Utilising hospital administrative datasets to identify patterns of service use in patients with a primary malignant glioma (#242)
The supportive and palliative care needs of patients with primary malignant glioma (PMG) are not well described despite patients having a high burden of symptoms, psychosocial distress and care. This study examined current patterns of hospital utilisation, thereby providing a base to plan appropriate care.
The Victorian Admitted Episodes Dataset, Emergency Minimum Dataset and Death Index are linked allowing patient care to be tracked over time and across sites of care. Patients admitted to hospital between January 2003 and December 2008 with an incident diagnosis of PMG were included. The outcomes of interest were total hospital bed-days in the last 30 days of life and place of death. Variables were analysed using descriptive statistics and multivariate regression analysis.
1,997 patients were identified, of whom 58.7% (n=1173) died during the follow-up period. During their illness, they had a mean of 4.9 hospital admissions, mean stay of 49 bed-days, and 2.6 emergency presentations. Analysis was restricted to those who died with more than 4 months of follow-up (n=678, 34%).
Supportive and palliative care utilisation steadily increased, particularly at 4-6 months prior to death, with 75.5% (n=512) receiving some palliative care by their death. Just one quarter (26.0%) of patients died outside of hospital. Palliative care in the 90 days preceding the last month of life was significantly associated with low hospital utilisation (RR 0.62, 95% CI 0.46,0.82), and an increased likelihood of dying outside of hospital (RRR:2.16, 95% CI 1.16-4.04, p 0.05).
Palliative care referral earlier in the illness trajectory enabled patients to be cared for and die at home. Linked data sets across hospitals in Victoria give valuable insights into the service use of patients with PMG to inform the development of relevant and responsive models of care.
Acknowledgement of Funding
Victorian Cancer Agency