HDR Research

Unemployment and Youth Mental Health Study

  • Researchers: Andrew Kellett, PhD Candidate under the supervision of Prof Janette Perz, Prof Jane Ussher, Dr Elizabeth Conroy and Dr Vanessa Rose (Adjunct).

Young people aged 17-25 experience higher levels of unemployment and psychological distress compared to the general population. This is concerning as unemployment has been associated with increased risk of suicide. Young people may also find it difficult to access services and support because of their unemployment and the need to spend time looking for work. Whilst cognitive behavioural therapy has proven effective in face-to-face settings in improving the mental health of youth unemployed, there is also the potential for more online self-help interventions given the high internet usage by people in this age group.

This mixed-methods study aims to address gaps in knowledge through an exploration of young people's experiences of unemployment. The study will explore young people's experiences of having participated in an online mental health intervention ('Walk the Talk') and any feedback they have on the program. The study will also explore gender differences in the experience of unemployment. Results will assist in designing more effective and appropriate mental health interventions with this group.

You are invited to participate. Find out more (opens in a new window)

A Qualitative Exploration of Treatment Experiences in Men with an Eating Disorder.

  • Researchers: Priyanka Thapliyal, Doctoral Candidate under the supervision of Phillipa Hay, Janette Perz, Paul Rhodes, Deborah Mitchison. Funded by Western Sydney University.

Eating disorders in men go unreported or undetected in many cases because it is considered as predominantly a female problem. The research project addresses the experiences of men with eating disorders, a neglected area in research. The study will use qualitative methodology in order to explore experiences of men seeking help for an eating disorder, and their experiences of treatment. It will investigate gender related issues and will explore the perceived barriers in men that prevent them from accepting eating disorder as a problem and seeking treatment and difficulties experienced as well as positive aspects of treatment. Larger studies focusing on treatment experiences are needed to explore and expand on the issues of the role and nature of men only services versus mixed gender services and what features of services may better engage and motivate men in therapy.

Homelessness and Women's Health Study

  • Researchers: Amy Dryden, PhD Candidate under the supervision of Prof Jane Ussher, Prof Janette Perz and Dr Elizabeth Conroy

Women comprise approximately 43% of the homeless population in Australia. An additional burden is placed on homeless women of reproductive age due to their unique sexual and reproductive health needs. This includes the need to manage menstruation privately and hygienically whilst living in public space; the need for adequate access to effective forms of contraception and family planning options; and the need for access to sexual and reproductive healthcare services which do not discriminate against women based on their homelessness. These areas of health have the potential to profoundly impact homeless women's lives, yet to date remain unstudied within an Australian context.

This mixed-methods study aims to address gaps in knowledge through exploration of homeless women's experiences of menstrual health, fertility and fertility control. The study will also explore any feedback women may have for the healthcare services which provide them with sexual and reproductive healthcare. Findings of this research will increase knowledge of the experiences and practices of homeless women regarding their menstrual health, fertility and fertility control. Results will highlight the specific needs homeless women have regarding their sexual and reproductive health, as well as women's suggestions for how to improve services which assist them with this area of healthcare.

You are invited to participate. Find out more (opens in a new window)

Providing sexual and reproductive health care to culturally and linguistically diverse women in Australia - Health care professionals perspectives

  • Researchers: Zelalem Mengesha, Doctoral Candidate under the supervision of Prof Janette Perz, Prof Jane Ussher and Dr Tinashe Dune

Cultural and linguistic diversity in Australia is increasing with more than one in four (26%) Australians being born overseas and an additional 20 percent have either one or both parents born in a foreign country. Women from refugee and migrant backgrounds often experience poor sexual and reproductive health outcomes in host countries. These Women also tend to face several barriers in accessing sexual and reproductive health care at individual, system and provider levels in host countries. In Australia, barriers to sexual and reproductive health services by refugee and migrant women and perspectives of improving access and utilization of sexual and reproductive health care are addressed exclusively from the perspectives of patients/clients. Understanding the challenges of providing these types of services to refugee and migrant women from the health care providers' perspectives is crucial in ensuring that the sexual and reproductive health needs of refugee and migrant women are met.

You are invited to participate in a research study. Find out more (opens in a new window)

Experiences of Transgender Parents in Australia

  • Researchers: Rosie Charter, Doctoral candidate under the supervision of Prof Jane Ussher and Prof Janette Perz

In Australia, little is known about transgender parents, their experiences of parenting and how they foster their own wellbeing. Since the mid 1990's there has been a growing body of research on lesbian, gay, bisexual and transgender (LGBT) families however, the unique experiences of trans parents are often lost in this broader framework LGBT framework. As such, the focus of this research is to explore the experiences of transgender parents in Australia.

You are invited to participate in the Experiences of Transgender Parents in Australia Survey (opens in a new window)

You are invited to participate in the Experiences of Transgender Parents in Australia - Partner Survey (opens in a new window)

Detecting clusters and anomalies in Medicare MBS claim trajectories

  • Researchers: Farshid Hajati, Federico Girosi, Shima Ghassem Pour and researchers from the National Health Performance Authority (NHPA). Funded by Capital Markets CRC ($50,000).

The provision of health care services is an activity that takes place over a period of time, during which the patient leaves a trail of administrative records usually referred to as "health trajectory" or "claim trajectory". It is of importance to be able to find groups (clusters) of patients whose trajectories are similar, as well as patients whose trajectories are different from the trajectories of everybody else. This problem is difficult because it requires defining a notion of similarity or distance between trajectories, a task that does not have a natural solution. The challenge of the project is therefore not in finding an appropriate clustering algorithm, but rather in finding a meaningful way of representing health trajectories and measuring their distances. We will develop an array of tools that can be used to find clusters and anomalies in Medicare MBS claims trajectories. The methods will draw on well tested techniques used in signal processing, such as Deep Learning and Deep Belief models. The project will proceed in three parts: 1) Develop algorithm for clustering; 2) Develop methods that analyse and interpret the newly found clusters; 3) Validate the methods on unseen data: predict what type of trajectory an individual is expected to follow and check whether this is observed in the data.

Tools for the design of customizable geographies and catchment areas

  • Researchers: Ludovico Pinzari, Federico Girosi, Shima Ghassem Pour and researchers from the National Health Performance Authority (NHPA). Funded by Capital Markets CRC ($50,000).

There is much interest among health care stakeholders in studying the geographic variation of health related indicators, since it is commonly agreed that part of the unexplained variation may point to inefficiencies and inequities in the provision and access of health care services. Geographic variation is usually studied using geographies defined by administrative boundaries, such as SLAs or Medicare locals. These, however, are not necessarily appropriate, since they were designed using criteria that are not related to most indicators of interest.

In this project we will develop tools that can be used to design geographies that are optimally suited to study the variation of specific indicators, and that allow users to "customize" them, that is specify properties that need to be satisfied. We will focus on catchment areas, that is geographies used to study variation in the utilization of services such as GP visits or hospital admissions. The design of catchment areas is notoriously difficult, and there is no general agreement on how catchment areas should be designed. The problem of designing catchment areas will be cast in the framework of graph theory, and the algorithm used will be akin to a graph partitioning algorithm that optimizes an objective function that encodes the user's requirements.

Customizable peer-grouping of geographic areas

  • Researchers: Ludovico Pinzari, Federico Girosi, Shima Ghassem Pour and researchers from the National Health Performance Authority (NHPA). Funded by Capital Markets CRC.

Studying geographic variation of health indicators can provide useful insight in many aspects of the health care system. Insights are derived by comparing indicators across different geographies such as SA2, SA3 and PHNs. However, it is only meaningful to compare a geography to a geography that is somewhat "similar", where similarity could refer, for example, to similarity in the distribution of socio-economic status, or age or any combination of other variables. In this project we develop a methodology to divide a set of geographies, such as SA3s, in "peer-groups", that is geographies that share certain similarities and such that comparisons within the peer-groups are meaningful and informative. The notion of similarity include characteristics such as population size, population age distribution, distribution of remoteness and distribution of socio-economic status.

Limitations of the Australian refined diagnosis-related groups (AR-DRGs) and their implications

  • Researchers: Michael Hart, Federico Girosi and Shima Ghassem Pour. Funded by Capital Markets CRC ($50,000).

The Australian Refined Diagnosis Related Groups (AR-DRGs) is the Australian classification system for episodes of hospital admissions. Its main purpose is to provide the basis of a prospective payment system in which payers reimburse the provider a fixed amount for each episode. Each AR-DRG represents a class of patients that, ideally, utilize the same amount of hospital resources. While there is anecdotal evidence that the current DRG system is not serving the public well, there is no systematic study of its shortcomings and how they could be improved.

In this project we will explore several issues related to the inefficiencies of AR-DRGs. One issue is the quality of ICD coding: if different hospitals code differently then the distribution of DRGs observed in the data may be inaccurate and lead to incorrect conclusions. At the same time we are exploring the role played by chronic conditions in explaining length of stay within the same DRGs. It is hypothesized that DRGs do not capture sufficiently well the complexity of cases involving multiple chronic conditions, and we are currently attempting to quantify to what extent this is the case. In addition we are also studying the distribution of DRGs across hospitals, attempting to cluster hospitals according to their case mix.

Machine learning methods for the definition of more efficient Australian refined diagnosis-related groups (AR-DRGs)

  • Researchers: Amir Marashi, Federico Girosi and Shima Ghassem Pour. Funded by Capital Markets CRC ($50,000).

The Australian Refined Diagnosis Related Groups (AR-DRGs) is the Australian classification system for episodes of hospital admissions. Its main purpose is to provide the basis of a prospective payment system in which payers reimburse the provider a fixed amount for each episode. Each AR-DRG represents a class of patients that, ideally, utilize the same amount of hospital resources. While it is relatively simple to define what it means for a DRG classification system to be "good", it is not easy to operationalize it, and the problem of defining homogenous categories of patients is difficult because it is plagued by a combinatorial explosion of possibilities. We anticipate that both unsupervised and supervised methods of machine learning will be required. We plan to use 3 years of data to build up to 1,000 groups (clusters) as homogeneous as possible. The groups will be constructed using rules that allow to classify any new patient, and the classification algorithm will be applied to a 4th year of test data. We will then compare, on the test data, the new rules and the current DRGs and validate to what extent the new groups could be used as the basis for a better DRG system.