was drawn up at the Red Cross War Memorial Children's Hospital in an effort to rationalise preoperative blood ordering. The aim of this audit was to assess how well the implemented guideline was adhered to, and how accurately the guideline predicted the need to have blood products available in theatre during burns surgeries of varying extent. Methods: After a guideline, based on expert opinion, had been drawn up and implemented, a prospective audit of practice was done from April 2014 to June 2015. Two hundred separate burn surgeries were audited. Data collected included haemoglobin levels, extent of surgery, pre-and intra-operative instructions to blood bank, and whether patients were transfused. Pre-operative instructions were compared to the guideline to test adherence, and to the ultimate need for blood to test accuracy. Additional data recorded were the adherence to surgical plan (extent of surgery). Results: Five of the 200 cases were excluded due to incomplete data, leaving 195 cases. Blood was ordered according to the guideline in 131 (67.2%) cases. There were two groups where adherence was particularly poor. In these patients the guideline suggested that only a group and screen was necessary - a category for which it would also be difficult to assess how accurately the guideline predicts the need for blood. After excluding these two groups, the preoperative instructions to the blood bank were appropriate in 119 (94.4%) of the 126 cases where the guideline was followed. Blood was ordered preoperatively in 83 of the 195 cases, but only used in 50 cases (60.2%). Of the 33 cases where blood was not used, 23 cases were not in keeping with the guideline. In 50 (83%) of the 60 cases where blood was ordered according to the guideline, it was appropriate. The performed surgery proceeded as planned in 162 (83.1%) cases. Discussion: Blood transfusion exposes the recipient to transfusion-related risks and is expensive. In an attempt to avoid these risks there has been a trend towards conservative transfusion practices. It has been shown to be cost effective and safe to employ a restrictive transfusion practice during major paediatric burn surgery. During our study period one unit of blood cost R1096,00 and a group and screen R172,00. Significant savings could therefore be incurred if blood is ordered according to the proposed guideline. Conclusion: This audit confirmed that the guideline is an appropriate one to use for preoperative ordering of blood products for burns surgery at the Red Cross War Memorial Children's Hospital.

Clinical Anaesthesia is an indispensable adjunct to the surgical treatment of disease. It directly affects the lives of a great number of people since every year no fewer than one in every twenty-five persons in the population is required to undergo a procedure which necessitates it. Of itself non-therapeutic, clinical anaesthesia must, above all else. be safe for the patient. Any adverse outcome to clinical anaesthesia and factors relevant to its administration results, at best, in postoperative morbidity for the patient and, at worst, his demise. Identification and examination of the factors and circumstances which have a material influence on the safety of anaesthesia for the patient, provided the motivation for and is the central theme of this collection of research publications which was submitted to the University of Cape Town in fulfilment of the requirements for the award of the Degree of Doctor of Science (Medicine). The publications submitted report the results of forty years of endeavour. In terms of their subject matter, these publications may be broadly grouped into the following five fields of interest: - 1) The Epidemiology of Death Attributable to Anaesthesia. 2) Pharmacogenetic States of concern to the Anaesthetist - a) The Malignant Hyperthermia Myopathy. b) The Acute Porphyrias. 3) The Effects of Anaesthetics on the Liver - Studies of Hepatic Drug Metabolism of relevance to post-Halothane Hepatitis and the hepatotoxicity of anaesthetic agents. b) The treatment of Fulminant Liver Failure. 4) Heat Homeostasis during Anaesthesia - a) Inadvertent Hypothermia during anaesthesia. b) Induced Hypothermia during anaesthesia. 5) Miscellaneous. Within these fields of interest, papers have been grouped in terms of related aspects of the main topic they cover. Editorial comment is included where appropriate. The nature and scope of many of the above investigations was such as to require, for their satisfactory conclusion, collaborative interdisciplinary research combining the endeavours of other clinicians and paramedical scientists. Appropriate recognition of such collaboration has resulted in the multiple authorship registered for many of the publications in this collection.

Background: A number of anticoagulants are available in clinical use to preserve blood samples in liquid form until a suitable time for laboratory testing. Rotational thromboelastography is usually performed on a blood sample that has been anticoagulated with sodium citrate and then recalcified immediately prior to testing. In our institution we have had shortages of citrated Vacutainer® sample tubes. The use of a single in vitro anticoagulant promises to cut costs, simplify laboratory processes as well as limit the amount of blood drawn from patients. This together with the known problems with using citrate as an anticoagulant for viscoelastic testing (VET) prompted us to investigate the suitability of EDTA as anticoagulant for VET. Method: Blood samples from 20 healthy volunteers were divided into citrated and EDTA Vacutainer® tubes. A ROTEM EXTEM® assay was performed on each sample in both groups following the manufacturer's guidelines. Clotting time (CT), clot formation time (CFT), alpha angle (α-angle) and maximum clot firmness (MCF) results were compared. Ionised calcium concentrations were measured on each sample before and after recalcification with CaCl2 to determine if there was a significant difference in post - recalcification ionised calcium concentrations between the groups. Results: The results from the two groups were treated by Bland-Altman analysis. Apart from MCF values there was significant bias between all parameters measured in the two groups. The limits of agreement for all parameters apart from MCF were unacceptable. Conclusion: We found that ROTEM EXTEM® results from EDTA samples were not comparable to or interchangeable with those from citrated samples. The difference in results is not due to differences in ionised calcium concentration levels in the samples post-recalcification as the ionised calcium concentrations in both groups post-recalcification were adequate for coagulation. EDTA samples did show superior consistency in all parameters and may be a suitable alternative for sample preservation for VET if reference ranges can be established.