Purchasing better, innovative and integrated health services

About 50 years ago, Paul McCartney and John Lennon complained that ‘money can't buy me love’; based on the Commonwealth Fund's 2014 review, it would appear that money cannot buy a credible health system either.1 Although comfort might be derived for the Fund's top ranked health systems (i.e. the UK, followed by Switzerland and Sweden), using the Institute for Healthcare Improvement's Triple Aim as a yardstick (i.e. better individual and population health, and reduced health costs),2 none of the Organisation for Economic Co-operation and Development (OECD) countries would be considered successful. The top ranked UK National Health Service (NHS) is already financially distressed,3, 4 reform is underway but is likely too little and too late5, 6 and further innovation is considered urgent.7, 8 Key processes of reform, such as a scheme to improve the quality of primary healthcare,9-15 and attempts to shift healthcare to community settings for older people16 have failed. If that is the state of the top ranker, little wonder the concern then that exists for the US, which was ranked last of the 11 nations considered by the Fund – despite an expenditure on healthcare that is more than twice that per capita in the UK.1 The New Zealand health-funding situation is typical. Between 2000 and 2008, productivity was apparently lost in most domains of the health service,17, 18 despite significant increases in costs.19 An unpublished analysis in 2013 by the New Zealand Institute of Economic Research showed that although pharmaceutical inflation was negligible – consequent to constraint on subsidies and bulk purchasing – total health cost growth was still estimated as 8.5% p.a., compared to the equivalent measure of gross domestic product (GDP) growth (nominal GDP) of about 6% p.a. For all OECD nations, healthcare is increasingly unaffordable.1 There are two issues. The first of these is health funding, which is challenged by an ageing demographic and increasing chronic disease burden. This challenge exists whether or not systems are largely insurance or taxation based, or blended. Not surprisingly, there is increasing attention to social insurance, employment-based, group and not-for-profit mutual schemes, such as those operated in Germany, the Netherlands, Singapore, South Korea and Switzerland. We are aware of, and encourage, putative models of health insurance that promote compliance in people with chronic disease and consequently reduce the consumption of healthcare. However, the subject of this editorial is the second issue, which is how health services are purchased, and in particular, how purchasing can result in better, innovative and integrated services. We can learn from historical approaches to purchasing (Tables 1,2). We have alluded to the difficulties faced by the NHS and in the US. The failure of the NHS primary care pay-for-performance scheme warrants analysis. Despite the framework title (i.e. Quality and Outcomes Framework), ‘activity’ was incentivised.9-15 Not surprisingly, these activities and consequent costs increased, but subsequent studies have not shown any improvement in acute or chronic disease, or in population health outcomes.9, 12-15 The conclusion is that whatever is incentivised needs to be intrinsically valuable.20 The observations that increased commitments in the UK to community-based healthcare do not result in fewer admissions of older people to NHS hospitals,16 and that a greater relative use of hospice care in the US increased the costs born by Medicare for the last year of life,21 also merit notice in an environment of constrained health funding where investment needs to be balanced by reduced future cost growth. The problem is that the counterfactual is hardly ever well-defined and what tends to happen is that hospital capacity is used at a lower level of acuity. Moreover, when it comes to investing closer to home to reduce hospital costs, unless hospital investment is actually slowed, the reliance is on hospital disinvestment and this is almost impossible politically. In addition, marginal hospital costs can be quite low once the overhead is sunk. It is important to be sure that cost avoided benefits are real (counterfactual) and that a credible plan exists for realising them. Pay-for-performance schema also have a poor track record in the US, perhaps because they are not generally based on sound behavioural economics principles.22-25 Although there has been a shift of doctors into salaried practice,26 and variable degrees of bundling payments,27 much of the healthcare in the US is still purchased by way of paying for transactions. The extraordinary cost of healthcare in the US, in comparison to determined quality,1 is well recognised and has led to: the American College of Physicians requiring in their ethics manual that physicians be parsimonious in regard to using healthcare resources;28 the Department of Health and Human Services revising the payment goals for Medicare and Medicaid;29 calls for funding that promotes quality healthcare and valuable outcomes and an effective balance between coordination and competition;29, 30 support for incentivising health practices that reduce readmissions31, 32 and the National Commission on Physician Payment Reform has suggested the phasing out of fee-for-service payments altogether.33 It is noteworthy in this context that changes in healthcare expenditure in the US largely mirror changes in GDP,34 and that nurse participation in the workforce is strongly influenced by the overall economic situation,35 which are factors that are not controllable by health purchasers. Finally, data from Medicare in the US suggests that the quality of healthcare increases and the costs of that care decrease as the number of general practitioners per capita in a region goes up and the number of medical specialists per capita goes down.36 The point is that not only does purchasing need to stimulate desired outcomes, but it also needs to have a positive impact on workforce composition and accessibility. Canada has a fragmented healthcare system that is linked nationally by principles of universality and comprehensive care.37 Despite significant recent increases in health funding to an annual commitment of almost 11% of GDP, only the US health system was ranked lower than that in Canada by the Commonwealth Fund.1 Notably, with the possible exception of some innovative groups in Ontario, the most common method of reimbursing physicians remains a fee-for-service; and, the purchasing of healthcare in Canada is largely focused on hospitals and doctors. Not surprisingly then, the Canadian health system is increasingly unable to satisfy the community-based healthcare needs of an ageing population.37 Canada also has a very mal-distributed medical workforce,38 which makes sophisticated purchasing difficult. We have five observations on healthcare purchasing in Australia that are germane to this editorial. First, the split of purchasing healthcare between private and public funders and, most importantly, between the federal and state governments significantly limits any integration of primary and secondary/tertiary healthcare. Second, most healthcare is purchased through payments for transactions (i.e. activity-based funding and efficient hospital pricing, and fees-for-service through the Medicare Benefits Schedule). In this context, there is an almost linear relationship between the number of Australian medical providers and the services delivered.39 A Cochrane Review examined the impact of payment systems in primary care.40 Although more services are delivered in a fee-for-service system compared with those that are capitated and or salaried (and hence at a greater cost), there is no difference in patient outcomes. Third, some population-based purchasing is in place, but this is essentially an outcome-free bulk fund for healthcare providers. Fourth, the track record of many health services is characterised by lax financial discipline, and this is in part the direct result of poor purchasing practice. The associated absence of a capital charge has led to perverse behaviours, such as the use of maintenance budgets for operational expenses, the running down of resources and the politicisation of replacement facilities and equipment. Ironically, at a time when the disease burden is increasingly because of chronic disease, Australian health systems still celebrate the opening of new (additional) hospital beds – perhaps because in the absence of quality indicators, the Australian public interprets more hospital beds as equating to better healthcare. Fifth, purchases of primary care services are generally neither quality nor outcome oriented – an example is paying providers for chronic care plans. What is needed is not the payment for a chronic care plan (i.e. ‘a piece of paper’), but rather the purchasing of a health outcome that is facilitated by a chronic care plan. By comparison, New Zealand purchases healthcare from District Health Boards (DHBs) through a population-based formula (PBF) and primary care is largely purchased by way of a capitation. Although the PBF is demographically adjusted, it is decreasingly effective as a proxy for population health need. It mixes existing service costs with need and, as such, it is hard to identify outcome measures that have real ‘bite’ at the aggregate level. It has also proved difficult to resist the temptation to protect districts with little population growth (by setting a minimum annual funding increase, effectively over-funding those populations). Finally, it does too little to encourage earlier intervention targeted at segments of the population where the future return to both the individuals and to the taxpayer is likely to be highest. As long as the DHBs under-invest in managing future cost pressure similarly, the implicit deal of living within a fixed real per capita budget is not enforceable and costs consequently accrue to future taxpayers. Primary care capitated funding was introduced to secure co-payment regulation and consequently to incentivise people to access primary care. However, it has had an unintended and perverse financial incentive effect on providers. Based on New Zealand Medical Council data,18 from the time primary care capitations were introduced in 2001 until the present, the average general medical practitioner has given up about a day of work per week in their usual workplace and the number of hours they do on-call or after-hours per week has dropped from an average of ten to about four. While other forces may have influenced these outcomes, they are consistent with the financial incentives created by capitation. We are also concerned about the extent to which the range of subsidies for primary care has actually addressed health inequities. Suggestions for improvement are aimed at better targeting subsidies to individual consumer's income and health needs. Consequently, more focus needs to be given to the impact of financial incentives on service design and delivery. Tying subsidies more closely to individuals' circumstances and having their enrolment choices drive funding will be much more effective the better informed consumers are about the quality of service delivered by different providers. Similarly, funding of primary providers that is not driven by consumer choice needs to be more closely linked to outcomes for patients and so encourage the sort of provider collaboration necessary to deliver good patient outcomes. Although it could be argued that combining capitation and fee-for-service approaches to primary care means that the worst of each scheme is modified, there is no evidence to this effect and the better solution is to align financial and non-financial incentives to obtain better patient outcomes. Funding and accountability arrangements can be characterised according to the specificity (i.e. ‘tight’ or ‘loose’) with which: desired outcomes are defined ex ante; funders define how those outcomes should be achieved and providers are held accountable for delivery. As such, a tight-loose-tight (TLT) approach means that providers are held accountable and have some financial risk around delivering a well-defined result, but are free to innovate in how the result is achieved. There are three basic types in practice. Our primary recommendation is that healthcare purchasers should use a TLT approach as much as is possible. This may well require funders specifying a few high level outcomes and working with providers to cascade those into more specific and/or clinically-determined contributory results (e.g. a multi-year contract for a diabetic population in which money is clawed back from the provider if agreed thresholds of hospital admissions and/or patients going into renal failure are exceeded). Transaction-linked payments will still be useful for high utility interventions, such as cataract surgery in an otherwise independent older person and a hip joint replacement in a working age person with significant hip disease. We also recommend the following. Because healthcare is complex and diverse, there is no universally successful purchasing method and the impact of purchasing will be modified by non-financial incentives and instruments. Purchasers need to use a process that can evolve and is blended (i.e. horses for courses), and that is behaviourally economically sound.22-25 For example, although regulation and compliance audits can ensure minimum standards of care provision, they are not as effective in improving quality as is the impact of informed consumers and informed referrers on a provider marketplace. This requires: alternative providers; the publication of useful outcome data that informs both practice and purchase and, that funding follows the choices that consumers and referrers make. Population-based funding or purchases by way of a capitation are likely to be more effective if they have a TLT basis. In addition to required minimum performance metrics (e.g. around access to primary care), purchasers should prefer healthcare plans that aim to deliver well-defined improvements in outcomes for specific population segments (with some of the payment conditional on success). Inevitably, that requires some judgement by funders on the likelihood of the plan succeeding – based on an assessment of provider capability and capacity (i.e. workforce, IT and facilities) and their proposed operating model. The desired outcomes will be determined by the funder's mandate. Those with a limited health mandate are likely to include familiar outcomes: for example, the patient experience and avoidable hospital admissions as well as disease and injury specific outcomes (e.g. rates of survival from prostate cancer with intact sexual function and urinary continence). However, there is increasing interest both in the contribution of social determinants to better health outcomes and in the contribution that improved health and independence can make to education, employment, welfare and justice outcomes. For those population segments where earlier intervention is most likely to yield substantial lifetime benefits, we are seeing more interest in taking an investment approach across the social sector. This investment will be conditional on delivering a return over time and across a broader range of social outcomes (e.g. improved educational performance, more employment and reduced welfare dependence, reduced offending and re-offending rates and so on). The New Zealand Accident Rehabilitation and Compensation Corporation successfully employs an investment approach to healthcare where the holistic lifetime cost of a problem is known.41 This enables the sizing of investments by likely return to early intervention: especially where investment in injury prevention and rehabilitation reduces the future actuarial cost of treatment, support and compensation for lost earnings. There is benefit in purchasers working with well functioning provider alliances not only in cascading high-level outcomes into deliverable milestones, but also in agreeing to effective contracting arrangements, such as: rolling contracts (to stimulate alliance performance and ensure enough revenue certainty for them to invest); pairing targets (to ensure real productivity gains); risk sharing with other alliances or providers to encourage cooperation where the latter adds value and ensuring that funding aimed at high-needs individuals and families are tagged to them. Sharing the risk with and between providers generates an incentive for them to produce a good plan and then to deliver on it. Alliance contracting is effective,42 but this has to be a genuine structural and operational alliance, rather than a loose coalition of purchasers and providers. In summary, healthcare is increasingly unaffordable.1 In our opinion, sophisticated purchasing can help address this situation by resulting in better, innovative and integrated services.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.