The subject matter of this thesis might appear to consist of eleven separate essays. These have been grouped under the headings repair, by-pass and replacement. However, it is believed that the same theme might be apparent throughout, based chiefly on the following considerations: a. Retrospective survey of case notes forms no part of this thesis. It is a record of personal observation of every patient mentioned. The evolution of ideas and techniques is the result of close observation of the progress of these patients, sometimes modified after a careful evaluation of the observations of other workers as reported in the medical literatures b. At no time was a consideration of the patient, as an integrated whole, and the evaluation of the entire urinary tract, sacrificed in the approach to a study of the more localized pathology. It is firmly believed that such an omission will be to the detriment of the patient. Whatever part of the urinary tract is involved, therefore, the approach to the problem remains fundamentally the same. c. In the management of these cases there has always been the closest cooperation between the urological surgeon and his colleague in gynaecology. This has always ensured an objective and critical check in all decisions regarding management of cases. Such a close combination of two disciplines has promoted further uniformity of approach to problems varying in detail. Therefore, although the subject matter has been subdivided into eleven sections, the problem in them all remains basically the same and there is considerable overlap between the sections. only to facilitate systematic discussion.

Includes bibliographical references.

Background. A multitude of data exists regarding global incidence of OASI as well as its contributing factors and complications. Little to no data exists regarding the incidence of OASI or its complications and the nature of these complications within South Africa. Objective. To describe the rate and nature of complications of OASI that occur within 6 weeks following primary repair of an OASI, followed up at the GSH perineal clinic. Secondarily, to investigate the incidence of OASI and follow up rate post primary repair Methods. This was a retrospective chart review. Participants were identified from theatre record books between January 2014 and December 2015. The charts of those that attended the perineal clinic follow up were reviewed and complications and possible associated risk factors were identified from the clinical notes. Primary aims were to identify the incidence and nature of complications seen in this population as well as possible related risk factors. The secondary aims were to determine the incidence of OASI and follow up rate for complication following primary OASI repair. Results. The mean age of participants was 25.85 years with a mean body mass index of 25.15kg/m2 . The mean birthweight seen was 3382.05 grams. Constipation (10.87%), pain with defaecation (11.96%) and anal incontinence (10.87%) were the most frequently reported complications. Wound infection was found on examination in 3.26% of participants and wound dehiscence was seen in 6.67%. Incidence of OASI in this study group was 8.64 per 1000 vaginal deliveries. The follow up rate of these participants was 26.20%. A total of 374 OASI were repaired within this region during the study period. Only 97 of these attended follow up, for which 93 folders were available to be included in analysis Conclusion. The Incidence of OASI in this Western Cape region is within the range seen worldwide but the proportion of complications seem to be less than global data suggests despite adverse conditions, theatre delays and the fact that repairs were mostly performed by specialists in training. It was not possible to identify any relationship with possible associated factors. There is a very poor follow up rate within this community which needs to be explored and systems need to be put in place to ensure all participants are given the opportunity of follow up.

Background: Venous thromboembolism is one of the leading causes of morbidity and mortality around the world. In addition to the immediate morbidity, there is significant implications on delivery plans, future options of contraception and thromboprophylaxis in subsequent pregnancies. At present, no pre-test probability assessments are being used to predict venous thromboembolism in pregnancy. This is the first study in South-Africa, addressing venous thromboembolism in the perinatal period which specifically examines the epidemiology and clinical presentation in pregnancy and the post-partum period. Objectives: To determine the incidence of venous thromboembolism in Groote Schuur Maternity Centre and to identify specific variables in the clinical presentation that had a predictive value of a thromboembolic event. Methods: A quantitative, retrospective study with a descriptive comparative research design, for a twelve-month period from January 2016 to December 2016. All pregnant and postpartum patients who were sent for a venous duplex ultrasound, ventilation perfusion study or computerized tomography pulmonary angiogram from the Groote Schuur Maternity center were included. A folder review was conducted and the diagnosis and clinical presentation of all the patients were documented and analyzed. Incidence of VTE were estimated as the number of events per 1,000 deliveries. The number of hospital deliveries in 2016 were used as the denominator for calculating this incidence. Results: A total of 41 (0.12%) patients had a venous thromboembolism. Six patients had a deep venous thrombosis (0.02%) and 37 had a pulmonary embolism (0.11%). Among the 186 retrieved medical records, 11 (28%) of the diagnosis occurred in the puerperal period and 28 (72%) during pregnancy. Among the 28 events during pregnancy, one (3%) was in the first trimester, nine (23%) in the second trimester and 18 (46%) in the third trimester. The majority of confirmed pulmonary emboli (72.22%) and deep venous thrombosis (66.67%) were diagnosed during the third trimester in pregnancy. Among individuals with deep venous thrombosis, the most frequently reported symptoms and signs were leg pain (66.7%), leg swelling (66.7%) and tachycardia (66.7%). Patients without deep venous thrombosis presented more with leg swelling (76.3%), red discolouration (10.5%) and cellulites (10.5%). The only presenting clinical features that were significantly different were haemoptysis (p=0.01) and coughing (p=0.03). Among those individuals without pulmonary embolus, tachycardia (77.3%) and dyspnoea (49.1%) were commonly reported. Among the patients with a PE, the most frequently reported symptoms were tachypnoea (78.4%), dyspnoea (64.9%), tachycardia (62.2%), chest pain (51.4%) and coughing (46%). Features in the clinical presentation that were statistically significant were chest pain (p=0.01), haemoptysis (p=0.07), tachypnoea (p=0.01) and tachycardia (p=0.03). The greatest statistically significant clinical feature was the symptom of coughing (p< 0.01). The stepwise logistic regression for the univariate analysis showed that coughing (OR=3.83; 95% CI: 1.71 to 8.58; P< 0.01), chest pain (OR=2.57; 95% CI: 1.2-5.53; P=0.02), tachycardia (OR=1.03; 95% CI: 1.0 to 1.06; P=0.03), tachypnoea (OR=1.06; 95% CI: 1.0 to 1.12; P=0.05) and a median symptom of 3.5 (1.58; 95% CI: 1.23 to 2.06; P< 0.01) were the best explanatory variables. The stepwise logistic regression for the multivariate analysis showed that both tachycardia (OR=1.03; 95% CI: 1.0 to 1.06; P=0.03) and coughing (OR=3.43; 95% CI: 0.88 to 11.30; P=0.05) predicted a positive pulmonary embolus. A logistic regression for tachycardia showed a 23% increase in pulmonary embolus for every increase of 5 beats per minute in the heart rate above 100Bpm. This association was statistically significant (OR=1.23; 95% CI:1.08 to 1.39; P=0.0004) A logistic regression analysis of the association between tachycardia, tachypnoea and chest pain and the risk of having a pulmonary embolus showed a 4% increase in the risk of pulmonary embolus for every single unit increase in heart rate. When controlling for tachycardia and tachypnoea, chest pain was also associated with a 3.8 times increase in the odds of having a pulmonary embolus. This association was statistically significant (p=0.0002) Conclusion: In this study, we found that the incidence of venous thromboembolism in the Groote Schuur Maternity Centre was the same as in other developed and developing countries around the world. The majority of confirmed venous thromboembolisms were diagnosed during the third trimester in pregnancy. This study found a lower incidence of deep venous thrombosis in comparison to other studies. The clinical features that had some predictive value for pulmonary embolism were chest pain, coughing, tachypnoea, tachycardia and more than three symptoms or signs. Tachycardia was significant in the univariate-, multivariate analysis and stepwise logistic regression. In addition, there was a statistically significant association between tachycardia, tachypnoea and chest pain and the risk of having a pulmonary embolus. This study has revealed the need to develop pre-assessment algorithms in pregnancy and postpartum patients to reduce maternal and fetal, morbidity and mortality. Until such algorithms are developed, clinicians should use their own clinical judgment and proceed to diagnostic imaging for suspected VTE, where indicated.

Poor invasion of trophoblast cells in early pregnancy has been associated with preeclampsia and intrauterine growth restriction as well as other adverse pregnancy outcomes such as miscarriage, preterm birth and intrauterine death. Hypertensive disorders of pregnancy, including pre-eclampsia are one of the leading causes of maternal mortality in South Africa (Third report on Confidential Enquiries into Maternal Deaths in South Africa (2002-2004)) and the rest of the world. The currently accepted mechanism underlying the development of preeclampsia implicates poor trophoblast invasion and inadequate transformation of the maternal spiral arteries. Despite extensive research in this area, the control of trophoblast invasion and early placental development remains poorly understood. A whole host of factors such as oxygen tension, activation of matrix metalloproteinases (MMPs), angiogenic factors (VEGF-A) and immunological factors such as TNF alpha, interleukins and TGFβ have been shown to be involved in the control of trophoblast invasion. Our knowledge of the molecular details of pregnancy is unfortunately limited to in-vitro experiments and animal studies. Recently kisspeptins and their cognate receptor GPR-54 originally involved in tumour metastasis suppression and regulation of puberty, have been implicated in the inhibition of trophoblast invasion. Expression levels of kisspeptin and its receptor in trophoblast cells are highest in the first trimester, when control of trophoblast invasion is critical, and lower towards term.

Background: The association between maternal haemoglobin concentration and the outcome of pregnancy has been a source of continual controversy. Preterm delivery and low birth weight are major causes of stillbirths and early neonatal deaths. Pre-eclampsia is a major complication which occurs during pregnancy and leads to significant maternal and fetal morbidity and mortality. This study aims to assess the association between maternal haemoglobin concentration and pregnancy outcome. Objectives: To assess the association between maternal haemoglobin concentration and pre-eclampsia, preterm birth and low birth weight. Methods: Retrospective analysis of 191 patients who delivered an infant of 28 weeks gestation or more from 1st to 22nd May 2017 at New Somerset Hospital with documented maternal haemoglobin concentration at 22 – 33 weeks gestation. The maternal haemoglobin concentration used was the lowest documented during 22 – 33 weeks gestation. Main outcome measures: Development of pre-eclampsia, low birth weight and preterm birth. Results: The distribution of pre-eclampsia, low birth weight and preterm birth were skewed towards the higher side of the maternal haemoglobin concentration spectrum. There were statistically significant differences in the development of pre-eclampsia, low birth weight and preterm birth between maternal haemoglobin concentrations =>13g/dl and that of < 13g/dl. Pre-eclampsia was associated with relatively lower birth weight, preterm birth and low placental weight independently of the haemoglobin status. Conclusion: High maternal haemoglobin concentration at 22 – 33 weeks gestation is associated with an increased prevalence of pre-eclampsia, low birth weight and preterm birth.

Introduction Advanced abdominal pregnancy (AAP) is a pregnancy of over 20 weeks gestation, with a foetus living, or showing signs of having once lived and developed, in the mother's abdominal cavity. It is a rare obstetric complication associated with high maternal and perinatal morbidity and mortality. The question of whether to leave the placenta in or to remove it has been the subject of debate. When such cases present outside high resource settings, where a multi-disciplinary approach may not be possible, there are many challenges to effective treatment. The management of this rare but serious complication of pregnancy at Rob Ferreira Hospital (RFH), a rural provincial tertiary hospital, in Mpumalanga, was investigated. Methods This was a retrospective observational study of patients with AAP identified from theatre registers from January 2011 to January 2018 at RFH. Data was collected on patient demographics, diagnostic challenges, preoperative evaluation, and surgical management particularly of the placenta, and outcomes. Delivery data for RFH was retrieved from the provincial database. Results There were 26 cases of AAP identified, for which 21 folders could be retrieved from the registry. The incidence of AAP was 66.2 per 100, 000 deliveries. Abdominal pain was a presenting complaint in 100% of patients. Ultrasound scan (USS) accuracy, prior to surgery, was eventually 90.5%. However, for 61.9% the diagnosis of AAP was missed at initial USS and for 33% of these, AAP was only diagnosed after failed attempts at induction of labour (IOL). AAP was diagnosed intra-operatively in 9.5%, for a supposed caesarean delivery and 19 (94.7%) were diagnosed pre-operatively. Intra-operatively, 36.8% patients had placental site bleeding for which partial placental removal was 2 necessary in 71.4%. This group had more adverse maternal outcomes. For 73.7% of patients there was no bleeding from placental bed and the placenta was left in situ; but 7.1% required relook laparotomy for haemorrhage. There was one maternal death. Overall neonatal survival rate was 14.3% and pregnancy loss rate was 71.4%. Conclusion This study shows that planned management of AAP in a resource constrained setting with tailored approach and management, can be performed without compromising maternal outcomes. However, poor outcomes occurred when pre-operative diagnosis was not made and AAP was discovered during emergency CS, emphasising the importance of antenatal diagnosis. USS and a high index of suspicion remains the best diagnostic tool. From this study it appears safe to leave the placenta undisturbed unless it can be safely removed, or the patient is already bleeding from the placental site. There were very low neonatal survival rates, and this needs to inform counselling of patients with AAP.

BACKGROUND Recurrent disease after cervical excisional procedures has been linked to many factors. We aim to determine if positive margins and crypt involvement increased the rate of recurrence of premalignant disease in patients who had excisional procedures. METHODS In this retrospective review of the colposcopy database, patient records and pathology database, women who had cervical excisional procedures at the Groote Schuur Hospital colposcopy clinic in 2010 were followed up until 2015. Recurrence was based on high grade cytology or histology at follow up. Chi-square tests were used to compare recurrence rates. RESULTS Two hundred and seventy women were included in the final analysis. 130 women had CIN 3 and 94 had CIN 2 at the excisional procedure. Eighty five (31.5%) had endo-margin involvement, 46 (17%) had ecto-margin involvement, and 24 (8.9%) had dual margin involvement. Two hundred and thirteen (79.2%) had crypt involvement. Recurrence occurred in 30 (19.4%) of the 155 patients we had follow up data on. Of those that recurred, 19 (P<0.001) had positive endo-margin involvement, 10 (P=0.007) had ecto-margin involvement, 9 (P< <0.001) had dual margin involvement, and 28 (P=0.058) had crypt involvement. 155 women (43%) were lost to follow-up CONCLUSION Positive margins at excisional procedure of the cervix have a statistically significant increased risk of recurrence of pre-malignant disease. There was a trend towards recurrence of disease in those who have crypt involvement. In limited resource setting follow up protocols can be adjusted so that women without margin involvement can be seen at longer intervals.

Includes bibliographical references (leaves 231-252). / Infertility is a common reproductive health problem in Africa. The experiences of men and women who are unable to conceive, their constructs of infertility, their motives for parenthood, and their health-seeking behaviour are, however, inadequately documented in South Africa. In order to improve our understanding of the patients' perspective of infertility, seven studies were conducted employing both qualitative and quantitative research methodology. Study participants were recruited from the infertility clinic at Groote Schuur Hospital, Cape Town, a tertiary referral centre within the public health care system. The central themes explored in the qualitative studies included reproductive health knowledge, health-seeking behaviour, barriers to health care, experiences related to involuntary childlessness, and the reality of infertility and HIV infection. Data from both men and women were collected through in-depth interviews, and the results were analysed according to the principles of descriptive analysis. In the quantitative studies psychological distress was measured and motives for parenthood were assessed with the use of two standardised instruments ( the Symptom Checklist-90-R for the measurement of acute psychological symptom status and the parenthood motivation list). In addition, participants' attitudes towards reproduction in HIV-infected individuals were evaluated. Standard statistical methods were used to analyse quantitative data. The results of these studies demonstrated that men and women had limited knowledge about fertility, infertility, and biomedical infertility management. Some men and women held traditional beliefs and had accessed traditional healers. Most informants appeared highly motivated to engage in biomedical infertility management. Treatment satisfaction varied and reasons for non-compliance were both service and patient-related. Infertile couples gave many reasons for wanting a child and expressed a strong desire for parenthood. For many men and women the inability to conceive was associated with negative emotions, marital instability, abuse, stigmatisation, and loss of social status. Psychological distress levels were significantly higher in infertile women when compared to women using contraception, and in infertile men when compared to fertile men. Infertile women who reported intimate partner abuse were particularly distressed. The diagnosis of HIV infection did not eliminate the wish for a child in infertile couples, and in the absence of medical assistance many continued to attempt conception. The concomitant experience of infertility and HIV infection was associated with considerable suffering The majority of HIV-negative, infertile men and women opposed reproduction in HlV-positive couples. Collectively, the results of these studies provide new insights into the manner in which men and women who access the public health system in South African construct, experience, and respond to infertility. Understanding those details of the patients' perspective should improve the management of infertility in this patient population.

Includes bibliographical references (leaves 29-32). / Venous thromboembolism (VTE) is a major cause of maternal mortality worldwide. In many developed countries, all maternal deaths are investigated, and accurate statistics are available. In United Kingdom (UK) for example, VTE is the leading cause- and is increasing despite heightened awareness of risk factors, and wider use of thromboprophylaxis (1,2). The 1994-96 UK Confidential Enquiries reported an overall maternal mortality rate of 12.2 per 100, 000 deliveries, with specific mortality from VTE at 2.2 per 100, 000 deliveries, with approximately 15 deaths a year due to VTE.