Reports from the maternal death notification system showed a high number of maternal deaths from anaesthesia in the Free State. I initiated this study in order to investigate and rectify factors identified as being associated with this problem. The study method used was an audit cycle. I selected from the literature on maternal deaths the five main factors most likely to be relevant ie manpower, use of regional anaesthesia, anaesthetic drugs and equipment, resuscitation protocols and the referral system. In the first part of the study (Phase 1), every level one and two hospital in the Free State in which Caesarean sections (CSâs) were being performed was inspected to determine whether the required standards for these factors were being met. The intention was to then implement remedial interventions to correct any problems identified in Phase 1 and then in Phase 2 to repeat the hospitals visits to assess the effect of the interventions. Phase 1 showed deficiencies in essential anaesthetic drugs and equipment in most hospitals. Hospital staff and managers were clearly informed regarding these problems but there was only a slight overall improvement in standards when the audit was repeated in Phase 2 and even a decline in certain aspects. Anaesthetic manpower was studied using self-completed questionnaires. Only a third of these questionnaires were returned and they showed that doctors were largely inadequately trained and supervised whilst administering obstetric anaesthesia. These results were presented to Free State Department of Health Top Management. Concern over possible bias regarding the doctors who returned the forms was raised and it was requested this study should be repeated using a different method to improve the response rate. This was done in Phase 2, a 69% response rate was achieved and the results confirmed the Phase 1 findings. Management was presented with these findings and interventions to resolve these problems was suggested to them. Use of regional anaesthesia was studied via inspection of theatre record books. In Phase 1 it was found that 71% of CSâs were performed using regional anaesthesia (RA), close to the proposed goal of 75%. However, in some hospitals no or few CSâs were performed using RA. There followed remedial interventions in which RA was promoted and in Phase 2 it was found that 84% of CSâs had been performed using RA. Unfortunately there is now concern that RA is being over utilised and general anaesthesia is not being administered when appropriate. In Phase 1, there were virtually no hospitals in which up-to-date resuscitation protocols were displayed. Protocols were distributed to all and in Phase 2 more, but not all, hospitals had the required protocols on display. The referral system was studied in Phase 1 using data collection sheets which were to be completed by the doctors when a patient had a CS. There was such poor compliance with this part of the study that it was decided that to resolve this would require a completely separate study utilizing greater research resources. Further investigation was therefore abandoned. Performing this audit had the desired result of improving the awareness of problems relating to obstetric anaesthesia in the Free State, at a Provincial and National level amongst both health care professionals and management. Although the remedial interventions did not âcureâ some of the problems identified, progress was made and where progress was not made there is now greater knowledge as to where the problems lie. It remains for those who have the resources to tackle these issues, which are largely concerned with manpower, to take cognisance of the results of this study and implement the suggested corrective measures.

Background: Low-cost, portable fingertip pulse oximeters are widely available to health professionals and the public. They are often not tested to ISO standards, or only undergo accuracy studies in healthy volunteers under ideal laboratory conditions. This study aims to pragmatically evaluate the agreement between one such device and a conventional bedside pulse oximeter in a clinical setting, in patients with varied comorbidities and skin pigmentations. Methods: A single-centre equipment comparison study was conducted. Simultaneous measurements were obtained in 220 patients with both a Contec CMS50D Fingertip Pulse Oximeter and a Nihon Kohden Life Scope MU-631 RK conventional bedside monitor. Peripheral oxygen saturations (SpO₂) and pulse rates were documented, and patient skin tone was recorded using the Fitzpatrick scale. Data was assessed using a Bland-Altman analysis with bias, precision and limits of agreement (LOA) calculated with 95% confidence intervals. A priori acceptability for LOA was determined to be 3%, in keeping with international standards. Results: Mean difference (therefore bias) between the conventional and fingertip oximeters for all data was -0,55% (95% CI -0,73 to -0,36%). Upper and lower limits of agreement (95% CI) were 2,16 (1,84 to 2,47) and -3,25 (-3,56 to -2,94) %. Regression analysis demonstrated worsening agreement with decreasing SpO₂. When samples were separated into “normal” (SpO₂ ≥ 93%) and “hypoxaemic” (SpO₂ < 93%) groups, the normal range displayed acceptable agreement between the two oximeters (bias -0,20 with LOA 2,20 to -2,27%), while the hypoxaemic group fell outside the study’s a priori limits. Heart rate measurements had mean difference (LOA) of -0,43 (-5,61 to 4,76) beats per minute. The study was not powered to detect difference among the skin tones, but demonstrated no trend for this parameter to alter the SpO₂ measurements. Conclusions: During normoxia, portable fingertip pulse oximeters are reliable indicators of SpO₂ and pulse rates in patients with various comorbidities in a pragmatic clinical context. However, they display worsening agreement with conventional pulse oximeters during hypoxaemia. Skin tones do not appear to adversely affect measurements.

Includes bibliographical references / Purpose: 1.To determine the effect of sevoflurane induction, followed by intravenous (IV) ketamine infusion on intraocular pressure (IOP) in the paediatric glaucoma population. 2. To establish the earliest time point at which IOP measurement most closely resembles awake values. Methods: A prospective, descriptive study of the IOP changes occurring in 25 children requiring IOP measurements at our institution. A standardised anaesthetic technique was employed; sevoflurane induction, intravenous cannulation, ketamine bolus (2mg/kg) and maintenance (4mg/kg/hr) for 15 minutes. IOP measurements and physiological variables were recorded after sevoflurane induction, then every 2 minutes for a period of 10 minutes, one at 15 minutes as well as 5 minutes after ketamine discontinuation. Results: IOP was measured in 25 patients (50 eyes). Twenty-six eyes (52%) had glaucoma. The mean patient age was 29 months (range 2-88 months). The mean IOP after sevoflurane induction was 3,68mmHg lower than that with ketamine maintenance (sevoflurane eliminated) (95% CI 1,35 to 6,02mmHg) (p=0,002)). Physiological variables return to baseline at 8 minutes, which correlates with the time taken for sevoflurane to be eliminated from exhaled gas. The difference in IOP between ketamine anaesthesia (time 15 minutes) and near wakefulness was only 0,28 mmHg (95% CI -2,23 to 2,79mmHg) (p=0,826). Mixed effects models showed similar trends but a higher baseline (7,85mmHg (6,19 to 9,51mmHg) (p<0,001)) in those with glaucoma when compared to those without. Conclusion: Sevoflurane lowers IOP significantly when compared to ketamine anaesthesia. While eyes with glaucoma had a higher baseline than those which did not have glaucoma, both groups follow similar trends in response to the anaesthetic agents. This standardised anaesthetic protocol allows reliable IOP measurement 15 minutes after termination of sevoflurane and commencement of ketamine infusion, with no reported adverse events.

Background: Preoperative assessment clinics have been employed in many institutions to manage perioperative risks1. These clinics provide an opportunity to stratify patients on the basis of risk prior to surgery, to make timely multidisciplinary referrals where appropriate, and to prescribe medical therapies according to the current best evidence resulting, in fewer last-minute cancellations for medical reasons2 and a shorter inpatient pre-operative stay3. The Anaesthetic Pre-Operative Clinic (APOC-GSH) was introduced to Groote Schuur Hospital in 2009 with the aim of assessing and optimising high risk patients undergoing intermediate or high risk surgery. The vision of the clinic is to decrease perioperative morbidity and mortality, rationally and costconsciously investigate patients, as well as to reduce theatre cancellations of inappropriately assessed and managed patients. The clinic, together with relevant role players, attempts to risk stratify patients in making an informed decision whether the intended perioperative risks are acceptable. Patients referred to the clinic fall into two categories. Either the surgical date has been scheduled and patients are referred to APOC-GSH for optimization, or the referral is for assessment of suitability to undergo an anaesthetic and thus the treatment modality hinges on the fitness for surgery. Objective The intention of the audit is to create a database of the patients seen at APOC-GSH during 2014. Primary objectives are: 1 Referral pattern: - What proportion of patients are referred for improvement of medical condition? - What proportion of patients are referred for an assessment of operability? 2 What interventions were recommended? - What was the influence of interventions on operability and timing of surgery? - Does the clinic improve theatre efficiency by reducing cancellations of patients who required further interdisciplinary discussion and investigations? Methods: Data collection will be based on the review of APOC-GSH clerking notes and hospital patient records. The following information will be collected and entered into an Excel spread sheet: the surgical discipline referring the patient, the proposed surgery, patient co-morbidities, the lead-time from the first APOC-GSH assessment to surgery, number of visits to APOC-GSH by each patient, the investigations and additional interventions instituted at APOC-GSH. We shall also calculate the proportion of patients that would have likely been cancelled had they not attended APOC-GSH but rather been assessed by an anaesthetist the day before surgery. The record of the APOC-GSH consultation will be documented on a clerking sheet established for use within the clinic (Addendum A). The original form will be included in the patients' file and a duplicate will be stored within the access controlled Department Of Anaesthesia offices at GSH. The duplicate records will be used to obtain the information for the audit. Ethical approval will be sought from the UCT Human Research Ethics Committee for the establishment of the database and retrospective review of the APOC-GSH records on the understanding that patient records would be kept confidential and that the data obtained would be stored in a password protected spreadsheet. Access to the spreadsheet will be limited to the investigators involved and no identifiable patient details will be included. Patients will be counselled as to the nature of the study and will be expected to submit signed consent forms allowing their records to be reviewed. Patients will be informed that they will not be disadvantaged by refusal to sign the consent form and that the appropriate standard of care will still be applied. No remuneration will be provided for partaking either. Consent will be documented on the consent form specifically designed for use in the APOC-GSH. (Addendum B) Statistical analysis will be performed using an Excel® spreadsheet. Means, medians, rates and percentages will be used to describe the discrete categorical data. Output and future work The audit will provide an objective assessment of the population profile referred to APOC-GSH. It may guide future implementation of APOC-GSH protocol changes and assist with resource allocation depending on the surgical discipline requirements. Patient and surgeon satisfaction ratings may be embarked at a later stage juxtaposed against an adequate assessment in a cost and time conscious manner.

Background Anaesthesia for ceasarean section (CS) in women with eclampsia is a major clinical challenge, and there are limited data concerning the rationale for the choice of technique, and the short-term outcomes. A retrospective audit was performed on practice at a tertiary referral centre in Cape Town. Methods The primary outcome of the audit was the proportion of patients with eclampsia receiving either spinal anaesthesia (SA) or general anaesthesia (GA) for CS, and the assessment of the rationale for the choice of method. A convenience sample of the records of one hundred consecutive patients with eclampsia admitted to the Maternity Centre at Groote Schuur Hospital from 2009- 2013, were studied. Patient demographics, clinical details and outcomes (including relevant results of special investigations), were recorded. The number and percentage of patients receiving GA or SA were documented. The presence or absence of factors influencing selection of the method of anaesthesia, including Glasgow Coma Scale (GCS), degree of urgency of CS, trauma to the oropharynx, number of seizures, control of blood pressure, and the availability of and/or platelet count prior to anaesthesia, were noted. In addition, complications of anaesthesia were described, and short-term maternal and neonatal outcomes were recorded. Results One hundred patient records were screened, and there were 11 exclusions; therefore 89 patient records were analysed. Seven/89 (7.9%) received SA and 82/89 (92.1%) GA. One patient required conversion to GA because of the occurrence of a seizure during attempted SA. All patients receiving spinal anaesthesia were stable and had no postoperative complications. Overall, 63 (70.8%) had a preoperative GCS< 14, and 26 (29.2%) patients had GCS ≥14. Seven patients in the subgroup with GCS ≥14 had SA, and the remaining 19 received GA. GA was performed because there was no platelet count available in 3, pulmonary oedema in 2, difficult airway due to a bitten tongue during a seizure in 2, fetal bradycardia in 2, HELLP syndrome in 1, renal failure in 1, and patient refusal in 1 patient. In 7 women, there was no clear reason why GA was chosen. The median [IQR] Apgar scores at 1 minute were lower in patients receiving GA [(5) (1-9) and 4 (1-9)] in patients with GCS ≥14 and < 14 respectively, versus 8 (6-9) for SA. The number (%) of neonates with Apgar scores ≥ 7 at 1 minute in patients who had SA, was 6 (85.7), compared with 3 (13.6) and 12 (17.9) in patients having GA with GCS ≥14 and < 14 respectively. There were 2 stillbirths and 2 neonatal deaths in the GA group. The median [IQR] length of stay in patients with GCS ≥ 14 was 1 day in the group receiving SA and 2 days in those who had GA; patients with GCS< 14 were ventilated for 3 days. No patients receiving SA required postoperative ventilation, whereas 5/19 (26.3%) patients with GCS ≥14 who received GA required ventilation, for 2 days. Seven patients with GCS< 14 had cerebral oedema, and 2 had cerebral infarcts. There were 2 maternal deaths. Conclusions In this audit, a small percentage of women with eclampsia received SA for CS. A review of the case records suggests that more patients could have safely received SA. A larger prospective audit is required to establish factors influencing safety of spinal anaesthesia for CS in women with eclampsia.

The incidence of intraoperative nausea and vomiting during spinal anaesthesia for caesarean section. Background: Nausea and vomiting (IONV) during spinal anaesthesia (SA) for caesarean section (CS) is unpleasant and may interfere with surgery. We studied the incidence of IONV during elective CS, as well as the influence of ethnicity on this outcome. Methods: Two hundred and fifty eight healthy term patients undergoing SA for elective CS were recruited to this prospective observational study conducted at two Cape Town Level 2 hospitals. Standard practice was employed for SA for CS at UCT (University of Cape Town): 2 mL hyperbaric bupivacaine plus 10 μg fentanyl at the L3/4 interspace, and 15 mL/kg cry stalloid co - load. Spinal hypotension was managed with phenylephrine boluses according to a standard protocol. Nausea and/or vomiting were treated by restoration of blood pressure, and metoclopramide. Intraoperative complaints of nausea, and vomiting, were noted. Patients were also interviewed postoperatively as to any experience of intraoperative - or previous history of nausea. Results: Of the 258 patients enrolled in the audit, 112 (43.4%) were non - African and 146 (56.6%) were Black African patients. The overall incidence (95% CI) of nausea was 32% (0.27 - 0.38), with 20% occurring prior to - and 11% after the delivery. The overall incidence of vomiting was 7% (0.05 - 0.11), with 3.2% occurring prior to, and 3.8% after, delivery. The incidence of nausea and/or vomiting was 33% (0.28 - 0.40). Black Africans experienced significantly less nausea than non - African patients (36/145 [24.8%] vs 47/112 [42.0%] respectively, p = 0.004). There was no significant difference in the incidence of vomiting (10/14 5 [6.8%] vs. 8/112 [7.1%] respectively, p = 0.865). The odds of experiencing intraoperative nausea for patients with any blood pressure value <70% of baseline, were 2.46 (95% CI 1.40 - 4.33). Conclusions Though in keeping with international standards, the clinically significant incidence of nausea and/or vomiting of 33% requires adjustments to the management protocol for spinal hypotension. The inclusion of ethnicity as a risk factor for nausea during SA for CS should be considered.

Background There is an unmet need for essential surgical services in Africa. Limited anesthesia services are a contributing factor. Non-physician anesthesia providers are utilized to assist with providing anesthesia and procedural sedation in order to make essential surgeries available. There is a paucity of data on outcomes following procedural sedation for surgery in Africa. We investigated the postoperative outcomes following procedural sedation by non-physicians and physicians in Africa. We hypothesized that the level of training of the sedation provider may be associated with the incidence of severe postoperative complications and death. Methods A secondary analysis of a prospective cohort of in-hospital adult surgical patients, representing 25 African countries was performed. An inverse probability of treatment weighting model was developed to assess the association between receiving procedural sedation conducted by a non-physician (vs physician) and in-hospital outcomes. All patients who only received procedural sedation for surgery were included. The primary outcome was the incidence of the composite of severe complications and death. Results 336 patients met the inclusion criteria, of which 98 (29.2%) received sedation from a non-physician provider. The incidence of severe postoperative complications and death was 10/98 (10.2%) in the non-physician group, and 5/238 (2.1%) in the physician group. The association between procedural sedation conducted by a nonphysician provider and in-hospital outcomes showed an eight-fold increase in the odds of severe complications and death (odds ratio 7.7; 95% CI 2.5 to 23.7). Conclusions The modest number of observations in this secondary data analysis, suggests that shifting the task of procedural sedation from physicians to non-physicians in order to increase access to care may be associated with severe postoperative complications and death in Africa. Research focusing on identifying factors contributing to adverse outcomes associated with procedural sedation is necessary to make this practice safer.

Background: The Histidine-tryptophan-ketoglutarate (HTK) solution is cardioplegic solution that confers myocardial protection during periods of ischemia in heart surgery. It has the advantage of allowing a longer protection and it is especially the preferred cardioplegic solution in complex paediatric heart surgery like transposition of great vessels. This has also been used in organ transplant as a preservative solution. Recently concern has been raised over its safety because of the increased incidence of transplant failure due to thrombosis in individuals whose transplanted organ were preserved with HTK solution. The main purpose of this study was to establish whether the HTK solution in the dose and volume used for myocardial protection during paediatric heart surgical procedures increases the thrombotic risk of patients. Patient and methods: This study was an experimental study conducted at the coagulation laboratory of the department of anaesthesia of the University of Cape Town. The human research ethical committee of this university approved this study. Twenty healthy individual were recruited for this study and each one of them read and signed the consent form before being enrolled for the study. The exclusion criteria were any known coagulopathy, liver diseases, alcoholism, kidney disease and intake of any drugs known to impair the coagulation. Ten millilitre of blood was drawn from the volunteer at the coagulation laboratory. The blood was diluted with either HTK solution or St Thomas' cardioplegic solution at 9:1 ratio to mimic the hemodilution due to HTK solution in paediatric population. The effect on coagulation was assessed by means of thromboelastography run on the native blood and each dilution separately. Results: The difference in the magnitude of change from baseline in r-time between the HTK solution and Saint Thomas' solution (mean difference 0.49 min) (p 0.014) was statistically significant. There was no statistically significant change in MA between the HTK solution and St Thomas' solution (mean difference -2.13 mm) (p 0.165). Compared to native blood there were no significant change in r-time with the HTK solution (mean -0.215 min), the same with St Thomas' solution. Compared to native blood, there was significant change in maximum amplitude with the HTK solution (mean 1.38 mm) the same with St Thomas' solution (mean 3.51). Although the difference in the magnitude of in R time change between HTK solution and St Thomas' was statistically significant, it is clinically not relevant. The data did not show a trend that might become significant with large sample. All variables showed a slight tendency towards decreased coagulation in the diluted samples, but not of sufficient magnitude to be clinically important. Our goal was the use of a dilution that is clinically relevant in the context of paediatric cardiac surgery. Conclusion: The results of this study suggest the HTK solution at 10% dilution does not cause significant changes in coagulation parameters. In comparison to normal saline, there were a tendency toward hypocoagulation. More research in this field is needed to clarify the pathophysiologic pathways of thrombosis observed in these transplants.

Background: Studies in healthy patients undergoing elective caesarean delivery show that ephedrine used for spinal hypotension is associated with increased fetal acidosis compared with phenylephrine. This has not been investigated prospectively in severe preeclampsia. Methods: Patients with severe preeclampsia requiring caesarean delivery for a non- reassuring fetal heart tracing were randomised to receive bolus ephedrine (7.5-15 mg) or phenylephrine (50-100 μg) for spinal hypotension. The primary outcome was umbilical arterial base deficit. Secondary outcomes were umbilical arterial (UA) and venous (UV) pH and lactate level, venous base deficit, and Apgar scores. Results: A total of 133 women were included;; 64 required vasopressor treatment and were randomised to 2 groups of 32 with similar patient characteristics. Pre- delivery blood pressure changes were similar in the 2 groups. There was no difference in mean [SD] UA base deficit (-4.9 [3.7] vs -6.0 [4.6] mmol·L⁻¹ for ephedrine and phenylephrine respectively;; P = 0.29). Mean [SD] pH (UA and UV) and lactate levels were also similar between groups (7.25 [0.08] vs 7.22 [0.10], 7.28 [0.07] vs 7.27 [0.10], and 3.41 [2.18] vs 3.28 [2.44] mmol·L⁻¹ respectively). In addition, UV PO₂ was higher in the ephedrine group (2.8 [0.7] vs 2.4 [0.62]) kPa, P = 0.02). There was no difference in 1- or 5-minute Apgar scores, numbers of neonates with 1-minute Apgar scores < 7 (10/32 [31%] vs 12/32 [38%]), or with a pH < 7.2 (6/31 [19%] vs 8/29 [28%]). Conclusions: In patients with severe preeclampsia and fetal compromise, fetal acid-base status is independent of the use of bolus ephedrine vs phenylephrine to treat spinal hypotension.

Includes bibliographical references / Background: The vast majority of patients will be admitted to general wards after their surgical procedures. Ward staff will provide the prescribed analgesia. The researchers would like to ascertain whether the patient population is satisfied with the analgesia that they receive. Methods: Fifty-two postoperative patients consented to taking part in a prospective audit that enquired about pain using a Numeric Rating Scale (NRS) on discharge from the theatre recovery room as well as on day one postoperatively. Additionally patients were asked to indicate whether the analgesia was 'good', 'fair' or 'poor' and were interviewed about their expectations regarding pain. Results: The mean age was 45 (SD 14) years and median surgical duration was 100 (IQR 75- 150) minutes. Mean NRS score was 3 (SD 3) on discharge from recovery as well as on day one postoperatively. 'Good' analgesia was reported by 69.2% of patients and 71.2% reported that they had less pain than expected. The median time from recovery room discharge to first dose of analgesia was 135 (IQR 65-400) minutes. Conclusion: Sixty seven per cent of patients indicated that they were satisfied with the analgesia provided. There are, however, still problems with long waiting times to first doses of analgesia. The relatively low overall pain scores and high levels of satisfaction are encouraging.