General practice health assessments of older Australians: Equitable, effective, or improved care?

Introduction By 2040, it is predicted that people aged over 85 years of age will form four percent of the Australian population. In 2005, they were only 1.5 percent of the Australia population. In this age group, a high proportion, with advanced frailty or disability, will be unable to live independently in the community. As a result, it is predicted that there will be a marked increase in the demand for support mechanisms and residential aged care facility (RACF) places for this age group. In 1999, the Australian government provided subsidies for all permanent residents of Australia aged 75 or more years to undergo a prescribed health assessment by their general practitioner (GP) with the expectation that this might lower the need for such care. Two types of assessment with identical content were funded. Item 700 assessments were to be performed solely in the GP's surgery while item 702 assessments required some components of assessment to be conducted in the patient's home. This thesis appraises whether these health assessment items for older Australians have resulted in better health outcomes over the three years since their introduction. Method The Health Insurance Commission (HIC) supplied data from medical service claims under Medicare and Department of Veterans Affairs (DVA) arrangements over the three twelve month periods commencing 1 October 1999 and ending 30 September 2002. The initial cohort sampling frame was all Australians aged 75 years and over at 30 September 1999, who claimed at least one primary care consultation item between 1 October 1999 and 30 September 2000, and who had not made a claim for an attendance at a RACF between 1 October 1998 and 30 September 1999. This sampling frame was chosen as it was the most complete sample of those eligible for a health assessment claim that could feasibly be obtained. In each successive 12 month period, those who made no primary care attendance claim or had a claim for an RACF attendance item were censored. Personal data was available for each claimant for gender, age group at the commencement of each 12 month period, Rural Remote and Metropolitan Area classification of post code of residence, Index of Relative Socio-economic Deprivation (IRSD) quartile of postcode of residence and whether the claimant was eligible for treatment under DVA arrangements. Using logistic regression, the personal characteristics associated with a claim for either an item 700 or 702 were determined to elucidate if heath assessments had been equitably distributed Similarly, using logistic regression to correct for the effects of these personal characteristics, the effect of a health assessment claim on first claims for a RACF attendance over the 36 months of the study was determined. This outcome measure was chosen, as it was a good proxy for permanent admission to a RACF, an outcome shown to be lessened by health assessment in randomised controlled trials. Finally, the effect of having a health assessment in the period 1 October 2000 to 30 September 2001, on medical service claims in that and the following 12 month period was explored using Poisson regression with correction for over dispersion to control for the effects of differences in personal characteristics between those who had each type of assessment and those who had none. Results The study cohort initially comprised 886,185 individuals. Cohort subjects were a virtually complete national sample with the number remaining in the cohort at the time of the 2001 Australian Bureau of Statistics Census being 98% of the census-derived corrected count for this age group living in the community. Over the 36 month period of data collection, 271,939 (31%) claimed for at least one of the two health assessment items, 72,425 (8%) claimed for at least two, 20,027 (2%) claimed for three and 114,040 (13%) individuals from the initial cohort made a first claim for a RACF GP attendance item. Initially, item 700 claims were more prevalent but by the third data collection period, the number of claims for item 700 and 702 were equal. Health assessment claims were more common in females, in the age group 80-84 years, in those living in metropolitan and regional Australia, in those living in the most disadvantaged IRSD quartile postcodes, and in those who had entitlement to medical care under the DVA. Those claiming item 700 were more likely to be male, aged from 75-84 years, live in regional Australia, live in the most disadvantaged IRSD quartile postcodes and be entitled to medical care under the DVA. Those claiming item 702 were more likely to be female, aged 80 or more years, live in metropolitan Australia, live in all but the most advantaged IRSD quartile postcodes and be entitled to medical care under the DVA. Claims for first RACF attendances were more common in those who claimed for an assessment. This was particularly marked for those who claimed an item 702 but was also seen in those who claimed for an item 700 over time. Those who claimed for an assessment item in the period 1 October 2000 to 30 September 2001, claimed for around 20 percent more medical services than those who made no claim. Those who claimed for an item 702 were more likely to make claims for home visits and after hours emergency visits. However, making a claim did not appear to affect the use of services in the following 12 months. Analysis With the exception of those who dwell in rural and remote Australia, health assessments may have improved equity as they were more likely to be claimed by those known to be at risk of poorer health outcomes. Claiming a health assessment did not protect against RACF admission, as indicated by an increased rate of first claims for a RACF attendance. In particular, an item 702 claim appeared to increase such claims. It seems unlikely that the intervention delivered differed between item 700 and 702. Two plausible explanations for this difference are: either the two items were targeted at different populations or item 702 was better at detection of disability requiring RACF admission. The differences in service usage suggest that those who claimed an item 702 were more likely to be frail. Better detection and grading of dementia, the commonest condition resulting in admission to RACFs in Australia and a condition GPs often under-diagnose or whose severity they underestimate, is the most likely reason why an item 702 claim is associated with an increased rate of admission to RACF care. Discussion Health assessments in Australia appear to be worthwhile. They improved the equity of health care services for high risk groups with the exception of those living in rural and remote Australia. While they did not lessen the rate of admission to RACFs as was initially hoped, they may have prevented carer distress and inappropriate crisis hospital admissions by earlier identification of those in need of RACF care. They do not appear to have increased the use of medical services. RCTs conducted in Australia, applying similar health assessment interventions, also did not show the lowering of RACF admission rates seen overseas. The advantages enjoyed by Australians are the most probable explanation for the failure of health assessment in Australia to lower the rate of admission to RACFs. Australians experience excellent health adjusted life expectancies, have low personal cost and easy accessibility to high standard health care and social support, and are only admitted to a RACF if they cannot be supported in the community. In other health systems, improvements to the health care system rather than health assessment alone may be a more beneficial way of improving health outcomes for older persons. An improved influenza immunisation rate is one factor, not reported in published studies, that might explain some of the improvement in health status seen in overseas RCTs of health assessment. Australia has a rate of uptake for influenza vaccine in those aged 65 or more years of around 80 percent, one of the highest rates in the world and well over twice that seen in many other similar developed countries. Influenza is known to accelerate functional decline, and thus result in admission to a RACF. Future trials of measures to lessen these adverse effects should examine the effect of influenza and influenza vaccination on these outcome measures. In summary, health assessments of older persons in Australia in the first 35 months following their introduction, were widely used, improved equity for at risk groups, appear to improve the identification of those at need of RACF admission, and did not increase claims for out of hospital medical services.