To what extent do the differences across SES depend on how SES is grouped into the 4 categories? Did you look at different types of SES classification?
Unfortunately we do not have access to the algorithm and data necessary to do the SES classification, DHHS staff were tasked to do this classification. It is not possible to vary the classification methods.
Is DHHS considering applying this method?
DHHS funded this research, but how they make use of the research results is an internal DHHS matter. Unless there are official announcements, we would not have any idea how the research is used for policymaking.
Shouldn't funding allow for preventive care? Rather than focusing purely on hospital care?
Yes, exactly the point, moving to capitation funding should help to prevent the worsening of chronic conditions so that patients do not end up in hospitals.
Were you worried at all about endogeneity of timing of onset?
Endogeneity of SES is an issue, but we are not able to deal with this issue given the nature of the data and that we do not have any information about the algorithm by which SES is classified.
What could be the reasons for the highest disadvantaged having the second highest increase in costs. Shouldn’t we see a continual increase from no disadvantage to highest disadvantage?
Yes, it is a surprise that the high disadvantaged group has somewhat lower utilisation than the moderate disadvantaged group. We suspect it is an access issue – patients in the high disadvantaged group possibly face greater access barriers. But we are not able to test this hypothesis with the data we have.
How would the capitation model decide how the funding should be spent? And how would we know if the amount allocated is appropriate?
In principle, funding should balance the benefit of managing and treating chronic conditions against the opportunity cost of fund. However, the design of a capitation scheme is a complex issue that touches on various health, economics and political considerations. It is beyond the scope of this research, which deals only with one element of the design, namely should SES play a role.
There have been previous examples worldwide that have factored in SES into their funding model. For example, Sweden has introduced a ‘bundled payments’ for particular chronic conditions and the payments have been stratified across 4 ASA scale (SES scale) to allow for more funding for low SES patients. Do you think it will be easier to move away from capitation to another payment structure such as ‘bundled payments’ to reduce inequalities?
Capitation payment models come in different forms and shapes. The main point is to avoid the pitfall of volume-based funding models, namely the misalignment of financial incentives and provision of care. Whether it is called blended payments or bundled payments is not important, the key issue is whether incentives are aligned.
Dr Brendan Murphy mentioned he will focus on policy reforms. Can you share any insight on what these may be?
I suspect “policy reforms” would encompass a broad spectrum of policy changes, of which funding reforms would likely be an integral part. But this is purely a guess.
Bundled payments are considered an intermediary step between fee for service and capitation. Is this a plausible step for the short term?
Some bundled payment schemes are designed to balance the pitfalls of volume-based funding by incorporating an element of capitation. Whether it works well or not depends on the context of the specific health system setting. I do not think there is a short-term or long-term consideration, funding models evolve over time.
In Sweden and other international examples, they have found no cream skimming and utilised very productive outcome measures. Is SES a useful measure because it cannot be easily gamed by the hospital/medical practitioner outcomes?
In the context of Australia’s mixed public-private hospital system, our earlier research found strong evidence of private hospitals engaging in cream skimming. Whether SES can be used to prevent gaming depends on the design and how SES is determined, and the context under which gaming occurs. For example, holding a health care card can be used as an indicator of low SES, and this is unlikely to be gamed.