Page 2 of 129

Chris Sampson’s journal round-up for 1st April 2019

Chris Sampson 1 Comment

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.

Toward a centralized, systematic approach to the identification, appraisal, and use of health state utility values for reimbursement decision making: introducing the Health Utility Book (HUB). Medical Decision Making [PubMed] Published 22nd March 2019

Every data point reported in research should be readily available to us all in a structured knowledge base. Most of us waste most of our time retreading old ground, meaning that we don’t have the time to do the best research possible. One instance of this is in the identification of health state utility values to plug into decision models. Everyone who builds a model in a particular context goes searching for utility values – there is no central source. The authors of this paper are hoping to put an end to that.

The paper starts with an introduction to the importance of health state utility values in cost-effectiveness analysis, which most of us don’t need to read. Of course, the choice of utility values in a model is very important and can dramatically alter estimates of cost-effectiveness. The authors also discuss issues around the identification of utility values and the assessment of their quality and applicability. Then we get into the objectives of the ‘Health Utility Book’, which is designed to tackle these issues.

The Health Utility Book will consist of a registry (I like registries), backed by a systematic approach to the identification and inclusion (registration?) of utility values. The authors plan to develop a quality assessment tool for studies that report utility values, using a Delphi panel method to identify appropriate indicators of quality to be included. The quality assessment tool will be complemented by a tool to assess applicability, which will be developed through interviews with stakeholders involved in the reimbursement process.

In the first place, the Health Utility Book will only compile utility values for cancer, and some of the funding for the project is cancer specific. To survive, the project will need more money from more sources. To be sustainable, the project will need to attract funding indefinitely. Or perhaps it could morph into a crowd-sourced platform. Either way, the Health Utility Book has my support.

A review of attitudes towards the reuse of health data among people in the European Union: the primacy of purpose and the common good. Health Policy Published 21st March 2019

We all agree that data protection is important. We all love the GDPR. Organisations such as the European Council and the OECD are committed to facilitating the availability of health data as a means of improving population health. And yet, there often seem to be barriers to accessing health data, and we occasionally hear stories of patients opposing data sharing (e.g. care.data). Maybe people don’t want researchers to be using their data, and we just need to respect that. Or, more likely, we need to figure out what it is that people are opposed to, and design systems that recognise this.

This study reviews research on attitudes towards the sharing of health data for purposes other than treatment, among people living in the EU, employing a ‘configurative literature synthesis’ (a new one for me). From 5,691 abstracts, 29 studies were included. Most related to the use of health data in research in general, while some focused on registries. A few studies looked at other uses, such as for planning and policy purposes. And most were from the UK.

An overarching theme was a low awareness among the population about the reuse of health data. However, in some studies, a desire to be better informed was observed. In general, views towards the use of health data were positive. But this was conditional on the data being used to serve the common good. This includes such purposes as achieving a better understanding of diseases, improving treatments, or achieving more efficient health care. Participants weren’t so happy with health data reuse if it was seen to conflict with the interests of patients providing the data. Commercialisation is a big concern, including the sale of data and private companies profiting from the data. Employers and insurance companies were also considered a threat to patients’ interests. There were conflicting views about whether it is positive for pharmaceutical companies to have access to health data. A minority of people were against sharing data altogether. Certain types of data are seen as being particularly sensitive, including those relating to mental health or sexual health. In general, people expressed concern about data security and the potential for leaks. The studies also looked at the basis for consent that people would prefer. A majority accepted that their data could be used without consent so long as the data were anonymised. But there were no clear tendencies of preference for the various consent models.

It’s important to remember that – on the whole – patients want their data to be used to further the common good. But support can go awry if the data are used to generate profits for private firms or used in a way that might be perceived to negatively affect patients.

Health-related quality of life in injury patients: the added value of extending the EQ-5D-3L with a cognitive dimension. Quality of Life Research [PubMed] Published 18th March 2019

I’m currently working on a project to develop a cognition ‘bolt-on’ for the EQ-5D. Previous research has demonstrated that a cognition bolt-on could provide additional information to distinguish meaningful differences between health states, and that cognition might be a more important candidate than other bolt-ons. Injury – especially traumatic brain injury – can be associated with cognitive impairments. This study explores the value of a cognition bolt-on in this context.

The authors sought to find out whether cognition is sufficiently independent of other dimensions, whether the impact of cognitive problems is reflected in the EuroQol visual analogue scale (EQ VAS), and how a cognition bolt-on affects the overall explanatory power of the EQ-5D-3L. The data used are from the Dutch Injury Surveillance System, which surveys people who have attended an emergency department with an injury, including EQ-5D-3L. The survey adds a cognitive bolt-on relating to memory and concentration.

Data were available for 16,624 people at baseline, with 5,346 complete responses at 2.5-month follow-up. The cognition item was the least affected, with around 20% reporting any problems (though it’s worth noting that the majority of the cohort had injuries to parts of the body other than the head). The frequency of different responses suggests that cognition is dominant over other dimensions in the sense that severe cognitive problems tend to be observed alongside problems in other dimensions, but not vice versa. The mean EQ VAS for people reporting severe cognitive impairment was 41, compared with a mean of 75 for those reporting no problems. Regression analysis showed that moderate and severe cognitive impairment explained 8.7% and 6.2% of the variance of the EQ VAS. Multivariate analysis suggested that the cognitive dimension added roughly the same explanatory power as any other dimension. This was across the whole sample. Interestingly (or, perhaps, worryingly) when the authors looked at the subset of people with traumatic brain injury, the explanatory power of the cognitive dimension was slightly lower than overall.

There’s enough in this paper to justify further research into the advantages and disadvantages of using a cognition bolt-on. But I would say that. Whether or not the bolt-on descriptors used in this study are meaningful to patients remains an open question.

Developing the role of electronic health records in economic evaluation. The European Journal of Health Economics [PubMed] Published 14th March 2019

One way that we can use patients’ routinely collected data is to support the conduct of economic evaluations. In this commentary, the authors set out some of the ways to make the most of these data and discuss some of the methodological challenges. Large datasets have the advantage of being large. When this is combined with the collection of sociodemographic data, estimates for sub-groups can be produced. The data can also facilitate the capture of outcomes not otherwise available. For example, the impact of bariatric surgery on depression outcomes could be identified beyond the timeframe of a trial. The datasets also have the advantage of being representative, where trials are not. This could mean more accurate estimates of costs and outcomes. But there are things to bear in mind when using the data, such as the fact that coding might not always be very accurate, and coding practices could vary between observations. Missing data are likely to be missing for a reason (i.e. not at random), which creates challenges for the analyst. I had hoped that this paper would discuss novel uses of routinely collected data systems, such as the embedding of economic evaluations within them, rather than simply their use to estimate parameters for a model. But if you’re just getting started with using routine data, I suppose you could do worse than start with this paper.

Credits

Chris Sampson’s journal round-up for 25th March 2019

Chris Sampson Leave a comment

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.

How prevalent are implausible EQ-5D-5L health states and how do they affect valuation? A study combining quantitative and qualitative evidence. Value in Health Published 15th March 2019

The EQ-5D-5L is able to describe a lot of different health states (3,125, to be precise), including some that don’t seem likely to ever be observed. For example, it’s difficult to conceive of somebody having extreme problems in pain/discomfort and anxiety/depression while also having no problems with usual activities. Valuation studies exclude these kinds of states because it’s thought that their inclusion could negatively affect the quality of the data. But there isn’t much evidence to help us understand how ‘implausibility’ might affect valuations, or which health states are seen as implausible.

This study is based on an EQ-5D-5L valuation exercise with 890 students in China. The valuation was conducted using the EQ VAS, rather than the standard EuroQol valuation protocol, with up to 197 states being valued by each student. Two weeks after conducting the valuation, participants were asked to indicate (yes or no) whether or not the states were implausible. After that, a small group were invited to participate in a focus group or interview.

No health state was unanimously identified as implausible. Only four states were unanimously rated as not being implausible. 910 of the 3,125 states defined by the EQ-5D-5L were rated implausible by at least half of the people who rated them. States more commonly rated as implausible were of moderate severity overall, but with divergent severities between states (i.e. 5s and 1s together). Overall, implausibility was associated with lower valuations.

Four broad themes arose from the qualitative work, namely i) reasons for implausibility, ii) difficulties in valuing implausible states, iii) strategies for valuing implausible states, and iv) values of implausible states. Some states were considered to have logical conflicts, with some dimensions being seen as mutually inclusive (e.g. walking around is a usual activity). The authors outline the themes and sub-themes, which are a valuable contribution to our understanding of what people think when they complete a valuation study.

This study makes plain the fact that there is a lot of heterogeneity in perceptions of implausibility. But the paper doesn’t fully address the issue of what plausibility actually means. The authors describe it as subjective. I’m not sure about that. For me, it’s an empirical question. If states are observed in practice, they are plausible. We need meaningful valuations of states that are observed, so perhaps the probability of a state being included in a valuation exercise should correspond to the probability of it being observed in reality. The difficulty of valuing a state may relate to plausibility – as this work shows – but that difficulty is a separate issue. Future research on implausible health states should be aligned with research on respondents’ experience of health states. Individuals’ judgments about the plausibility of health states (and the accuracy of those judgments) will depend on individuals’ experience.

An EU-wide approach to HTA: an irrelevant development or an opportunity not to be missed? The European Journal of Health Economics [PubMed] Published 14th March 2019

The use of health technology assessment is now widespread across the EU. The European Commission recently saw an opportunity to rationalise disparate processes and proposed new regulation for cooperation in HTA across EU countries. In particular, the proposal targets cooperation in the assessment of the relative effectiveness of pharmaceuticals and medical devices. A key purpose is to reduce duplication of efforts, but it should also make the basis for national decision-making more consistent.

The authors of this editorial argue that the regulation needs to provide more clarity, in the definition of clinical value, and of the quality of evidence that is acceptable, which vary across EU Member States. There is also a need for the EU to support early dialogue and scientific advice. There is also scope to support the generation and use of real-world evidence. The authors also argue that the challenges for medical device assessment are particularly difficult because many medical device companies cannot – or are not incentivised to – generate sufficient evidence for assessment.

As the final paragraph argues, EU cooperation in HTA isn’t likely to be associated with much in the way of savings. This is because appraisals will still need to be conducted in each country, as well as an assessment of country-specific epidemiology and other features of the population. The main value of cooperation could be in establishing a stronger position for the EU in negotiating in matters of drug design and evidence requirements. Not that we needed any more reasons to stop Brexit.

Patient-centered item selection for a new preference-based generic health status instrument: CS-Base. Value in Health Published 14th March 2019

I do not believe that we need a new generic measure of health. This paper was always going to have a hard time convincing me otherwise…

The premise for this work is that generic preference-based measures of health (such as the EQ-5D) were not developed with patients. True. So the authors set out to create one that is. A key feature of this study is the adoption of a framework that aligns with the multiattribute preference response model, whereby respondents rate their own health state relative to another. This is run through a mobile phone app.

The authors start by extracting candidate items from existing health frameworks and generic measures (which doesn’t seem to be a particularly patient-centred approach) and some domains were excluded for reasons that are not at all clear. 47 domains were included after overlapping candidates were removed. The 47 were classified as physical, mental, social, or ‘meta’. An online survey was conducted by a market research company. 2,256 ‘patients’ (people with diseases or serious complaints) were asked which 9 domains they thought were most important. Why 9? Because the authors figured it was the maximum that could fit on the screen of a mobile phone.

Of the candidate items, 5 were regularly selected in the survey: pain, personal relationships, fatigue, memory, and vision. Mobility and daily activities were also judged important enough to be included. Independence and self-esteem were added as paired domains and hearing was paired with the vision domain. The authors also added anxiety/depression as a pair of domains because they thought it was important. Thus, 12 items were included altogether, of which 6 were parts of pairs. Items were rephrased according to the researchers’ preferences. Each item was given 4 response levels.

It is true to say (as the authors do) that most generic preference-based measures (most notably the EQ-5D) were not developed with direct patient input. The argument goes that this somehow undermines the measure. But there are a) plenty of patient-centred measures for which preference-based values could be created and b) plenty of ways in which existing measures can be made patient-centred post hoc (n.b. our bolt-on study).

Setting aside my scepticism about the need for a new measure, I have a lot of problems with this study and with the resulting CS-Base instrument. The defining feature of its development seems to be arbitrariness. The underlying framework (as far as it is defined) does not seem well-grounded. The selection of items was largely driven by researchers. The wording was entirely driven by the researchers. The measure cannot justifiably be called ‘patient-centred’. It is researcher-centred, even if the researchers were able to refer to a survey of patients. And the whole thing has nothing whatsoever to do with preferences. The measure may prove fantastic at capturing health outcomes, but if it does it will be in spite of the methods used for its development, not because of them. Ironically, that would be a good advert for researcher-centred outcome development.

Proximity to death and health care expenditure increase revisited: a 15-year panel analysis of elderly persons. Health Economics Review [PubMed] [RePEc] Published 11th March 2019

It is widely acknowledged that – on average – people incur a large proportion of their lifetime health care costs in the last few years of their life. But there’s still a question mark over whether it is proximity to death that drives costs or age-related morbidity. The two have very different implications – we want people to be living for longer, but we probably don’t want them to be dying for longer. There’s growing evidence that proximity to death is very important, but it isn’t clear how important – if at all – ageing is. It’s important to understand this, particularly in predicting the impacts of demographic changes.

This study uses Swiss health insurance claims data for around 104,000 people over the age of 60 between 1996 and 2011. Two-part regression models were used to estimate health care expenditures conditional on them being greater than zero. The author analysed both birth cohorts and age classes to look at age-associated drivers of health care expenditure.

As expected, health care expenditures increased with age. The models imply that proximity-to-death has grown in importance over time. For the 1931-35 birth cohort, for example, the proportion of expenditures explained by proximity-to-death rose from 19% to 31%. Expenditures were partly explained by morbidity, and this effect appeared to be relatively constant over time. Thus, proximity to death is not the only determinant of rising expenditures (even if it is an important one). Looking at different age classes over time, there was no clear picture in the trajectory of health care expenditures. For the oldest age groups (76-85), health care expenditures were growing, but for some of the younger groups, costs appeared to be decreasing over time. This study paints a complex picture of health care expenditures, calling for complex policy responses. Part of this could be supporting people to commence palliative care earlier, but there is also a need for more efficient management of chronic illness over the long term.

Credits

Thesis Thursday: Alastair Irvine

Chris Sampson Leave a comment

On the third Thursday of every month, we speak to a recent graduate about their thesis and their studies. This month’s guest is Dr Alastair Irvine who has a PhD from the University of Aberdeen. If you would like to suggest a candidate for an upcoming Thesis Thursday, get in touch.

Title
Time preferences and the patient-doctor interaction
Supervisors
Marjon van der Pol, Euan Phimister
Repository link
http://digitool.abdn.ac.uk/webclient/DeliveryManager?pid=238373

How can people’s time preferences affect the way they use health care?

Time preferences are a way of thinking about how people choose between things that happen over time. Some people prefer a treatment with large side effects and a long chain of future benefits; others prefer smaller benefits but less side effects. These influence a wide range of health outcomes and decisions. One of the most interesting questions I had coming into the PhD was around non-adherence.

Non-adherence can’t be captured by ‘standard’ exponential time preferences because there is no way for something you prefer now to be ‘less preferred’ in the future if everything is held constant. Instead, present-bias preferences can capture non-adherent behaviour. With these preferences, people place a higher weight on the ‘current period’ relative to all future periods but weight all future periods consistently. What that means is you can have a situation where you plan to do something – eat healthily, take your medication – but end up not doing it. When planning, you placed less relative weight on the near term ‘cost’ (like medication side effects) than you do when the decision arrives.

In what way might the patient-doctor interaction affect a patient’s adherence to treatment?

There’s asymmetric information between doctors and patient, leading to an agency relationship. Doctors in general know more about treatment options than patients, and don’t know their patient’s preferences. So if doctors are making recommendations to patients, this asymmetry can lead to recommendations that are accepted by the patient but not adhered to. For example, present-biased patients accept the same treatments as exponential discounters. Depending on the treatment parameters, present-biased people will not adhere to some treatments. If the doctor doesn’t anticipate this when making recommendations, it leads to non-adherence.

One of the issues from a contracting perspective is that naive present-bias people don’t anticipate their own non-adherence, so we can’t write traditional ‘separating contracts’ that lead present-bias people to one treatment and exponential discounters to another. However, if the doctor can offer a lower level of treatment to all patients – one that has less side effects and a concomitantly lower benefit – then everyone sticks to that treatment. This clearly comes at the expense of the exponential discounters’ health, but if the proportion of present-bias is high enough it can be an efficient outcome.

Were you able to compare the time preferences of patients and of doctors?

Not this time! It had been the ‘grand plan’ at the start of the PhD to compare matched doctor and patient time preferences then link it to treatment choices but that was far too ambitious for the time, and there had been very little work establishing how time preferences work in the patient-doctor interaction so I felt we had a lot to do.

One interesting question we did ask was whether doctors’ time preferences for themselves were the same as for their patients. A lot of the existing evidence asks doctors for their own time preferences, but surely the important time preference is the one they apply to their patients?

We found that while there was little difference between these professional and private time preferences, a lot of the responses displayed increasing impatience. This means that as the start of treatment gets pushed further into the future, doctors started to prefer shorter-but-sooner benefits for themselves and their patients. We’re still thinking about whether this reflects that in the real world (outside the survey) doctors already account for the time patients have spent with symptoms when assessing how quickly a treatment benefit should arrive.

How could doctors alter their practice to reduce non-adherence?

We really only have two options – to make ‘the right thing’ easier or the ‘wrong thing’ more costly. The implication of present-bias is you need to use less intense treatments because the problem is the (relative) over-weighting of the side effects. The important thing we need for that is good information on adherence.

We could pay people to adhere to treatment. However, my gut feeling is that payments are hard to implement on the patient side without being coercive (e.g making non-adherence costly with charges) or expensive for the implementer when identification of completion is tricky (giving bonuses to doctors based on patient health outcomes). So doctors can reduce non-adherence by anticipating it, and offering less ‘painful’ treatments.

It’s important to say I was only looking at one kind of non-adherence. If patients have bad experiences then whatever we do shouldn’t keep them taking a treatment they don’t want. However, the fact that stopping treatment is always an option for the patient makes non-adherence hard to address because as an economist you would like to separate different reasons for stopping. This is a difficulty for analysing non-adherence as a problem of temptation. In temptation preferences we would like to change the outcome set so that ‘no treatment’ is not a tempting choice, but there are real ethical and practical difficulties with that.

To what extent did the evidence generated by your research support theoretical predictions?

I designed a lab experiment that put students in the role of the doctor with patients that may or may not be present-biased. The participants had to recommend treatments to a series of hypothetical patients and was set up so that adapting to non-adherence with less intense treatments was best. Participants got feedback on their previous patients, to learn about which treatments patients stuck to over the rounds.

We paid one arm a salary, and another a ‘performance payment’. The latter only got paid when patients stuck to treatment and the pay correlated with the patient outcomes. In both arms, patients’ outcomes were reflected with a charity donation.

The main result is that there was a lot of adaptation to non-adherence in both arms. The adaptation was stronger under the performance payment, reflecting the upper limit of the adaptation we can expect because it perfectly aligns patient and doctor preferences.

In the experimental setting, even when there is no direct financial benefit of doing so, participants adapted to non-adherence in the way I predicted.