Non-reimbursement for preventable health care-acquired conditions

Managing health care risk must be evidence based and not harm patients or the common good

The decision by Australia’s largest private health insurer Medibank to not reimburse hospitals for 165 hospital-acquired complications that it deems preventable has the stated objective of improving patient care and thereby containing costs.1 As with all health care claims, this requires assessment of potential efficacy as well as possible perverse or harmful outcomes.

Medibank’s policy appears to be an expansion of the 2008 decision of Medicare and Medicaid in the United States to improve hospitalised patient safety by refusing to pay for selected potentially preventable hospital-acquired conditions. These included care-related infections, pressure ulcers, retained surgical objects, and deep vein thrombosis or pulmonary embolism associated with total knee and hip replacement procedures.

It assumed that most of the adverse events might be prevented by widespread adoption of achievable practices; that the events could be measured accurately; that the events resulted in clinically significant patient harm; and that present-on-admission determination was feasible.2 Although the policy was seen as potentially effective, there was also recognition of the potential for unfairness, gaming and unforeseen consequences. Accordingly, the initial implementation was considered a bold experiment requiring careful monitoring.

A 2012 study found no evidence that financial disincentives reduced infection rates and no situations where patients appeared to benefit from the implementation of this policy.3 Instead, the authors concluded that careful evaluation is needed to determine when these programs work, when they have unintended consequences, and what might be done to improve patient outcomes. The article further reviewed the effectiveness of health insurers paying for performance and observed that evidence on improved patient outcomes is mixed, with concerns that any intended beneficial effect may be mitigated by providers avoiding the most seriously ill patients. Additional detriment may be caused by the pressure to inappropriately over-investigate and over-treat, such as the use of prophylactic antibiotics.4

Broader experience with the use of financial incentives to deliver health policy outcomes indicates that in many cases, such interventions have failed to deliver anticipated benefits. The lack of solid evidence has not deterred attempts to use them as levers for change, thus urgently requiring knowledge of how and why such schemes are implemented, as well as studies that go beyond black-box and control–resistance frameworks.5 Black-box situations are those where the output from a given input is apparent but knowledge of the actual and often complex determinants is opaque, thus risking erroneous inferences about cause and effect. Control–resistance is the dynamic between change and opposition to such change. Although resistance by medical professionals may be ascribed to self-interest, case studies do not support this.6 Instead, resistance may relate to a lack of policy fairness and legitimacy. Overall, there is currently no clear evidence to recommend widespread implementation of financial levers to drive health care quality improvements, and doing so may simply incur unintended and dysfunctional consequences. Genuine commitment to patients will involve applying health care initiatives that reliably improve quality and safety, while undertaking research that adds to a solid evidence base.

Current best practice is to proactively produce a health care environment in which controllable factors, such as inevitable errors and systematic failures, do not reach patients. This is achieved through adoption of a culture of safety, careful implementation of health information technology, robust process improvement, and clinical simulation.7 These measures should be instituted with the understanding that clinicians and support staff are already striving to provide patients with high-quality care, and the inability to meet performance measures is an opportunity for growth of the system rather than punishment of providers.8 Engineering principles and human factors analysis influence the design of these systems to ensure that they are safe and reliable. There are self-evidently also humanitarian and professional reasons that drive clinicians to deliver optimal patient outcomes, and powerful deterrents against harm such as suffering, reputational damage, legal liability, as well as potential professional and regulatory sanctions. These would seem to be of far greater concern to most practitioners than non-reimbursement.

This literature review suggests that Medibank’s policy to not reimburse selected hospital-acquired complications is unlikely to significantly contribute to the stated objective of patient safety. Instead, there are proven methods for addressing that goal, as well as established and powerful professional obligations to do so. The literature also identifies the need to consider additional consequences of such a policy, such as unfairness and unforeseen outcomes.

Medibank’s non-reimbursement policy targets conditions such as falls resulting in fractures or intracranial injury, health care-associated infections and surgical complications such as venous thromboembolism. Such events are more prevalent in older hospitalised patients, associated with the multimorbidity seen with longevity.9 These intrinsic age and frailty risks cannot be eliminated, even with the best culture and systems. For example, venous thromboembolism is one of the leading causes of post-operative morbidity and mortality in patients with hip fracture. Extensive knowledge and application of current evidence helps to reduce but cannot eliminate the risks of bleeding or thrombosis. Hospitalised older patients are also at an increased risk of infections, ulcers, delirium, refracturing and readmission, among other complications. Best-practice guidelines assist clinicians with their decision making and systems but cannot eliminate the risks from clinical and biological complexity and imperfections.10 Mobilisation following hip repair is equally fraught, with complications such as premorbid balance problems, cognitive and sensory impairments and medication side effects requiring complex benefit–risk decisions. Clinical teams working with older patients know that the risk of adverse events is a constant factor that must be recognised, managed and reasonably minimised, but it is inherent and therefore present in practically every therapeutic intervention.

Currently, patients or their surrogate decision makers consent to stated clinical risks in return for anticipated benefit. The intrusion of unrealistic and unavoidable non-reimbursement risk could oblige providers to make an economically rational decision to avoid interventions associated with this imposed financial risk, irrespective of patient clinical need or risk acceptance. In practice, this will disadvantage policy holders most at risk of non-reimbursable events, such as elderly, obese or seriously ill patients. Non-reimbursement could also be perceived as a statement of negligence, exposing the provider to malpractice claims and effectively further discouraging acceptance of high-risk patients. Even if individual clinicians were prepared to risk non-reimbursement, this is unlikely to be sustainable for private and not-for-profit hospitals with significant overheads.

Indirectly restricting access to health care insurance policies maintained over many years in anticipation of later life or critical need is an issue for the Australian government, which must ensure equity and fairness for particularly older policy holders, especially those who have invested in lifetime cover. Further, the legislated principle of community rating prevents private health insurers from directly or indirectly discriminating between people on the basis of their health or for any other reason. Taxpayers should also be concerned if high-risk Medibank patients are restricted through application of this policy to accessing public hospitals as public patients, thus effectively shifting significant costs from Medibank to predominantly state and territory governments.

It will be apparent that this policy has been introduced following the public listing of Medibank Private on 25 November 2014. This has brought with it pressure to deliver on share market expectations. Although a non-reimbursement policy might deliver short-term gains, in the longer term informed consumers are likely to avoid health insurance policies that do not offer sustained customer value.

The long-term interests of both Medibank’s shareholders and policy holders would arguably be better served by evidence-based initiatives that are also likely to be ethical and socially responsible. This requires constructive engagement with health care providers, supporting proven proactive quality practices that acknowledge inherent and unavoidable clinical risk. Like other payers in the health care system, the private health insurance sector has an important role in engaging ethically with the health professions and industry to ensure a positive health care return on finite investment in our nation’s health.