Estimating unplanned and planned hospitalization incidents among older Australian women aged 75 years and over: The presence of death as a competing risk.

OBJECTIVES To estimate the incidence of unplanned and planned hospitalization and identify associated factors among older women aged 75 years and over. METHODS This study is a prospective longitudinal cohort study (over the period 2001-2016). Women born between 1921 and 1926 were included from the Australian Longitudinal Study on Women's Health (ALSWH). ALSWH self-reported data were linked with New South Wales's state Admitted Patient Data Collection. A competing risk analysis was performed using SAS v 9.4. RESULTS Overall, during the 15-year observation period, 86.7% of women experienced at least one unplanned admission and 60.3% experienced at least one planned admission. The complement of Kaplan-Meier survival function overestimates hospitalization incidence compared to cumulative incidence function in the presence of a competing risk, that is, "death." Predisposing factors (older age and not partnered) and need factors (hospital doctor visit, GP or family doctor visit, poor perceived general health, and having at least one chronic disease) were associated with increased unplanned admission. First language other than English and not having private health insurance showed a negative association with planned admission; specialist doctor visits had a positive association. CONCLUSIONS Hospital admission was influenced by predisposing and enabling factors, as well as need. Intervention aiming to support older people who do not have private health insurance may be required to improve the goal of equal access to healthcare through planned admissions people with equal need.