from The Academic Health Economists’ Blo… at http://bit.ly/2IAH5IE on March 9, 2020 at 12:05PM
Every Monday our authors provide a round-up of some of the most recently published peer reviewed articles from the field. We don’t cover everything, or even what’s most important – just a few papers that have interested the author. Visit our Resources page for links to more journals or follow the HealthEconBot. If you’d like to write one of our weekly journal round-ups, get in touch.
There’s comfort and safety in equal treatment. The idea that a ‘QALY is a QALY is a QALY’ will probably always be important as a baseline in discussions about efficiency and equity in health care. This paper reports on a major study of some of the most important ways in which we might like to consider deviating from the idea that all QALYs are made equal.
The overall purpose of the study is to describe a new method that can be used to identify weights that could be attached to particular ‘types’ of QALYs. Underlying the approach is the notion that QALYs can be generated in different ways. They might be generated solely (or mostly) by improvements in quality of life (QoL). Alternatively, they might be generated by extensions of life. Thus, there are life-extending QALYs, QoL-improving QALYs, QALYs generated as a mix of life extension and QoL improvement, and QALYs that involve trading life years for QoL-improvement (or vice versa). Society might feel differently about the value of these QALYs, and the resulting weights could be used in resource allocation decision-making.
This is an impressive study. The work is grounded in a review of the literature, three phases of qualitative research, and survey piloting. This knowledge base is used to design a discrete choice experiment (DCE) in which a representative sample of the Australian public (n=1000) were asked about treatments resulting in the different types of QALY gains described above. In addition to attributes relating to QoL gain and life extension, the DCE also included attributes for the age of people who would receive treatment, QoL without treatment, life expectancy without treatment, and the absolute number of QALYs gained. All of this was plugged into a variety of choice models to estimate the relative weights.
The findings are not necessarily what you would expect. In general, people preferred QALYs generated as a mixture of QoL improvement and life extension. The influence of severity (i.e. baseline QoL) is complicated, with people in severe states least favoured and people with a moderate health state most favoured. Similarly, there doesn’t seem to be an overall preference for end of life treatment. With respect to age, gains for younger people were preferred, especially when life expectancy was short. On average, the value of gains for infants and adults was around twice that for the oldest people.
There’s so much in this paper that I can’t do it justice. If you’re interested in DCEs or social preferences, there’s very likely some important methodological features that you’ll want to explore in this paper. A potentially important overall finding is that social preferences – on average – neither maximise QALYs nor prioritise the worst off. Rather, they seem to aim for the achievement of some moderate threshold.
But there’s one thing that worries me. Considering the matter of efficiency and equity, it isn’t clear in which camp this work sits. I suspect the authors would put it in the former, as the study elicits preferences. But, if that’s the case, we surely have a problem. The ‘types’ of QALYs being traded in this exercise were already generated on the basis of trade-offs. Individuals have already been asked to trade-off life years and QoL according to their own preferences, and these trade-offs have been lumped together to obtain a social set of values to estimate QALYs. What does it mean to give society a second chance to trade-off types of QALYs? I’m not sure.
Estimating social variation in the health effects of changes in health care expenditure. Medical Decision Making [PubMed] Published 15th February 2020
The work by Claxton et al that came up with the infamous £13,000 per QALY threshold (subsequently reframed as ‘marginal productivity’) has facilitated a range of further studies. Here is one of the more interesting examples. If we accept (just for a moment) the idea that £13,000 represents the average cost of a QALY in the NHS, then a logical next step is to consider what that average is made up of, and the sources of variation. This study considers the social gradient in the marginal productivity of health spending.
The research is built on the foundations laid by the Claxton et al threshold work, but builds in ‘equity-relevant’ variables, namely age, sex, and deprivation. To achieve this, the authors estimated the health benefits accruing to age-sex groups by disaggregating those reported by Claxton et al according to incidence estimates, and then linking this to utilisation levels in hospital episode statistics to approximate expenditure.
The headline result is that more QALYs are generated by spending in deprived areas than spending in well-off areas; nearly twice as many in the most deprived quintile as in the least deprived quintile. The effect seems to be more pronounced for men than for women and for younger than for older. Across disease areas, there is greater inequality in marginal productivity of expenditure in mental health, while spending on cancer seems to be more productive in the least deprived areas.
These findings are potentially useful in distributional cost-effectiveness analysis and could in principle inform budgeting decisions. But the policy implications aren’t simple. One thing that the authors don’t discuss in great detail is the multi-sectoral considerations. All else equal, it makes sense to spend more on health care in less well-off areas. But what is the opportunity cost of health expenditure within an area if all else is not equal? What if the relative productivity of spending on education and social welfare is even greater than that on health care? In that case, it might make sense to actually spend less on health care in less well-off areas and more on other programmes.
Reproducible research practices, openness and transparency in health economic evaluations: study protocol for a cross-sectional comparative analysis. BMJ Open [PubMed] Published 13th February 2020
I’m all for more transparent research practices. Who isn’t, right? So I was drawn to this protocol for a study of how transparent cost-effectiveness studies have been over the years.
The researchers plan to review 600 economic evaluations from 2012, 2019, and 2022. They’ll extract information about each study, with a focus on ‘transparency, openness and reproducibility’. The authors don’t define exactly what they mean by these things, which could be defined in a variety of ways, but state that they will extract information relating to matters such as whether the study was registered, whether data are available, and the quality of the reporting.
The authors are overly ambitious. Not only is 600 studies a lot from which to extract more than 30 pieces of information (including the corresponding author’s gender!), but the authors also claim that they will continue to surveil the literature into the future. They also believe that they can, to some extent, evaluate the impact of the publication of the CHEERS statement in 2013. I would be amazed if all of this could be achieved, so hopefully the core contributions of the study won’t fall by the wayside. The beauty of open research is that we can keep track!