Chris Sampson’s journal round-up for 4th May 2020

from The Academic Health Economists’ Blo… at on May 4, 2020 at 12:11PM

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.

Learnings for health economics from the early stages of the COVID-19 pandemic. PharmacoEconomics – Open [PubMed] [RePEc] Published 10th April 2020

I’ve included several editorials and brief commentaries on COVID-19 in these round-ups and, last week, Ewan included some relevant older papers. There isn’t much real research to write about yet. This article considers what might be the implications of COVID-19 for our discipline.

The author starts with a pat on the back for health economics, suggesting that we do well with respect to transparency and sensitivity analyses. Relative to what? The epidemiological models published on COVID-19 to date. These have (we are told) been based largely on point estimates and with little explanation of the model inputs and data structures. I haven’t been following developments closely enough to know whether this has changed since the writing of this article.

Potential lessons for health economics relate to the communication of uncertainty and to our influence on policy and the media. Clearly, this is of huge importance when the impact of communication could manifest in changes in people’s behaviour, which could influence health outcomes for them and for the rest of humanity. The author cites the example of the Intergovernmental Panel on Climate Change (IPCC), which uses standardised evidence-based language for the description of probabilities. A similar approach has been adopted for the context of COVID-19. When it comes to engagement with the media and policymakers, we should be supporting public understanding of opportunity cost. That’s because we are the discipline most familiar with the challenge of trading off health and economic impacts, which are the basis for ‘lockdown’ policy responses. There are also questions about our understanding of the ‘rule of rescue’ in the public’s preferences for the allocation of resources. In view of the importance of pre-print publications in crisis times, the author also suggests that the current situation presents an opportunity for us to “get our own house in order” with respect to efficient publication and peer review practices.

It’s easy to agree with all of this. If you’re looking for COVID-related opportunities for our field, this is a reasonable place to start.

Using QALYs versus DALYs to measure cost-effectiveness: how much does it matter? International Journal of Technology Assessment in Health Care [PubMed] Published 28th April 2020

I’ve never really understood DALYs, which is probably because I haven’t done any work in low resource settings. Part of their definition is that they are upside down versions of QALYs, with the best health state being at 0 and dead being at 1. If that was all, they would be directly comparable with QALYs. But the value of different health states is also identified differently, using (as far as my ignorance goes) arbitrary weights. So we might expect cost-effectiveness analyses using DALYs to give different results to those using QALYs. But do they? The authors of this study conducted a review to find out.

The CEA Registry includes cost-per-QALY studies and the Global Health CEA Registry includes cost-per-DALY studies. Essentially, this review identified the overlap in the two, looking at cost-effectiveness studies that used both QALYs and DALYs within a single study. In total, only 11 studies were included, which is the main limitation of the review. There’s not much we can tell from 11 studies. And four of these didn’t report cost-effectiveness ratios because the interventions were cost-saving. Most were conducted in the context of infectious diseases and around half were conducted in high-income settings. Eleven cost-effectiveness ratios were reported across the studies, which were roughly evenly split between being higher for QALYs or DALYs.

The authors summarise some of the literature that explores (in theory and in practice) reasons for differences in QALY and DALY estimates. Knowing very little about DALYs, I found this useful. But the authors aren’t able to add much to this literature or to support or contradict previous findings. Thus, the authors conclude on the fence, suggesting that it probably doesn’t make much difference whether you choose QALYs or DALYs, except it might. For instance, one study showed outcomes almost twice as great in terms of QALYs, such that the implied decision alters according to the choice of outcome.

This study will probably be conducted again in another five years or so, by which time there may be enough data points to say something practically useful. But even then, a very big question mark remains over the usefulness of the research. If QALYs and DALYs are measuring different constructs (which, surely, they are, or else DALYs wouldn’t need to exist) and are therefore not interchangeable, then what is the point in comparing them in this way?

International valuation protocol for the EQ-5D-Y-3L. PharmacoEconomics [PubMed] [RePEc] Published 16th April 2020

Finally! The EQ-5D-Y descriptive system has been at the disposal of our discipline for as long as I have. But, to date, it has been of limited practical use. That’s because there has been no value set. This paper summarises a suite of research studies that were funded by EuroQol with the intention of figuring out how best to go about developing value sets for the EQ-5D-Y, and the resulting protocol.

Numerous research questions were identified, but principal among them were i) whether we can get away with using adult EQ-5D value sets, ii) whether we should refer to adults or children as the arbiters of value, and iii) how can we anchor on the full health to dead scale when people are averse to trading away life years for children? Four studies are briefly described, involving various formulations of discrete choice experiment (DCE) and time trade-off (TTO) exercises. Each of these studies has been (or will be) reported separately, so this paper is more about the resulting protocol.

Inevitably, the research showed that we cannot get away with using adult EQ-5D value sets, which is what most people have been doing in lieu of an EQ-5D-Y value set. The wording of the instruments is important. DCEs seem to be a feasible and appropriate method by which to elicit the relative value of different health states described by the EQ-5D-Y, while TTO data are problematic because the technique cannot handle the prospect of a dead child. But with DCE data only, values are on an unknown scale, not related to being dead or in full health, and so QALYs can’t be estimated. Therefore, a second stage is required, for which TTO seems to be OK.

The resulting protocol is essentially an adapted version of the current (adult) EQ-5D-5L valuation protocol. As with the adult protocol, values are supposed to be obtained from a representative sample of the (adult) general public. Valuation tasks are framed such that participants should consider the health of a 10-year old child. Like the 5L protocol, both (online) DCE and (face to face) composite TTO tasks should be completed. However, the composite TTO in the EQ-5D-Y valuation protocol has the specific purpose of anchoring on the full health to dead scale.

It may have taken ten years, but the work described is testament to EuroQol’s thorough approach. Questions will remain, and methods will continue to be refined, but we’re now at the point where researchers can go forth and create EQ-5D-Y value sets, as well as, no doubt, raising more research questions along the way.

Keeping pace with pharmaceutical innovation: the importance of the NICE methods review. PharmacoEconomics [PubMed] [RePEc] Published 24th April 2020

For many people, the NICE methods review is an opportunity. In particular, it’s an opportunity for researchers to get their favourite methodology recognised by the methodology gods, and all of the citations that go along with that. It’s an opportunity for patient groups and charities to influence minor changes in methodology that could lead to big changes in access to medicines for their members. And, in some respects, it’s an opportunity for the pharmaceutical industry, who would like to see more patients get access to more medicines at a faster pace. This paper (written, I should say, by some people that I work with quite a bit) provides a perspective from industry.

We’re all guilty, at times, of special pleading, but I don’t have time for it. So I was relieved to see none in this article. Rather, the authors highlight several matters that have been debated in the literature for quite some time and on which – the authors argue – the NICE methods review needs to provide a firmer stance. The main three matters raised are i) a broader value framework, ii) a clearer characterisation of uncertainty, and iii) differential discount rates. Most people would agree that NICE should not adopt a rigid cost-per-QALY approach and, indeed, it doesn’t. But the authors would like to see greater transparency on matters of QALY weighting. In particular, they raise the possibility of a severity-based modifier in place of binary criteria such as the end of life ‘premium’. On discounting, the authors would like to see alignment with the Treasury Green Book. Any number of positions are tenable when it comes to discounting, but there is at least some logic in consistent policymaking. The most interesting discussion is around uncertainty and the notion that not all uncertainty should be treated equally. In some contexts (e.g. rare diseases) there is an inherent uncertainty that is qualitatively (and, perhaps, normatively) different to uncertainty that exists due to a lack of research. Patients shouldn’t lose out simply by virtue of research being more difficult in their context.

As you might expect, this article discusses the innovative nature of new technologies and the importance of achieving high levels of access to medicines in the UK compared to other countries. But you needn’t share those priorities to recognise the worth of the arguments in this article. NICE has a tough job on its hands with the methods review.