Long term health outcomes of Indigenous and non-Indigenous Australians in the Northern Territory after serious illness
: a survival analysis approach

  • Vincent Yaofeng He

    Student thesis: Doctor of Philosophy (PhD) - CDU


    This PhD thesis used survival analysis methods in a novel way to examine differences in long-term outcomes of patients diagnosed with serious, acute conditions in the Northern Territory of Australia. In particular, the thesis compared long-term outcomes for Indigenous patients against those for non-Indigenous patients. It is intended to function at two levels: at the theoretical level, this thesis investigates the potential of using survival analysis to produce useful information from existing data to improve our knowledge about long-term outcomes after acute care. At the practical level, it undertakes four illustrative studies to examine the long-term health outcomes for Indigenous patients compared with non-Indigenous patients after diagnosis and treatment for four severe and acute health conditions: sepsis, stroke, cancer and rheumatic heart disease (RHD).

    All four studies found that long-term health outcomes after serious illness were poorer for Indigenous people than they were for non Indigenous people. Indigenous sepsis patients had higher excess mortality. Indigenous stroke patients had higher case-fatality, stroke recurrence and long-term mortality; comorbidities explained part of the Indigenous disparity in longterm survival after a stroke, but did not explain the disparity in stroke recurrence. Indigenous RHD patients had higher mortality; of which 28% of the excess mortality was explained by comorbid renal failure and hazardous alcohol use. Indigenous cancer patients had a lower net survival rate for breast, colorectal, lung, head and neck cancers. In the presence of competing risk, Indigenous cancer patients with all types of cancer had higher probability of cancer death and higher probability of non-cancer death (except for those with lung cancer and head and neck cancers - both smoking-related cancers).

    This thesis has demonstrated the great potential of survival analysis to produce useful information from linked administrative data to improve our knowledge about long-term health outcomes.

    Date of AwardDec 2015
    Original languageEnglish
    SupervisorJohn Condon (Supervisor), Yuejun Zhao (Supervisor) & Joshua Davis (Supervisor)

    Cite this