A few weeks ago* I headed to Rome for the EuroQol plenary meeting. I arrived at the hotel around 06:30 a.m. off the back of a 36-hour train adventure and soon got to the usual routine of catching up with old friends and discussing all things EQ-5D (et al.). Plenary meetings include numerous strategic discussions and business matters for EuroQol members to tackle. I’ll spare you those and focus on the scientific programme that ran over two days on the 20th and 21st of September.
EuroQol plenary meetings employ an HESG-style discussant model with full papers shared in advance, but with shorter sessions for discussion and – for the most part – a single stream of sessions. Many papers are authored by EuroQol members, though there are plenty of guests and newcomers. There’s also a growing number of EuroQol PhD students who tend to be well-represented. This meeting was chaired by Juanita Haagsma and Nils Gutacker, with expert input on local matters from proud Roman Aureliano Finch. A game of EuroQol bingo lasted about 10 minutes as the speakers and audience members blasted through all of the usual clichés.
The first day started with a collection of papers on the valuation of EuroQol instruments, opening with an unusually hostile discussion. Paul Kind and Ling-Hsiang Chuang‘s paper challenged EuroQol’s EQ-VT valuation protocol. This was on the grounds that time trade-off (TTO) and discrete choice experiment (DCE) data are not commensurable, yet the EQ-VT demands their combination. The discussion paper arose from an attempt to collate information from valuation studies worldwide. In some ways, this kind of paper shows the EuroQol plenary at its best, encouraging debates (occasionally existential, often rambling) about the fundamentals of using the EQ-5D to value health states and support resource allocation. On this occasion, the paper was discussed by Benjamin Craig, who took it as an opportunity to grind an axe about discounting (the paper makes no reference to discounting).
In another paper that combined data from different valuation studies, Edward Webb showed that EQ-5D value sets have changed over time, particularly in terms of the relative value of different domains of health. This suggests that value sets should perhaps have a shelf life. A paper in the first session also launched an ongoing discussion about the Online Elicitation of Personal Utility Functions: the OPUF tool. Throughout the meeting, participants (sometimes with tongue in cheek) presented ‘OPUF’ as the answer to every question. This persisted until the last word of the scientific programme, with Bram Roudijk‘s mic-drop statement: “OPUF is not a preference-based valuation method.” Boom!
Another hot topic at the meeting, as it has been at other conferences, was decolonisation. Lucky Ngwira‘s paper presented some novel qualitative research with children and adolescents in Malawi, exploring how they conceptualise health. The study gives the impression that the kids taking part were super smart (an impression that I often get from qualitative research with children), and it raised some notions that we Europeans might not have considered important or even relevant to health, such as feelings of accepting illness and of health being a god-given gift. Elsewhere in the programme, we had a couple of posters on Māori and other indigenous people’s perspectives on health. There are important questions to be asked about the hegemony of the EQ-5D and the scope and reach of the EuroQol Group’s work.
Research on EuroQol instruments for children comprised a large part of the programme. There was an entire parallel session that I did not attend, which included a paper from some OHE colleagues using the OPUF to explore the role of different perspectives in the valuation of the EQ-5D-Y. We’re on the brink of the EQ-5D-Y-5L going mainstream, and the EQ-TIPS instrument is heading-up a new horizon for valuing infants’ health. A particularly interesting discussion paper led by Nancy Devlin explored how we might approach the challenging task of valuing babies’ health states.
A long-standing recurring theme at EuroQol plenaries is research on the possible approaches to (and arguments about whether we should be) extending the scope of EuroQol instruments. Bolt-ons are a long-standing favourite (/bugbear), and this meeting included a paper exploring the impact of 9 possible bolt-ons. As is almost always the case, the study demonstrated the potential value of having these bolt-ons to understand more about respondents’ health states. There was also a poster (that I discussed) reporting on a study that assessed 20 bolt-ons. That’s too many bolt-ons! A more innovative idea presented at the meeting was to ‘bolt on’ an entire pre-existing instrument. Brendan Mulhern and colleagues outlined a possible framework for doing this with an application to the Adult Social Care Outcomes Toolkit (ASCOT). Personally, I don’t like the idea, but I do appreciate such radical lines of thought.
When working on health outcomes research, we sometimes ask ourselves, “Where is the economics?” At this meeting, it was in Michał Jakubczyk‘s paper (drafted in LaTeX) on the maths and moral philosophy of averaging. The proposal is simple; to support interpersonal comparisons, we should consider the geometric mean of utility values instead of the arithmetic mean. And the idea is old; Michał borrows and builds upon the idea as presented in a QJE paper from 1971. I’m convinced by the arguments but don’t yet completely understand them. Nevertheless, I suspect that, in time, once we’ve all got our heads around it, the geometric mean will routinely be adopted in the aggregation of EQ-5D health state values.
There was plenty of other quality stuff on the programme, including a collection of papers on using PROMS and dashboards in clinical settings, but I didn’t have the brain power to effectively engage with everything. One off-programme benefit of attending the EuroQol plenary meetings is the corridor chat, which allows you to find out what’s on the horizon or going on behind the scenes in the world of EuroQol. At this meeting, there was much discussion of the UK EQ-5D-5L value set study, which has apparently completed data collection.
My lowlight of the meeting was being eaten alive by insects, no doubt employed by the purveyors of the ‘itching’ bolt-on. I also had my first experience in living memory of not eating a plate of food served to me; an inedibly (and unfathomably) burnt pasta dish left the vegans hungry on the big night out.
This was an enjoyable and valuable EuroQol plenary, and the first I’ve been able to properly attend as a EuroQol member (I got COVID on the first day of the last one). I learned a lot and was able to meet plenty of new people. Sometimes, EuroQol meetings can seem like a closed shop, but they don’t mean to. So prepare your research for submission to the next plenary meeting in Noordwijk in September 2024.
* Yes, it took me a while to publish this. That’s because I got sick after the meeting. Post-conference sickness seems to be an increasing problem (for me, at least). I’d like to see a study on it.