High-cost users of hospital beds in Western Australia: a population-based record linkage study.

OBJECTIVE: To describe how high-cost users of inpatient care in Western Australia differ from other users in age, health problems and resource use. DESIGN AND DATA SOURCES: Secondary analysis of hospital data and linked mortality data from the WA Data Linkage System for 2002, with cost data from the National Hospital Cost Data Collection (2001-02 financial year). OUTCOME MEASURES: Comparison of high-cost users and other users of inpatient care in terms of age, health profile (major diagnostic category) and resource use (annualised costs, separations and bed days). RESULTS: Older high-cost users (> or = 65 years) were not more expensive to treat than younger high-cost users (at the patient level), but were costlier as a group overall because of their disproportionate representation (n = 8466; 55.9%). Chronic stable and unstable conditions were a key feature of high-cost users, and included end stage renal disease, angina, depression and secondary malignant neoplasms. High-cost users accounted for 38% of both inpatient costs and inpatient days, and 26% of inpatient separations. CONCLUSION: Ageing of the population is associated with an increase in the proportion of high-cost users of inpatient care. High costs appear to be needs-driven. Constraining high-cost inpatient use requires more focus on preventing the onset and progression of chronic disease, and reducing surgical complications and injuries in vulnerable groups.

Health care financing and public responses: use of private insurance in Western Australia during 1980-2001.

OBJECTIVES: The aim was to identify and explain trends and cut points in payment classification (privately insured or otherwise) for episodes of hospitalisation in Western Australia. METHODS: Hospital morbidity data from 1980 to 2001 were used to produce trend lines of the proportion of hospital separations in each payment category in each year in age and clinical subgroups. RESULTS: The most significant changes in payment classification over time were found in all groups between 1980 and 1984, corresponding to a period when free public hospital care in Australia was abolished (Sep 1981 to Feb 1984). The trend associated with this policy change rebounded significantly just before the introduction of Medicare in 1984. These observations were consistent over all age groups except in the oldest group (70+ years). This trend was more pronounced for the surgical subgroup compared with other broad clinical categories. More recently, a trend towards increasing public episodes was reversed from 2000 following introduction of incentives for private health cover and sanctions against deferred uptake in younger people. CONCLUSION: The public appeared to adopt a short-term crisis reaction to major policy change but then reversed towards past patterns of behaviour. The implications for policy makers include the need to understand the underlying culture of the population; to realise that attitudes become fixed as people age; and to recognise the difference in the effectiveness of incentive- and deterrent-based policies.

The use of end-quintile comparisons to identify under-servicing of the poor and over-servicing of the rich: a longitudinal study describing the effect of socioeconomic status on healthcare.

BACKGROUND: To demonstrate the use of end-quintile comparisons in assessing the effect of socio-economic status on hospital utilisation and outcomes in Western Australia. METHODS: Hospital morbidity records were extracted from the WA Data Linkage System for the period 1994-99, with follow-up to the end of 2000. Multivariate modelling was used to estimate the effect of socio-economic status on hospital admission rates, average and total length of stay (LOS), cumulative incidence of readmission at 30 days and one year, and case fatality at one year. RESULTS: The study demonstrated higher rate ratios of hospital admission in the more disadvantaged quintiles: rate ratios were 1.31 (95% CI 1.25-1.37) and 1.32 (1.26-1.38) in the first quintile (most disadvantaged) and the second quintile respectively, compared with the fifth quintile (most advantaged). There was a longer total LOS in the most disadvantaged quintile compared with quintile 5 (LOS ratio 1.24; 1.23-1.26). The risk of readmission at 30 days and one year and the risk of death at one year were also greater in those with greater disadvantage: the hazard ratios for quintiles 1:quintile 5 were 1.07 (1.05-1.09), 1.17 (1.16-1.18) and 1.10 (1.07-1.13) respectively. In contradiction to the trends towards higher hospital utilisation and poorer outcomes with increasing social disadvantage, in some MDC's the rate ratio of quintile 1:quintile 2 was less than 1, and quintile 4:quintile 5 was greater than 1. For all surgical admissions the most disadvantaged had a significantly lower admission rate than the second quintile. CONCLUSION: This study has shown that the disadvantaged within Western Australia are more intensive users of hospital services but their outcomes following hospitalisation are worse, consistent with their health status. Instances of overuse in the least disadvantaged and under use in the most disadvantaged have also been identified.

Health outcomes in people with type 2 diabetes. A record linkage study.

INTRODUCTION: This study pilots a method of measuring health outcomes in a general practice population of patients with type 2 diabetes. METHOD: The Diabetic Register of the Perth and Osborne Divisions of General Practice was linked to the Western Australian Health Services Research Linked Database. RESULTS: Of the 487 patients in the study, 332 (68%) had been admitted before their diagnosis of diabetes (40% with a diabetes related condition), and 56% were admitted postdiagnosis (55% with a diabetes related condition). The admission rate increased with age and duration of diabetes. DISCUSSION: The data show that a large proportion of diabetic patients suffer from serious comorbidity both pre- and post-diagnosis and demonstrate that their hospital admission rate is higher than that in the general population. CONCLUSION: The project demonstrates that linked hospital morbidity data can be used to monitor health outcomes in a general practice population of diabetic patients.