Variation in spending on primary care
Recently published programme budgets contain, for the first time, the split of spending between primary and secondary settings. Overall, in 2006/07 primary care trusts (PCTs) spent around a third of their budget on primary care, but there is remarkably wide variation with 14 spending over £600 per head while 11 others spend less than half that. Why?
To be fair, this is the first year that the Department has collected this information, and so data quality problems are to be expected - particularly in the dozen or so PCTs at the extremes.
Given the drive to reduce health inequalities and to move resources from secondary to primary care, it could be that PCTs in more deprived areas are investing more of their resources in GPs and community-based activities. However, the data doesn't support that hypothesis.
Whatever the reason for the variation, we might expect PCTs spending more on primary care to achieve better outcomes in due course. Comparison of current spending with the latest outcome information is far from ideal because of the time delay between investment and outcome. But, assuming that current spending patterns are at least in part a reflection of historical patterns, it may nevertheless be informative.
The results are interesting. Those PCTs with higher spending per head on primary care have no better outcomes in terms of mortality (from amenable causes, 2003-2005), after taking account of age, sex, and deprivation, than those with lower spending.
There is a significant link with disease prevalence, however. Higher spending in primary care is actually associated with higher (not lower) prevalence rates as measured in the Quality and Outcomes Framework 2006/07. The example shown opposite is coronary heart disease. One explanation is that higher spending in primary care is still predominantly a reaction by PCTs to treat the incidence of disease, rather than an investment in prevention. Another possibility is simply that the more GPs in an area, the more disease is detected. There seems to be evidence for both.