Browsing by Author "Davies, Mary-Ann"

Now showing 1 - 20 of 41
  • Thumbnail Image
    Item
    Open Access
    A longitudinal analysis of neonatal and infant diagnostic HIV-PCR uptake and associations during three sequential policy periods in Mitchell’s Plain, Cape Town
    (2018) Kalk, Emma; Davies, Mary-Ann
    Background: Despite technological and programmatic advances in the prevention of vertical transmission of HIV and early infant diagnosis (EID), paediatric HIV continues to have a significant impact on infant and child survival in low- and middle-income countries. Many EID programmes follow the WHO recommendation of initial infant HIV testing with a nucleic acid assay at 4-6 weeks old. In general this strategy has been poorly implemented with substantial attrition after birth such that, according to UNAIDS, only 51% of HIV-exposed infants received a virological test in the first two months of life in 2015. In addition, there is concern about the sensitivity of the nucleic acid assays at six weeks in the context of exposure to prolonged multidrug antiretroviral therapy as infant post-exposure prophylaxis, and in breast milk. HIV-PCR testing at birth has been promoted as a means of maximizing the number of infants who receive an HIV test as well as identifying in utero-infected infants in whom HIV infection may follow an aggressive course. Evidence from pilot studies and modelling data was sufficiently compelling for the WHO to include a conditional recommendation for the addition of a birth HIV-PCR (either routine or targeted at high risk groups) to its EID algorithms in 2015. The Western Cape introduced targeted birth HIV testing for high risk infants in August 2014 and expanded this in line with the South African National Prevention of Mother-to-Child Transmission Guidelines, to include all HIV-exposed infants in December 2015. Methods: Between 2013 and 2016 we conducted an implementation science project to iteratively assess the implementation and effectiveness of the vertical transmission prevention of HIV in a chain of referral facilities in Cape Town (i.e. from primary to tertiary care). The e-register provided a single longitudinal record for each woman (linked to her infant after birth) and enabled assessment of HIV testing and treatment from first antenatal visit through delivery to infant HIV testing. Using a cohort of HIV-exposed live infants from this database, a protocol was designed (Section A: Protocol) to assess the implementation and outcome of effectively three different EID policy periods in the facility chain. i.e. an initial period of birth HIV-PCR at the clinician’s discretion if evidence of HIV infection; a period of targeted birth testing of high risk infants and lastly, of routine birth HIV-PCR for all HIV-exposed infants. A critical review of the literature appraised published assessments of birth HIV testing programmes in low- and middle-income countries (Section B: Literature Review) with the aim of assessing in utero transmission rates, follow-up testing and transmission rates and the resources required for implementation. Studies that modelled the impact of birth HIV testing were specifically reviewed. The manuscript (Section C: Manuscript) presented an analysis of the HIV-infected/exposed mother/infant dyads from the e-register of the Closing the Gaps study. Using this database adherence to guidelines in each period was assessed as well as the outcome of HIV-PCR at four delivery sites and the impact of the policies on return for follow-up EID. Results: South Africa is the first country in sub-Saharan Africa to implement birth HIV testing and most of the studies in support of this strategy were generated here. There was limited literature which highlighted the need for further investigation into the implementation and effectiveness of such programmes. No prospective data addressed targeted birth testing and those reporting on more routine birth HIV-PCR demonstrated success in timeous diagnosis and treatment although significant additional project resources were required. The retrospective laboratory data indicated that receipt of a birth HIV-PCR reduced the likelihood for follow-up at later testing time-points. This is important as the modelling studies suggested that the clinical and financial benefits of adding birth testing to existing algorithms would be lost if follow-up was poor. In the cohort of 2012 HIV-exposed infants in the study presented in the manuscript, the proportion who received birth testing increased with the progression of the EID policies but guideline implementation was poor, especially in primary care, with only 60% of infants being tested as recommended. The proportion of infants with positive HIV-PCR decreased as the pool of HIV-exposed infants undergoing testing expanded, being highest during the periods of targeted birth testing. In concurrence with the South African literature, receipt of a birth HIV-PCR decreased the likelihood of follow-up testing at 6-10 weeks. Among infants tested at 6-10 weeks old, the proportion who were positive for the first time at this time- point increased with the introduction of routine birth testing for all HIV-exposed infants, emphasizing the importance of the follow-up EID time-points. Conclusion: Virological testing at birth effectively increased the number of HIV-exposed infants who received an HIV test and was effective in identifying in utero infection in high risk infants (who could start treatment early with the attendant benefits.) The Western Cape EID policies were incompletely implemented in the study facilities over this time with many infants not being tested as indicated. Birth HIV-PCR decreased follow-up testing, an unintended consequence of serious concern. Adherence to the provincial and national guidelines needs to be re-enforced at delivery sites and at the primary care facilities where follow-up EID occurs.
  • Thumbnail Image
    Item
    Open Access
    A longitudinal analysis of the completeness of maternal HIV testing, including repeat testing, during pregnancy, and the predictors thereof, in Mitchell’s Plain, Cape Town
    (2019) De Beer, Shani; Davies, Mary-Ann; Kalk, Emma
    HIV testing during pregnancy is the gateway to the HIV-related services that are part of the prevention of mother-to-child transmission (PMTCT) cascade. The virtual elimination of vertical HIV transmission cannot be achieved without universal antenatal care (ANC) HIV testing. Furthermore, women are at an increased risk of HIV infection and subsequent mother-to-child transmission (MTCT) during pregnancy. Emphasis has thus been placed on repeat testing during pregnancy among women who have a HIV-negative result at their first ANC test. Very little has been published on the current uptake and adherence to antenatal and repeat HIV testing in sub-Saharan Africa (SSA) countries. In line with the World Health Organization Guidelines, the Western Cape Prevention of Mother-to-Child Transmission of HIV (PMTCT) Clinical Guidelines in 2014 recommended a repeat HIV test between 32 - 34 weeks gestation and again at delivery in addition to testing at “booking” (< 20 weeks gestation), meaning that there were three “testing windows” during which pregnant women not previously diagnosed as HIV-infected should undergo testing. Between 2013 and 2016 the Closing the Gaps study established an electronic PMTCT register (e-register) that consolidated routine care data from a primary healthcare facility and its secondary and tertiary referral sites in Cape Town, South Africa. This provided a single longitudinal record, from antenatal care to delivery, for each pregnant woman which enabled the longitudinal assessment of maternal HIV testing uptake and treatment. Utilizing these data, we conducted a retrospective sub-analysis investigating the implementation of PMTCT HIV testing guidelines (until delivery), in Cape Town, for the period 1 July 2014 - 31 December 2016. The main objectives of the study were to assess the coverage and timing of initial HIV testing during pregnancy, the completion of HIV testing at “booking” and within the recommended testing windows (including delivery), HIV prevalence and incidence at the recommended testing windows, and the predictors of missed testing opportunities. The research protocol (Part A) was designed to describe the proposed significance, objectives and methodology of the study. The literature review (Part B) critically evaluated available literature on: antenatal and repeat HIV testing proportions, HIV positivity, the feasibility and acceptability of repeat iv testing, and the predictors of testing completeness within different SSA countries, for the period 2010 - June 2018. Its aim was to inform this study. The need for post-Option B+ implementation, longitudinal studies that analyze antenatal and repeat HIV testing coverage and implementation within SSA was identified. In Part C I present the methods, results and interpretation thereof for the analysis of individual-level, longitudinal, maternal HIV-testing patient data from the Closing the Gaps study e-register as a manuscript to be submitted for publication. Among 8558 women who delivered at either the primary care facility or its referral sites, 7213 were not diagnosed HIV-positive prior to their first visit and thus eligible for testing in pregnancy. Among these women, 91% received ≥1 HIV test and 85% “booked” >5 days before delivery with 98% testing completeness at “booking”. Only 49% of women eligible for testing “booked” ≤22 weeks gestation. Among women that “booked” ≤22 weeks gestation who weren’t diagnosed HIV-positive before delivery and delivered >5 days after the start of the third trimester, 10% received tests in all three recommended windows. Thirty-one percent of women that had not been diagnosed HIV-positive before delivery had an uncertain (i.e. last tested ≥3 months before delivery) or unknown (i.e. never tested) HIV status after delivery. Out of the women that had a known HIV status at delivery, 21% were HIV-positive of whom 95% were known HIV-positive before current pregnancy and 4% were diagnosed at “booking”. Overall, HIV incidence in those with ≥2 HIV tests during pregnancy/at delivery was estimated to be 0.2% between “booking” and delivery. Women who enrolled after 2014 were less likely to miss ≥1 of the three recommended tests (aOR: 0.70; CI: 0.55 - 0.90) and not test at delivery (aOR: 0.63; CI: 0.55 - 0.71) compared to those who enrolled in 2014. Conclusion: In our study, HIV testing completion at “booking” was high, but women tended to “book” late during pregnancy resulting in late initial testing and missed opportunities for early HIV diagnosis. Implementation of repeat HIV testing is poor, particularly at delivery. HIV incidence between first negative ANC test and delivery is very low and therefore future studies to assess the most cost-effective number and timing of HIV tests, and feasibility of implementation, should be considered. Overall, maternal HIV testing within the PMTCT programme in Cape Town has matured post 2014 with improved implementation over time.
  • Thumbnail Image
    Item
    Open Access
    Adherence to antiretroviral therapy in young children in Cape Town, South Africa, measured by medication return and caregiver self-report : a prospective cohort study
    (2010) Davies, Mary-Ann; Boulle, Andrew; Eley, Brian
    Excellent adherence to antiretroviral therapy (ART) is necessary if HIV-infected children are to experience the dramatically improved outcomes that this treatment affords. However, there is very limited data on adherence to antiretroviral therapy in Africa, with few studies that examine the predictive value of low technology measures of adherence in terms of viral and immune outcomes. In addition there are no long terms studies of adherence in young children in Africa.
  • Thumbnail Image
    Item
    Open Access
    Adherence to antiretroviral therapy in young children in Cape Town, South Africa, measured by medication return and caregiver self-report: a prospective cohort study
    (BioMed Central Ltd, 2008) Davies, Mary-Ann; Boulle, Andrew; Fakir, Tanzeem; Nuttall, James; Eley, Brian
    BACKGROUND:Antiretroviral therapy (ART) dramatically improves outcomes for children in Africa; however excellent adherence is required for treatment success. This study describes the utility of different measures of adherence in detecting lapses in infants and young children in Cape Town, South Africa. METHODS: In a prospective cohort of 122 HIV-infected children commenced on ART, adherence was measured monthly during the first year of treatment by medication return (MR) for both syrups and tablets/capsules. A questionnaire was administered to caregivers after 3 months of treatment to assess experience with giving medication and self-reported adherence. Viral and immune response to treatment were assessed at the end of one year and associations with measured adherence determined. RESULTS: Medication was returned for 115/122 (94%) children with median age (IQR) of 37 (16 - 61) months. Ninety-one (79%) children achieved annual average MR adherence [greater than or equal to] 90%. This was an important covariate associated with viral suppression after adjustment for disease severity (OR = 5.5 [95%CI: 0.8-35.6], p = 0.075), however was not associated with immunological response to ART. By 3 months on ART, 13 (10%) children had deceased and 11 (10%) were lost to follow-up. Questionnaires were completed by 87/98 (90%) of caregivers of those who remained in care. Sensitivity of poor reported adherence (missing [greater than or equal to] 1 dose in the previous 3 days) for MR adherence <90% was only 31.8% (95% CI: 10.7% - 53.0%). Caregivers of 33/87 (38.4%) children reported difficulties with giving medication, most commonly poor palatability (21.8%). Independent socio-demographic predictors of MR adherence [greater than or equal to] 90% were secondary education of caregivers (OR = 4.49; 95%CI: 1.10 - 18.24) and access to water and electricity (OR = 2.65; 95%CI: 0.93 - 7.55). Taking ritonavir was negatively associated with MR adherence [greater than or equal to] 90% (OR = 0.37; 95%CI: 0.13 - 1.02). CONCLUSION: Excellent adherence to ART is possible in African infants and young children and the relatively simple low technology measure of adherence by MR strongly predicts viral response. Better socio-economic status and more palatable regimens are associated with better adherence.
  • Thumbnail Image
    Item
    Open Access
    Antiretroviral treatment for children
    (Health and Medical Publishing Group, 2006) Eley, Brian; Davies, Mary-Ann; Apolles, Patti; Cowburn, Carol; Buys, Heloise; Zampoli, Marco; Finlayson, Heather; King, Spasina; Nuttall, James
    Objective: To describe the response of children during their first year on highly active antiretroviral therapy (HAART). Design: Retrospective, descriptive. Setting. Tertiary, referral hospital. Subjects: All HIV-infected children commenced on HAART from 1 August 2002 until 31 December 2004. Outcome measures: Children were retrospectively restaged using the WHO 4-stage clinical classification and CDC immunological staging system. After commencing HAART, patients were assessed at monthly intervals for the first 6 months and thereafter mostly 3-monthly. Baseline and 6- monthly CD4 counts and viral loads were performed. Results. Of 409 children commenced on HAART, 50.6% were < 2 years old, 62.7% had severe clinical disease and 76.6% had severe immune suppression. After 1 year, 65.8% were alive and continued HAART at the hospital, 11.2% had been transferred to another antiretroviral site, 15.4% had died, 4.6% were lost to follow-up and treatment had been discontinued in 2.9%. Kaplan-Meier survival estimate for 407 children at 1 year was 84% (95% confidence interval (CI) 80 - 87%). On multivariate analysis, survival was adversely affected in children with WHO stage 4 v. stage 2 and 3 disease (adjusted hazard ratio (HR): 5.26 (95% CI 2.25 - 12.32), p = 0.000), age < 12 months (adjusted HR: 2.46 (95% CI 1.48 - 4.09), p = 0.001) and CD4 absolute count (per 100 cell increase) (adjusted HR: 0.93 (95% CI 0.88 - 0.98), p = 0.013). In a separate multivariate model including only children with an initial viral load (N = 367), viral load r 1 million copies/ml (adjusted HR: 1.84 (95% CI 1.03 - 3.29)) and taking a protease inhibitor (PI)-based regimen (adjusted HR: 2.25 (95% CI 1.10 - 4.61)) were additionally independently associated with poorer survival; however, young age was not a significant predictor of mortality, after adjusting for viral load (p = 0.119). After 1 year of HAART 184/264 (69.7%) of children had a viral load < 400 copies/ml. Comparative analysis showed significant improvements in growth, immunological status and virological control. Conclusion: HAART can improve the health of many HIVinfected children with advanced disease, including those aged less than 2 years in resource-limited settings.
  • No Thumbnail Available
    Item
    Open Access
    Burden and causes of ongoing paediatric infectious disease morbidity and mortality in the Western Cape Province of South Africa
    (2025) Kehoe, Kathleen; Davies, Mary-Ann; Eley Brian
    Under-five mortality has significantly decreased globally over the past 28 years, halving to 39 deaths per 1,000 live births in 2018, yet remains high, necessitating further progress to meet the Sustainable Development Goal of reducing it to below 25 deaths per 1,000 live births by 2030. Improvements in healthcare access, nutrition, vaccinations, and socioeconomic conditions have been key drivers of this observed reduction, but infectious diseases such as lower respiratory tract infections (LRTIs) including pneumonia, diarrhoea, and malaria continue to cause substantial childhood mortality. In South Africa, LRTIs, diarrhoea, meningitis, and tuberculous meningitis (TBM) remain leading causes of childhood morbidity and mortality. The Western Cape continues to bear a substantial burden from these infectious diseases, but the available data is outdated and lacks granularity. Therefore, the aim of this thesis was to explore the morbidity and mortality of LRTIs, diarrhoea, meningitis and TBM among children younger than five years in the public sector in the Western Cape. After a brief background chapter which lays out the key issues and overview of the South African healthcare system, Chapter 2 provides a comprehensive literature review discussing the morbidity and mortality of LRTIs, diarrhoea and meningitis (including TBM where appropriate) in children under five years globally and in South Africa with a focus on the Western Cape. Chapter 3 provides a detailed description of the data management required to develop the de-duplicated dataset that was used for each of the results chapters (Chapter 4-7). Chapter 4 explores causes of death using various death data sources, including routinely collected and detailed audits, and found that routine hospital information systems had accurate causes of death relying on International Classification of Diseases 10th Revision codes, particularly for LRTIs and diarrhoea. Chapters 5 and 6 explore the impact of COVID-19 public health and social measures (PHSM) on LRTI and diarrhoea admissions. COVID-19 surges and their associated measures, including PHSM, were linked to declining LRTI admissions and in- facility deaths, likely driven by a combination of reduced infectious disease transmission and reduced use of healthcare services. Lastly, Chapter 7 identified associations with repeat infectious disease admissions (LRTIs, diarrhoea and meningitis) among children who were first admitted for an infectious disease in the first six months of life. Male children with lower birthweight, whose first admission was due to LRTI or diarrhoea (versus meningitis), experienced a longer length of stay during their initial admission, and were living with HIV were more likely to be re-admitted for an infectious disease. Both individual- and population level interventions are needed to reduce the prevalence and impact of factors associated with infectious disease re-admissions and reduce infectious disease morbidity. This thesis concludes that infectious disease morbidity and mortality persist among children under five years in the Western Cape by presenting up-to-date and comprehensive data. It highlights the need to address existing gaps to improve data quality and comprehensiveness, as well as healthcare and health outcomes for these children.
  • No Thumbnail Available
    Item
    Open Access
    Characteristics and outcomes of adolescents living with HIV transitioning to adulthood in different health care models across Southern Africa
    (2025) Tsondai, Priscilla; Davies, Mary-Ann
    Adolescents and young people constitute an increasing proportion of the global population living with HIV, reflecting the shift in the burden of paediatric HIV towards older age groups. The recognition of an emerging cohort of young individuals living with HIV is indisputable. Nevertheless, there exists a notable dearth of understanding concerning the identification and timing of their transition to adulthood within the sub-Saharan Africa context. Although progress has been made in research on the transition of adolescents and youth living with HIV (AYLH) into adult care within developed nations, the field remains in its nascent stages in low- and middle-income countries (LMICs), where the majority of this population resides. Particularly concerning is the lack of information regarding the outcomes experienced by AYLH as they progress into adulthood in settings where the transition to adulthood necessitates the assumption of greater responsibility for their healthcare, without the physical transfer of care from paediatric or adolescent clinics to specialized adult HIV clinics, as may be the prevalent scenario within sub-Saharan Africa. To effectively evaluate and compare transition outcomes, it is imperative to establish optimal measurement approaches and implement comprehensive data collection practices across diverse settings. National programs and research cohorts in LMICs have made limited efforts to document the transition experience and associated factors. Moreover, the absence of a standardized approach to measure and compare transition-related processes and outcomes across different settings severely impedes the surveillance of outcomes specific to transitioning adolescents. Furthermore, most studies that have scrutinized transition-related outcomes have predominantly focused on post-transition outcomes, paying limited attention to the engagement in care among transitioning youth during the pre-transition period. Given that transitioning adolescents often undergo a protracted period of HIV care and ART prior to the formal transition, it is necessary to establish a comprehensive "adolescent transition cascade" that assesses outcomes both preceding and succeeding this critical phase. Therefore, there is a need for comprehensive longitudinal studies that track AYLH. Analyses that combine rigorous quantitative data collection, such as linked medical records with qualitative techniques such as in-depth interviews, will provide a more nuanced understanding of the complex transition process. Encouraging collaboration and networking among researchers across different countries could foster development of standardized measurement tools that accurately assess transition-related processes and outcomes. This will enhance the comparability of data across various studies and settings, facilitating a more comprehensive analysis of the factors that influence successful transitions. Aim and objectives: This thesis aims to provide a comprehensive description of the characteristics and outcomes of adolescents living with HIV across Southern Africa throughout the adolescent transition cascade. Additionally, it aims to identify the factors that contribute to these outcomes. Given the limited availability of standardized measurement tools that accurately assess transition-related processes and outcomes in resource-limited settings, this thesis also endeavours to develop and pilot a widely applicable and accessible data collection and management platform that will facilitate the measurement of transition processes and outcomes in adolescent HIV care.
  • No Thumbnail Available
    Item
    Open Access
    Characteristics and outcomes of children, adolescents and young adults on antiretroviral therapy in Southern Africa, incorporating additional outcome ascertainment through linkage and tracing studies
    (2024) Nyakato, Patience; Cornell, Morna; Davies, Mary-Ann
    Despite significant progress in pediatric HIV care and treatment, children, adolescents, and young adults living with HIV (CAYHIV) continue to face challenges in achieving optimal outcomes compared to adults due to challenges like virologic non-suppression (VNS) among those in care and loss to follow-up (LTFU). Adolescents, in particular, face psychosocial and structural barriers that hinder their adherence to antiretroviral therapy (ART), leading to VNS and associated negative consequences, such as increased morbidity, drug resistance, mortality and a higher risk of HIV transmission to sexual partners and, for those who are pregnant, to their infants. LTFU is concerning as CAYHIV who are not in care are also likely not to be on ART resulting in faster disease progression, VNS, and increased morbidity and mortality. LTFU also poses a challenge to accurately measuring programme outcomes as the true outcomes of those LTFU are unknown. Accurate estimation of mortality rates among CAYHIV requires ascertaining outcomes in those LTFU and is important for the effective management of HIV care programmes. The thesis therefore aimed to describe the characteristics and outcomes of CAYHIV in Southern Africa, including additional outcomes ascertained from linkage and tracing studies among those who had been reported as LTFU at the original sites of ART initiation. The thesis consists of five papers (three published, two submitted) reporting the results from observational HIV cohorts of the International epidemiology Database to Evaluate AIDS -Southern Africa (IeDEA-SA) in six Southern African countries of Lesotho, Malawi, Mozambique, South Africa, Zambia and Zimbabwe. Chapter 1 (Introduction) lays the foundation for key issues and concepts. This is followed by the literature review (Chapter 2) which gives a comprehensive discussion on virologic outcomes among adolescents, LTFU, ascertaining outcomes among CAYHIV reported as LTFU and correction of programme-level mortality estimates for LTFU among all CAYHIV using outcomes ascertained through tracing or linkage studies. Chapters 3 and 4 examine virologic outcomes and early LTFU among younger adolescents (10-14 years, Chapter 3) and older adolescents (15-19 years, Chapter 4), with a sub-analysis among those initiating treatment during pregnancy. Chapter 3 reports increasing 75th quantile viral load values with a three-fold increase at age 14 vs age 10 years, but no specific age at which this increase is more marked, and no differences observed by sex. Chapter 4 reports a relatively low rate of virologic non suppression (15%), but a high proportion of early LTFU following ART initiation (around 30%) irrespective of pregnancy status. Chapters 5 and 6 provide results on outcomes of CAYHIV previously reported as LTFU and either traced (Chapter 5) or linked to a health information exchange (Chapter 6). We defined tracing as the physical tracking of patients reported as LTFU using text messages, phone calls and home visits while linkage was defined as the process of linking patient unique identifiers to different healthcare data platforms like pharmacy records, laboratory records, hospital admissions to identify if they have had any interaction with the healthcare system within the province outside of their original facility or ART registration. The tracing approach reveals a high proportion of unreported mortality (9%) and a low proportion of self-transfers (21%) among CAYHIV while the linkage approach reveals a low proportion of mortality (3%), and a high proportion of self transfers (47%). Chapter 7 consolidates the results in Chapters 5 and 6 alongside routinely collected data to correct mortality estimates comparing three uncorrected and three corrected methods. There is a two-fold increase in estimated mortality after incorporating deaths among successfully traced CAYHIV due to the high mortality in traced patients. In contrast, incorporating linkage data has minimal impact on mortality estimates as there were few deaths but a high number of self transfers. Tracing and linkage-informed studies both show substantial variability in mortality among retained children and those LTFU across countries and sites, respectively. The thesis concludes that virologic response among CAYHIV, particularly adolescents, has greatly improved in more recent years with improved ART regimens and is expected to continue improving with the introduction of dolutegravir based therapies. However, this can easily be jeopardized by the persistent high proportion of CAYHIV reported as LTFU across the entire continuum of HIV care. Mortality estimates can also be substantially impacted if no additional outcome ascertainment is conducted among those reported as LTFU. Tracing and linkage informed studies are, therefore, important for accurate estimation of mortality and retention estimates.
  • No Thumbnail Available
    Item
    Open Access
    Characteristics of and factors associated with infectious disease hospital admissions in children by HIV-exposure status in an era of high coverage of prevention of vertical HIV transmission in the Western Cape Province of South Africa
    (2025) De Beer, Shani; Davies, Mary-Ann
    In South Africa, an estimated 22% of children under age 15 years are exposed to HIV and uninfected (HEU). Since the implementation of lifelong antiretroviral therapy (ART) for all pregnant women living with HIV in 2013 (Option B+) and all individuals living with HIV in 2016 (Universal ART), the proportion of children conceived while their mothers are on ART has increased. Previous research suggests that improved maternal health, due to ART, reduces the excess risk of poor health outcomes, including infectious disease hospital admissions, among children HEU vs. children unexposed to HIV and uninfected (HUU). Conversely, ART initiation before conception may result in higher risk of adverse birth outcomes (e.g. preterm delivery, which may drive increased risk of subsequent mortality and morbidity) compared to ART initiation after conception. Using routine healthcare data from the Western Cape (WC) Provincial Health Data Centre, this research aims to characterise the temporal changes in HIV-related characteristics of hospitalised children and their mothers and investigate whether child infectious morbidity and healthcare utilisation vary by HIV exposure and maternal ART use status, during the Option B+ and Universal ART eras. The sample size of children included in this research ranges across chapters from N=52,811 (children with infectious disease admissions 2008 – 2021) to N=549,782 (all children born 2008 – 2018). This thesis uses standardised definitions to systematically classify HIV exposure status of each child. It then describes changes in HIV-related characteristics of children with infectious disease hospitalisations across the WC from 2008 to 2021, showing a decrease in the proportion of admitted children living with HIV and an increase in the proportion of HIV- exposed children exposed to ART from conception, which highlights the success of HIV vertical transmission programmes. A comparison of infection-related hospitalisation rates demonstrates that children HEU vs. HUU experience higher rates of hospitalisation, irrespective of maternal ART history, during the first year of life. Finally, complete routine child healthcare clinic visit attendance among children HEU is seen to be associated with maternal ART start before pregnancy with no gaps in ART care, compared to maternal ART start later in pregnancy with or without gaps in ART. This thesis concludes that better coverage and implementation of existing interventions, as well as the introduction of more multi-faceted interventions, is required to ensure that both children HEU and HUU survive and thrive. Furthermore, this thesis demonstrates that electronic healthcare platforms play an important role in identifying children that require additional support and targeted interventions.
  • Thumbnail Image
    Item
    Open Access
    Clinical mentorship of nurse initiated antiretroviral therapy in Khayelitsha, South Africa: a quality of care assessment
    (Public Library of Science, 2014) Green, Ann; de Azevedo, Virginia; Patten, Gabriela; Davies, Mary-Ann; Ibeto, Mary; Cox, Vivian
    Introduction To combat the AIDS epidemic and increase HIV treatment access, the South African government implemented a nurse-based, doctor-supported model of care that decentralizes administration of antiretroviral treatment (ART) for HIV positive patients through nurse initiated and managed ART. Médecins Sans Frontières (MSF) implemented a mentorship programme to ensure successful task-shifting, subsequently assessing the quality of clinical care provided by nurses. METHODS: A before-after cross-sectional study was conducted on nurses completing the mentorship programme in Khayelitsha, South Africa, from February 2011-September 2012. Routine clinical data from 229 patient folders and 21 self-assessment questionnaires was collected to determine the number of patients initiated on ART by nurses; quality of ART management before-after mentorship; patient characteristics for doctor and nurse ART initiations; and nurse self-assessments after mentorship. RESULTS: Twenty one nurses were authorized by one nurse mentor with one part-time medical officer's support, resulting in nurses initiating 77% of ART eligible patients. Improvements in ART management were found for drawing required bloods (91% vs 99%, p = 0.03), assessing adherence (50% vs 78%, p<0.001) and WHO staging (63% vs 91%, p<0.001). Nurse ART initiation indicators were successfully completed at 95-100% for 11 of 16 indicators: clinical presentation; patient weight; baseline blood work (CD4, creatinine, haemoglobin); STI screening; WHO stage, correlating medical history; medications prescribed appropriately; ART start date; and documented return date. Doctors initiated more patients with TB/HIV co-infection and WHO Stage 3 and 4 disease than nurses. Nurse confidence improved for managing HIV-infected children and pregnant women, blood result interpretation and long-term side effects. CONCLUSIONS: Implementation of a clinical mentorship programme in Khayelitsha led to nurse initiation of a majority of eligible patients, enabling medical officers to manage complex cases. As mentorship can increase clinical confidence and enhance professional development, it should be considered essential for universal ART access in resource limited settings.
  • Thumbnail Image
    Item
    Open Access
    Clinical mentorship of nurse-initiated Antiretroviral Therapy (ART) in Khayelitsha, South Africa: a quality of care assessment
    (2013) Green, Ann; Cox, Vivian; Davies, Mary-Ann
    South Africa’s National Strategic Plan calls for task shifting of HIV care from doctors to nurses by 2016 to increase access to ART. There is little research demonstrating sustainable success of competent HIV management by nurses after training and mentorship. In February 2011, Medecins Sans Frontieres (MSF) partnered with the City of Cape Town Department of Health to implement a NIMART mentorship programme and assess quality of clinical care provided by nurse graduates. A before-after cross-sectional study was conducted on nurses completing mentorship from February 2011-September 2012. Routine clinical data from 229 patient folders and 21 selfassessment questionnaires was collected to determine the number of patients initiated on ART by individual nurses; quality of ART-management before and after mentorship; patient characteristics for doctor and nurse ART-initiations; and nurse self-assessments. 21 nurses were authorized by one nurse mentor with one part-time medical officer’s support, resulting in nurses initiating 77% of eligible patients. Improvements in ART management were found for drawing required bloods (91% vs 99%, p=0.03), assessing adherence (50% vs 78%, p<;0.001) and WHO staging (63% vs 91%, p<;0.001). Nurse ART initiation indicators were successfully completed 95-100% of the time for 10 out of 16 indicators: clinical presentation, past medical history, prior ART history, WHO stage, STI screening, patient weight, baseline blood work (CD4 count, creatinine, and haemoglobin), treatment plan, and medications prescribed. Doctors initiated more patients with TB/HIV co-infection and Stage 3 or 4 disease than nurses; 24% of patients initiated by nurses were seen or discussed with a doctor in the month prior to initiation. Nurse confidence improved for the management of HIV-infected children and pregnant women, blood result interpretation, and long-term side effects following completion of the mentorship programme.
  • Thumbnail Image
    Item
    Open Access
    Cohort Profile: The Western Cape Pregnancy Exposure Registry (WCPER)
    (2022-06) Kalk, Emma; Heekes, Alexa; Slogrove, Amy; Phelanyane, Florence; Davies, Mary-Ann; Myer, Landon; Euvrard, Jonathan; Kroon, Max; Petro, Greg; Fieggen, Karen; Stewart, Chantal; Rhoda, Natasha; Gebhardt, Stefan; Osman, Ayesha; Anderson, Kim; Boulle, Andrew; Mehta, Ushma
    Purpose: The Western Cape Pregnancy Exposure Registry (PER) was established at two public sector healthcare sentinel sites in the Western Cape province, South Africa, to provide ongoing surveillance of drug exposures in pregnancy and associations with pregnancy outcomes. Participants: Established in 2016, all women attending their first antenatal visit at primary care obstetric facilities were enrolled and followed to pregnancy outcome regardless of the site (ie, primary, secondary, tertiary facility). Routine operational obstetric and medical data are digitised from the clinical stationery at the healthcare facilities. Data collection has been integrated into existing services and information platforms and supports routine operations. The PER is situated within the Provincial Health Data Centre, an information exchange that harmonises and consolidates all health-related electronic data in the province. Data are contributed via linkage across a unique identifier. This relationship limits the missing data in the PER, allows validation and avoids misclassification in the population-level data set. Findings to date: Approximately 5000 and 3500 pregnant women enter the data set annually at the urban and rural sites, respectively. As of August 2021, >30 000 pregnancies have been recorded and outcomes have been determined for 93%. Analysis of key obstetric and neonatal health indicators derived from the PER are consistent with the aggregate data in the District Health Information System. Future plans: This represents significant infrastructure, able to address clinical and epidemiological concerns in a low/middle-income setting.
  • Thumbnail Image
    Item
    Open Access
    A comparison of death recording by health centres and civil registration in South Africans receiving antiretroviral treatment
    (Journal of the International AIDS Society, 2015-12-16) Johnson, Leigh F; Dorrington, Rob E; Laubscher, Ria; Hoffmann, Christopher J; Wood, Robin; Fox, Matthew P; Cornell, Morna; Schomaker, Michael; Prozesky, Hans; Tanser, Frank; Davies, Mary-Ann; Boulle, Andrew
    Introduction: There is uncertainty regarding the completeness of death recording by civil registration and by health centres in South Africa. This paper aims to compare death recording by the two systems, in cohorts of South African patients receiving antiretroviral treatment (ART). Methods: Completeness of death recording was estimated using a capture-recapture approach. Six ART programmes linked their patient record systems to the vital registration system using civil ID numbers, and provided data comparing the outcomes recorded in patient files and in the vital registration. Patients were excluded if they had missing/invalid IDs or had transferred to other ART programmes. Results: After exclusions, 91 548 patient records were included. Of deaths recorded in patients files after 2003, 94.0% (95% CI: 93.3-94.6%) were recorded by civil registration, with completeness being significantly higher in urban areas, older adults and females. Of deaths recorded by civil registration after 2003, only 35.0% (95% CI: 34.2-35.8%) were recorded in patient files, with this proportion dropping from 60% in 2004-2005 to 30% in 2010 and subsequent years. Recording of deaths in patient files was significantly higher in children and in locations within 50km of the health centre. When the information from the two systems was combined, an estimated 96.2% of all deaths were recorded (93.5% in children and 96.2% in adults). Conclusions: South Africa’s civil registration system has achieved a high level of completeness in the recording of mortality. However, the fraction of deaths recorded by health centres is low and information from patient records is insufficient by itself to evaluate levels and predictors of ART patient mortality. Previously-documented improvements in ART mortality over time may be biased if based only on data from patient records.
  • Thumbnail Image
    Item
    Open Access
    A description of HIV-exposed uninfected infants in the IeDEA Southern Africa Cohort and an examination of growth outcomes
    (2014) Morden, Erna; Davies, Mary-Ann
    Since the successful use of antiretroviral therapy for the prevention of mother-to-child transmission of human immunodeficiency virus (HIV), there has been a steady increase in the number of infants born to HIV-infected mothers who remain uninfected. The characteristics of these HIV-exposed uninfected infants are not well known, including growth and other health outcomes. The International Epidemiologic Databases to Evaluate AIDS Southern Africa (IeDEA-SA) research strategy 2011-2016 includes specific studies in pregnant women, infants, children and adolescents. This study addresses one of the IeDEA-SA objectives, namely to establish and describe a sub-cohort of HIV-infected pregnant women and their exposed infants. Part A, the protocol, includes background information on sites contributing to this cohort of HIV-exposed uninfected (HEU) infants. It also details the aims, objectives and methodology of this study. Part B, the literature review, discusses what is known about HIV-exposed uninfected infants to date. It includes maternal disease factors, the use of antiretroviral therapy and the association between feeding modality and growth, focussing on studies conducted on the African continent. Part C, the manuscript, details the methodology, results and their interpretation of longitudinal analysis of growth among HEU infants in the IeDEA-SA collaboration. This cohort of HEU infants included 2621 infants from two South African sites. The median birth WAZ was -0.65 (IQR -1.46; 0.0), 51% were male and there was a median of 2 visits per infant. The feeding modalities practised were as follows: 0.5% exclusive breastfeeding, 7.9% unknown breastfeeding exclusivity, 78.6% mixed breastfeeding and 10.6% formula feeding. Mothers with a CD4 <200 cells/μl delivered infants with a lower birth WAZ (adjusted ß -0.253 [95% CI -0.043; -0.072], p = 0.006) compared to mothers with a CD4 ≥500 cells/≧l. Similarly, iv mothers who did not receive antiretroviral (ARVs) drugs delivered infants with a lower birth WAZ (adjusted ß -0.49 [95% CI -0.78; -0.20], p = 0.001) compared to mothers who received antenatal ARVs. Antenatal maternal ARVs and CD4 cell count did not have an effect on postnatal growth. Mixed effects models using maximum likelihood estimation for the longitudinal analysis of growth showed that exposure to breast milk positively influenced growth, albeit the effect was small. Infants with a birth weight <2 500g (ß 0.069 [95% CI 0.061; 0.078], p <0.0001) experienced faster growth within the first 28 weeks of life compared to infants with a birth weight ≥2 500g. In this cohort of South African HEU infants, less severe maternal disease and the use of ARVs positively impacted birth weight. Mixed feeding was common, and any breastfeeding may have a positive effect on longitudinal growth.
  • Thumbnail Image
    Item
    Restricted
    Growth and mortality outcomes for different antiretroviral therapy initiation criteria in children ages 1–5 Years: a causal modeling analysis
    (Lippincott, Williams & Wilkins, 2016) Schomaker, Michael; Davies, Mary-Ann; Malateste, Karen; Renner, Lorna; Sawry, Shobna; N’Gbeche, Sylvie; Technau, Karl-Günter; Eboua, François; Tanser, Frank; Sygnaté-Sy, Haby; Phiri, Sam; Madeleine, Amorissani-Folquet; Cox, Vivian; Koueta, Fla; Chimbete, Cleophas; Lawson-Evi, Annette; Giddy, Janet; Amani-Bosse, Clarisse; Wood, Robin; Egger, Matthias; Leroy, Valeriane
    Background: There is limited evidence regarding the optimal timing of initiating antiretroviral therapy (ART) in children. We conducted a causal modeling analysis in children ages 1–5 years from the International Epidemiologic Databases to Evaluate AIDS West/Southern-Africa collaboration to determine growth and mortality differences related to different CD4-based treatment initiation criteria, age groups, and regions. Methods: ART-naïve children of ages 12–59 months at enrollment with at least one visit before ART initiation and one follow-up visit were included. We estimated 3-year growth and cumulative mortality from the start of follow-up for different CD4 criteria using g-computation. Results: About one quarter of the 5,826 included children was from West Africa (24.6%).The median (first; third quartile) CD4% at the first visit was 16% (11%; 23%), the median weight-for-age z-scores and height-for-age z-scores were −1.5 (−2.7; −0.6) and −2.5 (−3.5; −1.5), respectively. Estimated cumulative mortality was higher overall, and growth was slower, when initiating ART at lower CD4 thresholds. After 3 years of follow-up, the estimated mortality difference between starting ART routinely irrespective of CD4 count and starting ART if either CD4 count <750 cells/mm3 or CD4% <25% was 0.2% (95% CI = −0.2%; 0.3%), and the difference in the mean height-for-age z-scores of those who survived was −0.02 (95% CI = −0.04; 0.01). Younger children ages 1–2 and children in West Africa had worse outcomes. Conclusions: Our results demonstrate that earlier treatment initiation yields overall better growth and mortality outcomes, although we could not show any differences in outcomes between immediate ART and delaying until CD4 count/% falls below 750/25%.
  • Thumbnail Image
    Item
    Open Access
    Growth of HIV-exposed uninfected infants in the first 6 months of life in South Africa: The IeDEA-SA collaboration
    (Public Library of Science, 2016) Morden, Erna; Technau, Karl-Günter; Giddy, Janet; Maxwell, Nicola; Keiser, Olivia; Davies, Mary-Ann
    BACKGROUND: HIV-exposed uninfected (HEU) infants are a growing population in sub-Saharan Africa especially with the increasing coverage of more effective prevention of mother-to-child transmission (PMTCT) antiretroviral therapy regimens. This study describes the characteristics of South African HEU infants, investigates factors impacting birth weight and assesses their growth within the first 28 weeks of life. METHODS: This is a retrospective cohort based on routine clinical data from two South African PMTCT programmes. Data were collected between 2007 and 2013. Linear regression assessed factors affecting birth weight-for-age z-scores (WAZ) while growth (longitudinal WAZ) was assessed using mixed effects models. RESULTS: We assessed the growth of 2621 HEU infants (median birth WAZ was -0.65 (IQR -1.46; 0.0) and 51% were male). The feeding modalities practised were as follows: 0.5% exclusive breastfeeding, 7.9% breastfeeding with unknown exclusivity, 0.08% mixed breastfeeding and 89.2% formula feeding. Mothers with CD4 <200 cells/μl delivered infants with a lower birth WAZ (adjusted ß -0.253 [95% CI -0.043; -0.072], p = 0.006) compared to mothers with aCD4 ≥500 cells/μl. Similarly, mothers who did not receive antiretroviral drugs delivered infants with a lower birth WAZ (adjusted ß -0.39 [95% CI -0.67; -0.11], p = 0.007) compared to mothers who received antenatal antiretrovirals. Infants with a birth weight <2 500g (ß 0.070 [95% CI 0.061; 0.078], p <0.0001) experienced faster growth within the first 28 weeks of life compared to infants with a birth weight ≥2 500g. Infants with any breastfeeding exposure experienced slower longitudinal growth compared to formula fed infants (adjusted ß -0.012 [95% CI 0.021; -0.003], p = 0.011). CONCLUSION: Less severe maternal disease and the use of antiretrovirals positively impacts birth weight in this cohort of South African HEU infants. Formula feeding was common with breastfed infants experiencing marginally slower longitudinal growth.
  • Thumbnail Image
    Item
    Open Access
    How accurately do routinely reported HIV viral load suppression proportions reflect progress towards the 90-90-90 target in the population on ART in Khayelitsha, South Africa?
    (2018) Euvrard, Jonathan George; Davies, Mary-Ann; Schulz, Tanja
    Background: In 2016, Khayelitsha reported almost 89% viral load (VL) suppression but less than 56% completion on routine quarterly reports, casting doubt on the validity of reported suppression. Objectives: To assess the validity of reported VL suppression as a measure of progress towards the 90-90-90 target and identify barriers to routine VL completion. Methods: A retrospective cohort study including all patients on antiretroviral treatment (ART) in Khayelitsha with a routine VL expected between 1 July 2015 and 30 June 2016 was conducted. ART programme and laboratory data were obtained and a sample of 1 035 patient folders were reviewed. Suppression was calculated using laboratory data and compared to reported suppression. A VL cascade from “expected” to “done”, “filed”, “noted” and “captured” was constructed to reflect the steps a VL must complete to be included in reported suppression and successful progression to each step was estimated. Logistic regression models were used to calculate adjusted odds ratios (aOR) with 95% confidence intervals (95%CI) for completion among different patient groups. Results: Using laboratory data, VL suppression was estimated to be 82%, 87%, 89% and 91% at the 50, 200, 400 and 1 000 copies/mL thresholds respectively, but reported suppression would have been 80%, 86%, 88% and 89% at those thresholds. Of 22 991 patients with a routine VL due in the study period, 84% were done, 79% filed, 76% noted, and 55% captured. Routine VL were more likely to be done among children< 15 years old (aOR 1.89, 95%CI 1.45–2.48) and pregnant women (aOR 1.90, 95%CI 1.28–2.81) compared to adult men, adjusted for facility. Conclusion: Despite low reported completion, actual completion was high and reported suppression was similar to suppression calculated using laboratory data, thus providing an accurate measure of progress towards the 90-90-90 target. More work is needed to reach the 16% of patients missed by routine testing. Most of the VL done were available to the clinician at the next assessment, and further research is needed to assess how effectively these VL results are used in clinical decision-making.
  • Thumbnail Image
    Item
    Open Access
    Initial experience of a public sector antiretroviral treatment programme for HIV-infected children and their infected parents
    (2004) Eley, Brian; Nuttall, James; Davies, Mary-Ann; Smith, Lara; Cowburn, Carol; Buys, Heloise; Hussey, Gregory
    Objective. To describe the initial experience of treating HIVinfected children and their infected parents with antiretroviral therapy. Design. Prospective, cohort study. Setting. Tertiary, referral hospital. Patients. HIV-infected children and their parents. Methods. This report focuses on the early response of children to highly active antiretroviral therapy (HAART). Children were followed up at 4-weekly intervals. Monitoring included initial and yearly viral load measurements, baseline and 6- monthly CD4 counts and 4-weekly adherence checks. Results. Between August 2002 and June 2003, 80 children were enrolled in the programme, representing a follow-up period of 23.9 patient-years. Seventy-five children had severe clinical disease, severe immune suppression, or a combination of the two. The response of children who had received HAART for ≥ 6 months (N = 17) was assessed. There was no change in mass z-score (p = 0.11) or length z-score (p = 0.37), but a significant increase in CD4 percentage (p < 0.0001) during the first 6 months of therapy. Six-month viral loads were available for 12 children. There was a significant drop in viral load (p = 0.001) and 9 achieved undetectable levels by 6 months. Most children achieved ≥ 85% adherence. By June 2002, 67 children (84%) were relatively well, 1 had B-cell lymphoma, 7 (8.8%) had died, 4 (5%) were lost to follow-up and 1 was withdrawn from the programme. Of 57 children who completed 3 months of HAART, 12 were admitted a total of 17 times for infectious complications. There were no severe drug reactions. Three of 7 mothers on HAART received treatment through the programme. Conclusion. These initial results suggest that many HIVinfected children in the public sector will benefit from antiretroviral therapy. However, both ambulatory and inpatient facilities are required to manage children on HAART comprehensively
  • Thumbnail Image
    Item
    Open Access
    "Injury-prone areas" Repeat Presentations for Injuries in Childhood
    (2019) Peters, Shrikant Maurice; Davies, Mary-Ann
    Background: Injury remains a leading cause of childhood morbidity and mortality in the developing world. Probability of injury occurrence is influenced by agent, host and environmental factors. Studies of repeat injuries in childhood thus provide insight into factors in the epidemiological triad predisposing children to injury. Objectives: The study objectives were to determine the proportion of children and the factors associated with repeat presentations to Red Cross War Memorial Children’s Hospital Trauma Unit (RCWMCH TU) for all non-transport related injuries in childhood. Methods: This was a retrospective cohort study using data from RCWMCH TU. We included children aged 0-10 years with first presentation from January 1997 to June 2013, and followed up until the earliest of age 13 years or June 2016. We assessed individual and population-level factors associated with repeat injury using multilevel Poisson regression. Child Dependency Ratios were derived from the 2011 national census. Results: Between 1997 and 2013, 72 490 children under 10 years of age (59% male) presented to RCWMCH TU for the first time with injuries. After the initial injury, 9 417 (13%) presented with a repeat injury by 2016 and before age 13 years. After adjusting for health Sub-District, distance from RCWMCH TU and age at first presentation, factors associated with reduced repeat presentation were: injury identified as due to abuse (adjusted incidence rate ratio [aIRR] 0.6; 95% confidence interval [CI]: 0.4 - 0.7), fluid burn (aIRR 0.6; 95% CI: 0.6 - 0.7), foreign body ingestion (aIRR 0.7; 95% CI: 0.7 - 0.9), moderate and severe (vs minor) initial injury (aIRR 0.9; 95% CI: 0.8 - 0.9) and (aIRR 0.7; 95% CI: 0.6 - 0.8 respectively), whilst boys were more likely to have repeat injury presentations (aIRR 1.4; 95% CI: 1.4 - 1.5). Conclusion: Repeat presentations constituted a substantial proportion of disease burden. Factors associated with repeat presentations were identified, strengthening the argument that injuries arise due to sustained exposure to host, agent and environmental risk factors. While it is reassuring that children with initial injuries due to abuse and severe initial injuries are less likely to present again, injury prevention education should not neglect patients with minor and unintentional injuries. The findings of this study suggest that post-injury health promotion activities should not exclude patients who present with minor injuries, and that targeted education and further research is indicated for specific injuries, including those due to non-height falls and injuries sustained whilst playing sport.
  • Thumbnail Image
    Item
    Open Access
    Knowledge, attitudes and perceptions of antibiotic use and antibiotic resistance among private sector patients and prescribers in South Africa
    (2017) Farley, Elise; Boyles, Tom H; Stewart, Annemie; Davies, Mary-Ann
    Antibiotic resistance (ABR), alternately referred to as antimicrobial resistance, has been labelled as the next big global health crisis. If current levels of ABR continue along the same trajectories, by 2050 ABR will cost the lives of 10 million people a year, ABR cannot be stopped but it can be slowed down. ABR occurs because the bacteria evolve to protect themselves from antibiotics. One of the main causes of ABR is the misuse and over prescription of antibiotics. The primary objective of the study is to ascertain the level of knowledge, attitudes and perceptions of appropriate antibiotic use and ABR, among prescribers and patients in private health care in South Africa. The secondary objective of the study is to explore associations between knowledge, attitudes and perceptions of prescribers and patients regarding antibiotic use and resistance. This project consists of three main sections, a proposal, literature review and a journal ready article. All sections focus on ABR. The proposal lays a foundation for the need for the research, and explains how the research will be conducted. The literature review explores the existing evidence on the topic, and the final section is a secondary analysis of cross sectional study data, in which private practice patients and prescribers in South Africa completed a once-off anonymous survey. Data was analysed using Stata,T-tests, chi-squared tests and logistic regression models were used to assess associations between knowledge, attitudes and perceptions of both patients and prescribers. We found that mean knowledge scores among patients (n=403, mean 9 out of 14, standard deviation [SD] 3) and providers (n=175, median 5 maximum 7, IQR 4, 6), were suboptimal and that poor knowledge was associated with perceptions and behaviours as well as prescribing practices that could lead to ABR. Associations between knowledge, attitudes and perceptions of patients and prescribers were explored in multivariate logistic regression models. After adjusting for education and sex, a 1-unit increase in patient knowledge score was associated with the belief that antibiotics will work less well in future if we over-use them now (aOR 1.3; 95% CI: 1.18, 1.43; pvalue <0.001). Prescribers with higher knowledge scores were less likely to report that they prescribe antibiotics when not necessary as antibiotics cannot harm the patient (aOR 0.55; 95% CI: 0.33, 0.91; pvalue 0.02). We also identified a large proportion (58%) of patients who were interested in alternatives to antibiotics and a large proportion (91%) of prescribers wanting educational material to facilitate conversations about resistance with patients. Our study demonstrates gaps in patient and prescriber knowledge that are associated with potentially harmful perceptions and destructive behaviours regarding antibiotic use. These associations, together with our finding that patients and prescribers would like more education on ABR, suggest that educational tools and patient-provider communication tools could promote rational antibiotic use.