Several aspects of this direction could catch out those who treat it as a routine procedural instrument.
The definition of “avoidable hospital readmission” is circular. It is defined as readmission within 28 days “with a particular focus on avoidable readmission within 5 days”. The term “avoidable” itself is not defined. The direction then says the focus should be on readmissions “for a condition or conditions arising from complications of the management of the original condition”. This is a proxy for avoidability, but it still relies on causal linkage. Proving that a readmission is due to complications of management rather than the natural course of disease is clinically contentious and often ambiguous. The model will need a robust adjudication mechanism to avoid penalising hospitals for unavoidable readmissions.
The sentinel events model is to be implemented from 1 July 2017, only ten months after the direction’s issue - and the advice was due 30 November 2016. That leaves just over a year from the advice deadline to implementation. This timeframe was extremely ambitious, particularly given the design principles require a “comprehensive and risk-adjusted model”. Risk adjustment for rare events like sentinel events is statistically challenging because the small numbers make reliable hospital level comparisons difficult. A hospital might go years without a sentinel event and then have two in a year without any change in quality. The model would need to account for random variation, and the direction does not specify how to handle this.
The direction requires IHPA to “have regard to” the Parties’ intention to implement models, but the Parties are not bound to implement anything. This means IHPA must design models that the Parties said they intend to adopt, but the advice could be ignored. IHPA’s effort could be wasted if political priorities shift. The direction does not guarantee implementation even if IHPA produces excellent work.
The requirement to “give consideration to any probable known costs and expected benefits” is sandwiched between aspirational language. The direction does not say what happens if costs exceed benefits. It does not make implementation conditional on a positive cost benefit analysis. It simply requires IHPA to consider them, leaving the final decision to the Minister and COAG.
The definition of “hospital acquired condition” includes the phrase “for which clinical risk mitigation strategies may reduce (but not necessarily eliminate) the risk.” This is vague. It could cover almost any complication. The Australian Commission on Safety and Quality in Health Care will define the specific conditions, but the direction does not limit the list to conditions that are highly preventable. If the list includes conditions with small attributable fractions, hospitals may be penalised for outcomes they can barely influence, creating perverse incentives to avoid admitting high risk patients.
The design principle (c)(i) says “the financial levers are designed to ensure there is transparency between the approach and the intended outcome.” This is unclear. It might mean that the funding adjustments should be clearly linked to the adverse event (i.e., the amount deducted should be transparent and understandable). Alternatively, it could mean the methodology for calculating adjustments should be transparent. In either case, achieving transparency when using complex risk adjustment models is difficult. The models may become black boxes, undermining trust.
The direction does not specify an appeals or review process for hospitals, states, or clinicians. If a hospital believes it was incorrectly penalised for a sentinel event or readmission, there is no mechanism set out in this direction. That would need to be designed as part of the model. The lack of such detail in the direction means the IHPA advice must include a robust governance framework, or else implementation will face resistance.
The direction refers to “the Parties’ intention to send a signal at the health system level” but also says models should “not compromise state system financial sustainability”. Sending a strong signal requires meaningful financial consequences. If the adjustments are small enough to preserve state financial sustainability, the signal may be too weak to change behaviour. This tension pervades the direction.
The advice deadline of 30 November 2016 fell after the 2016 federal election (July 2016) and before the 2017-18 budget cycle. It is a tight deadline for what is essentially a major policy development requiring data analysis, consultation, and consensus building among states. IHPA would have had to reprioritise its staffing and resources, potentially at the expense of its core pricing work.
The direction requires IHPA to “have regard to” submissions from “other parties deemed relevant”. This gives IHPA discretion to invite or ignore submissions from interest groups like the Australian Medical Association, private hospitals, or patient advocacy organisations. The advice could be skewed if IHPA chooses not to consult broadly. The direction does not mandate public consultation.