Medicine & Health

Publication Search Results

Now showing 1 - 8 of 8
  • (2023) Keller, Elena
    Infertility affects 1 in 6 couples and >180 million people worldwide. It represents an increasingly important public health problem, amplified by the continuing global trend to later childbearing. Fertility treatment including in vitro fertilization (IVF) is not suited to traditional health technology assessment (HTA) methods, because its value is derived by its ability to create life, rather than extend, improve, or save existing lives. Consequently, there is a lack of guidance, and satisfactory HTA methods to determine whether fertility treatment provides good value for money. Moreover, the ever-increasing demand for elective egg freezing (EEF) to preserve female fertility poses additional challenges for economic assessments. This thesis describes 5 studies that move the research agenda forward for guiding the economic evaluation of fertility treatment. Study 1, a systematic review, identified and quantified 5 methodological categories for value-of-statistical-life elicitation. Based on these categories, Study 2 investigated methods for eliciting the value of a statistical baby (VSB) and concluded that discrete choice experiments (DCEs) are the most appropriate method in a fertility treatment context. Study 3 applied DCE outputs to derive a VSB estimate, which was used to assess value for money of publicly funded IVF in a cost-benefit analysis, finding that at least 5 IVF cycles likely provide good value for women <42 years. Study 4 elicited patient preferences for fertility treatment based on a DCE and Study 5 performed an incentive-compatible lab experiment to assess the impact of patient and treatment characteristics on the demand for IVF and EEF. Both experiments indicate that the demand for fertility treatment is price-inelastic and unresponsive to income level, which might explain why women continue fertility treatment once they have commenced despite their financial capacity. This research makes several methodological contributions and provides an evidence base to assess the public investment in fertility treatment. Overall, patients and society were found to value fertility treatment highly. New knowledge generated includes: (1) identifying the number of cost-beneficial IVF cycles by female age; (2) quantifying price and income elasticities for IVF and EEF; (3) bridging the gap between the proliferation of DCEs and policy by applying DCE outputs to HTA; and (4) demonstrating that government funding decisions can be explored in a lab experiment.

  • (2023) Eastick, Jessica
    Aims: The first study investigated the agreement between embryologists when assessing cytoplasmic strings (CS) in human blastocysts. The second study investigated the idea that embryos originating from fresh ejaculate sperm is comparable to those originating from frozen-thawed ejaculate sperm. The third study investigated and expanded on the association between the presence of CS, blastocyst quality and live birth rate. The fourth study used timelapse technology to investigate whether the presence of CS in human blastocysts was associated with a clinical pregnancy. Methods: In the first study, CS activity in one hundred blastocysts was assessed to measure inter- & intra-observer agreement. In the second study, timelapse was used to observe embryo development in those originating from fresh ejaculate sperm compared to those originating from frozen-thawed ejaculate sperm. In the third study, timelapse was used to investigate the association between the presence of CS and their characteristics, with blastocyst quality, development and live birth in one thousand day 5 human blastocysts. In the fourth study, timelapse was used to assess CS presence in blastocysts to investigate the link between those blastocysts that formed a fetal heart after transfer. Results: From the first study, a moderate level of inter and intra observer agreement was observed when five embryologist assessed day 5/6 blastocysts for the presence of CS and their vesicles. From the second study, it was found that no differences were detected in key developmental events between embryos originating from fresh ejaculate sperm compared to their frozen-thawed counterparts. The third study confirmed that CS presence in human blastocysts is associated with blastocyst quality. The fourth study, the presence of CS in human blastocysts was found to be associated with a fetal heart. Conclusions: The first study showed a moderate level of inter- and intra-observer agreement when the presence of CS and their vesicles was assessed. The second study found that there are no differences in the morphokinetic parameters of early embryo development when either fresh or frozen ejaculate sperm was used for ICSI insemination. The third study confirmed that CS presence in human blastocysts is associated with blastocyst quality. The fourth study found that the presence of cytoplasmic strings in human blastocysts is associated with the probability of clinical pregnancy with fetal heart.

  • (2023) Odutola, Michael
    Follicular lymphoma (FL) accounts for one-third of incident non-Hodgkin lymphomas in Western countries, but its etiology is largely unexplained. I performed systematic reviews and meta-analyses, and used a population-based family case-control study to investigate the relationship between lifestyle, environmental and occupational risk factors of FL. Meta-estimates from my random-effect models showed a non-significant association with smoking, heterogeneous results for alcohol, modest increased risk with obesity, and positive associations with exposure to polychlorinated biphenyls (PCBs), chlorinated solvents and dichlorodiphenyldichloroethylene (DDT, a pesticide). The population-based case-control study included 770 FL cases and 490 family controls (siblings, partners). Participants completed a lifetime residential and work calendar and health, lifestyle, and diet questionnaires. I used unconditional logistic regression to examine associations with FL risk, including group-based trajectory modeling to examine associations with body shape and outdoor hours over the life course. I identified deaths using record linkage and applied Cox proportional regression to estimate hazard ratios for all-cause and FL-specific mortality. I observed a positive association between smoking and FL risk and mortality. Associations with recent alcohol intake and FL were null. Being obese 5 years prior to enrolment and higher body mass index 5 years prior to enrolment was associated with a modest increased FL risk, but there was no association with body shape trajectory. Body size was not associated with mortality. I observed an elevated FL risk with consumption of oily fish, but no association with mortality. I found no significant association between occupational exposure to pesticides, or extremely-low frequency magnetic fields, and FL risk. For sun exposure, I observed an inverse association with high cumulative outdoor hours and high outdoor hour maintainers over the life course, and FL risk. Policies on tobacco control, maintaining body weight within normal range, and safe use of solvents and pesticides are crucial in reducing the burden from FL. Sun exposure is not recommended as a cancer control policy, but the association may inform research on targeted therapies for this malignancy. My key findings have advanced our understanding of FL etiology and help guide risk reduction strategies and future research.

  • (2023) Liu, Leibo
    The widespread adoption of Electronic Medical Records (EMRs) in hospitals continues to increase the amount of patient data that are digitally stored. Although the primary use of the EMR is to support patient care by making all relevant information accessible, governments and health organisations are looking for ways to unleash the potential of these data for secondary purposes, including clinical research, disease surveillance and automation of healthcare processes and workflows. EMRs include large quantities of free text documents that contain valuable information. The greatest challenges in using the free text data in EMRs include the removal of personally identifiable information and the extraction of relevant information for specific tasks such as clinical coding. Machine learning-based automated approaches can potentially address these challenges. This thesis aims to explore and improve the performance of machine learning models for automated de-identification and clinical coding of free text data in EMRs, as captured in hospital discharge summaries, and facilitate the applications of these approaches in real-world use cases. It does so by 1) implementing an end-to-end de-identification framework using an ensemble of deep learning models; 2) developing a web-based system for de-identification of free text (DEFT) with an interactive learning loop; 3) proposing and implementing a hierarchical label-wise attention transformer model (HiLAT) for explainable International Classification of Diseases (ICD) coding; and 4) investigating the use of extreme multi-label long text transformer-based models for automated ICD coding. The key findings include: 1) An end-to-end framework using an ensemble of deep learning base-models achieved excellent performance on the de-identification task. 2) A new web-based de-identification software system (DEFT) can be readily and easily adopted by data custodians and researchers to perform de-identification of free text in EMRs. 3) A novel domain-specific transformer-based model (HiLAT) achieved state-of-the-art (SOTA) results for predicting ICD codes on a Medical Information Mart for Intensive Care (MIMIC-III) dataset comprising the discharge summaries (n=12,808) that are coded with at least one of the most 50 frequent diagnosis and procedure codes. In addition, the label-wise attention scores for the tokens in the discharge summary presented a potential explainability tool for checking the face validity of ICD code predictions. 4) An optimised transformer-based model, PLM-ICD, achieved the latest SOTA results for ICD coding on all the discharge summaries of the MIMIC-III dataset (n=59,652). The segmentation method, which split the long text consecutively into multiple small chunks, addressed the problem of applying transformer-based models to long text datasets. However, using transformer-based models on extremely large label sets needs further research. These findings demonstrate that the de-identification and clinical coding tasks can benefit from the application of machine learning approaches, present practical tools for implementing these approaches, and highlight priorities for further research.

  • (2023) Talbot, Benjamin
    Chronic kidney disease (CKD) affects more than 1/10 people worldwide with a disproportionately high burden in disadvantaged communities. As CKD severity increases, the associated morbidity, mortality and treatment costs also increase. In the case of kidney failure, the most severe form of CKD, the costs of treatment, including life prolonging treatment with dialysis or kidney transplant, are often unaffordable in under-resourced healthcare settings. Data has been central to improving the outcomes of patients with CKD, but there continue to be important data gaps, especially in low- and lower-middle-income countries (LLMICs). In order to more comprehensively understand the burden of kidney disease, it is necessary to overcome the many challenges to data collection which exist globally. To explore how this could be achieved, this thesis examines how four different data sources can contribute to addressing gaps in understanding CKD. Firstly, the role of kidney replacement therapy (KRT) registries in LLMICs were assessed through a review of the literature and explored further by implementing a dialysis registry in Fiji. Secondly, extending data collection of a randomised controlled trial to examine how differing practice patterns across regions might impact outcomes was assessed through analysis of the extended follow-up of the Study of Heart and Renal Protection (SHARP). Thirdly, the role of administrative data was explored through a literature review and through two novel data linkage analyses. Lastly, semi-structured interviews were conducted with patients and clinicians to understand their perspectives on remote patient monitoring (RPM), a novel approach to patient data collection for dialysis treatment. The analyses examining the role of focused KRT registries in LLMICs and the utility of long-term follow-up of clinical trials to compare outcomes between regions suggest that whilst useful at describing the burden of disease and treatment, these data sources are unlikely to be central to solving major knowledge gaps due to their cost and complexity. The use of administrative data and data linkage offer an opportunity for efficient data collection in CKD and may represent a cost-effective investment for developing healthcare systems in the future. Novel data capture techniques, such as RPM, may improve CKD data collection, but a thorough understanding of the perspectives of user populations should be considered before their wider implementation.

  • (2023) Bharat, Chrianna
    Despite strong evidence for the effectiveness of a range of interventions to improve the health and wellbeing of people who are dependent on opioids, morbidity and mortality in this population remains higher than that of the general population. There is a need for innovative approaches to monitor and improve the quality use and safety of available medicines, and to better understand risk factors impacting adverse outcomes in this population. In this thesis, routinely collected administrative data on people with opioid dependence in New South Wales (NSW), Australia, were used to investigate medicine use, including Opioid Agonist Treatment (OAT), opioid analgesics, and other psychotropic medicines. Studies in this thesis examined novel methodological approaches to evaluate medicine exposure and quantify risk, both observed and predicted. This thesis used a diverse range of data sources, including controlled drug registries and pharmaceutical claims databases, linked with health service use and mortality records; and implemented a range of statistical methodologies, including generalised estimating equations, Cox proportional hazards models, and deep learning algorithms. Specifically, this thesis aimed to: (i) estimate retention in OAT and identify person, treatment, and prescriber characteristics that are associated with retention; (ii) develop, evaluate and compare models predicting OAT cessation risk at entry to treatment; (iii) examine trends in opioid analgesic utilisation during periods in and out of OAT; (iv) review methods for generating exposure periods from pharmaceutical dispensing data; and (v) evaluating the all-cause and cause-specific mortality risk associated with opioid analgesics, benzodiazepines, gabapentinoids, and OAT. The first study found retention in OAT to be affected not only by characteristics of the person and their treatment, but also of their prescriber, with longer prescribing tenure associated with increased retention of people in OAT. The results from the second study indicated time-to-event prediction models may be limited in their ability to identify individuals at high cessation risk on entry to OAT. Of the methods used in model development, machine learning approaches performed similarly to traditional statistical methods. In the third study, people with opioid dependence were found to have high rates of recent psychotropic medicine utilisation at the time of opioid analgesic initiation, and reduced opioid analgesic dispensing while engaged in OAT. The fourth study describes a novel method for generating medicine exposure periods from dispensing claims data, developed especially for application to medicines with complex and variable dosing regimens. Finally, in the fifth study, benzodiazepines and gabapentinoids appear to increase mortality risk when used in combination with opioid analgesics, although the risk may be reduced when engaged in OAT. This thesis demonstrates the utility of person-level data linkage and innovative analytical methods to generate real-world evidence about the use and outcomes of prescribed medicines among people with opioid dependence. Awareness of harms in clinical settings and evaluating outcome risk during medicine use would give clinicians the ability to understand who needs prevention and treatment services, ensuring efforts and resources are targeted towards those most at-risk. These represent important strategies for improving quality medicine use and reducing harms among people with opioid dependence.

  • (2023) Fitzgerald, Oisin
    Glycaemic control is a core aspect of patient management in the intensive care unit (ICU). Blood glucose has a well-known U-shaped relationship with mortality and morbidity in ICU patients, with both hypo- and hyper-glycaemia associated with poor patient outcomes. As a result, up to 40-90% of ICU patients receive insulin, depending on illness severity and variation in clinical practice. Generally, clinical guidelines for glycaemic control are based on a series of trials that culminated in the NICE-SUGAR study in 2009, a multicentre study demonstrating that tight glycaemic control (a target of 80-110 mg/dL) did not improve patient outcomes compared to moderate control (<180 mg/dL). However, there remain open questions around the potential for more personalised blood glucose management, which real-world evidence sources such as electronic medical records (EMRs) can play a role in answering. This thesis investigates the role that EMRs can play in glycaemic control in the ICU using open access EMR databases, covering a heterogenous 208 hospital USA based patient cohort (the eICU collaborative research database, eICU-CRD) and a large tertiary medical centre in Boston, USA (MIMIC-III and MIMIC-IV). This thesis covers: i) curation and characterisation of the eICU-CRD cohort as a data resource for real-world evidence in glycaemic control; ii) investigation of whether blood lactate modifies the relationship between blood glucose and patient outcome across different subgroups; and iii) the development and comparison of machine learning and deep learning probabilistic forecasting algorithms for blood glucose. The analysis of the eICU-CRD demonstrated that there is wide variety in clinical practice around glycaemic control in the ICU. The results enable comparison with other data resources and assessment of the suitability of the eICU-CRD for addressing specific research questions related to glycaemic control and nutrition support. Informed by this descriptive analysis, the eICU-CRD was used to examine whether blood lactate modifies the relationship between blood glucose and patient outcome across different subgroups. While adjustment for blood lactate attenuated the relationship between blood glucose and patient outcome, blood glucose remained a marker of poor prognosis. Diabetic status was found to influence this relationship, in line with increasing evidence that diabetics and non-diabetics should be considered distinct populations for the purpose of glycaemic control in the ICU. The forecasting algorithms developed using MIMIC-III and MIMIC-IV were designed to account for the intrinsic statistical difficulties present in EMRs. These include large numbers of potentially sparsely and irregularly measured input variables. The focus was on development of probabilistic approaches given the measurement error in blood glucose measures, and their potential conversion into categorical forecasts if required. Two alternative approaches were proposed. The first was to use gradient boosted tree (GBT) algorithms, along with extensive feature engineering. The second was to use continuous time recurrent neural networks (CTRNNs), which learn their own hidden features and account for irregular measurements through evolving the model hidden state using continuous time dynamics. However, several CTRNN architectures are outperformed by an autoregressive GBT model (Catboost), with only a long short-term memory (LSTM) and neural ODE based architecture (ODE-LSTM) achieving comparable performance on probabilistic forecasting metrics such as the continuous ranked probability score (ODE-LSTM: 0.118±0.001; Catboost: 0.118±0.001), ignorance score (0.152±0.008; 0.149±0.002) and interval score (175±1; 176±1). Further, the GBT method was far easier and faster to train, highlighting the importance of using appropriate non-deep learning benchmarks in the academic literature on novel statistical methodologies for analysis of EMRs. The findings highlight that EMRs are a valuable resource in medical evidence generation and characterisation of current clinical practice. Future research should aim to continue investigation of subgroup differences and utilise the forecasting algorithms as part of broader goals such as development of personalised insulin recommendation algorithms.

  • (2023) Carson, Joanne
    Background: Direct-acting antiviral (DAA) treatment is being scaled up to eliminate hepatitis C virus (HCV) as a major public health threat. Aims: The aim of this research was to assess potential barriers to HCV elimination in Australia, including DAA discontinuation, treatment failure, and HCV reinfection. Specific aims included assessing (1) reinfection incidence in prison; (2) real-world effectiveness of retreatment for reinfection and (3) treatment failure; (4) national trends in retreatment for reinfection and treatment failure; and (5) national trends in treatment discontinuation. Methods: In Chapter 2, HCV reinfection incidence was assessed in a prospective cohort of people treated for HCV in four prisons. In Chapters 3 and 4, the effectiveness of retreatment for reinfection and treatment failure were assessed using standard of care data from a national observational cohort that included a broad range of treatment settings. In Chapter 5, a machine learning model was developed and applied to national pharmaceutical administrative data to assess trends in retreatment for reinfection and treatment failure. In Chapter 6, treatment discontinuation trends were assessed in national pharmaceutical administrative data. Key Findings: High HCV reinfection (13/100 person-years) was observed in prisons. Retreatment for reinfection was highly effective (95%), but high losses to follow-up during treatment (25%) were observed. The effectiveness of retreatment for treatment failure (81%) was not impacted by treatment setting, supporting decentralisation of HCV care. A machine learning model with high predictive accuracy (96%) to classify retreatment reason was developed. When applied, an increasing national trend of retreatment for reinfection was observed. Half of retreatment for treatment failure was among individuals discontinuing treatment. National treatment discontinuation rates doubled between 2016-2021, rising to 15%, despite increasing use of simplified shorter duration DAA regimens. Conclusions: Retreatment of reinfection and treatment failure will be crucial to reduce HCV transmission and HCV-related morbidity and mortality. Retreatment can be effectively delivered through decentralised models of HCV care. Increasing trends of retreatment for reinfection and treatment discontinuation correspond with increasing treatment uptake among people who inject drugs. Additional strategies are needed to ensure vulnerable populations achieve and maintain HCV cure.