Archive | January, 2017

The ‘triple aim’: a reality check

27 Jan

by Jonathan Stokes, Research Associate in the Centre for Health Economics at The University of Manchester


In December, I completed the final exam (or ‘viva’) for my PhD, which was funded by the NIHR Greater Manchester PSTRC. In the discussion section of my thesis, I presented an argument that my examiners suggested I should explore further, so I’m using this blog post to try and record some initial thoughts, which I hope to turn into an academic paper at some point (so feel free to post any feedback you have below).

So, my argument, briefly, was this:

  • Health systems are recognised to have a so-called ‘triple aim’, aiming to improve health and satisfaction of patients, while reducing costs. With the move towards ‘integrated care’, policymakers argue that integration can improve all triple-aim outcomes simultaneously.
  • In the NHS, integrated care is defined around the patient’s feeling of joined-up care (i.e. around the outcome of satisfaction). However, policymakers, when setting out the aims of their integrated care models aim primarily to reduce costs, usually by reducing emergency hospital admissions (probably unsurprising given the ‘NHS crisis’ currently being covered in the media).
  • But the literature tells us that the ‘triple aim’ goals are quite complexly interlinked, so achieving one does not necessarily have a beneficial effect on all of the others. For example, the highest patient satisfaction (good) is associated with the highest overall healthcare expenditure (bad), and may be also those experiencing worse health outcomes, e.g. a higher mortality rate (very bad).
  • Therefore, there is a potential conflict between the patient-level (the feeling of more joined up care for the patient) and system-level (reducing costs) goals of integrated care in the NHS, and potentially a need to prioritise aims.

Basically, there is little evidence that the ideal scenario of not having to pay for improvements in care quality is plausible (note: that doesn’t mean we can’t have this improvement, it just implies that we would need to invest more in our health system to get it). And, if we design interventions that address the NHS definition of integration and increase satisfaction with care, then we won’t necessarily meet the pressing system aims of reducing costs in the current NHS funding crisis.

So, to my mind, if we want to contain costs, this would mean that the patient satisfaction aim (while still important, and should continue to be measured) should be bumped down the priority. This is partly because people who are healthier tend to be more satisfied in any case (so it is partially a secondary outcome of better health and should take care of itself if people get better health outcomes), but also because in a tax-funded system like the NHS, where the overarching values are delivering equity and social justice, sustainability of the system is more important than preferences of the individual (i.e. ‘consumer-focused’ healthcare). Especially when aiming primarily for satisfaction of the individual harms the system.

To try and explain why this might be, imagine that if we’re primarily aiming to improve satisfaction with care, implementing an intervention where we gave everyone a dram of whisky as they waited for their appointment might send satisfaction rates through the roof, but it would cost a fortune and might not do people’s health a lot of good.

Healthcare is a complicated process (which is why it takes so long to train healthcare professionals, who in fact never stop receiving training during their working lives). One of the ‘market failures’ in healthcare (i.e. why a normal economic free-market system, where the consumer decides, doesn’t work) is ‘information asymmetry’ (i.e. the expert knows more than others, so can make the more informed decisions). While this asymmetry is clearly in patients’ favour in some aspects (e.g. experience of disease/ experience of using health services/ treatment burden etc.), it favours the physician/expert in others (e.g. treatment course in many cases/ health system setup or organisation/ cost of care etc.). The patient is therefore not likely to make the most informed decision when it comes to overall health system costs.

In view of this, I think there’s an interesting debate to be had about the role of ‘patient-centred’ care more widely, and how this fits with our system goals. For example, I went to a seminar the other day where an NHS Vanguard programme talked about how they were shifting their approach in patient care to one where they ask the patients to define their goals, asking them ‘What matters to you?’. But, before that, we’d had a talk from NHS England about the national metrics that the Vanguards would be assessed on, centred on system goals (again, primarily about reducing avoidable emergency hospital admissions). This got me thinking: what if the patient’s goal is something that doesn’t affect our system goals whatsoever, or is even completely contrary to these? If it was in the USA, the patients themselves (or their health insurer) would pay for the option and it wouldn’t matter so much, but in a tax-funded system there is more need to think resourcefully at this wider system level (and try being poor and getting decent healthcare in the USA if you think that’s the way to go instead).

The role of the patient in influencing/defining research priorities could be another debate. For example, the PSTRC and wider NIHR funders are very focused on Public and Patient Involvement (PPI) in research. But, what if the researcher (while employing common sense, and not purely as a stereotypical ivory-tower-type) has access to more knowledge about where the current knowledge gaps are, and which questions are possible to answer and how? Is it a good use of resources to pay for the ‘lay perspective’ in all cases (or at all)? (But, of course, this is just one simplified question in the debate of the role of PPI in influencing and defining research priorities, the balance of which I hope to explore further in a future post).

In sum, I think some of the assumptions we make (in good faith, I’m sure), about the aims and delivery of health services, and research to support these services, are still open to debate. While we dwell on ideally achieving all outcomes, we neglect the reality of the need for trade-offs and priority setting. But, I’d be interested in hearing your thoughts.