Chris Sampson’s journal round-up for 2nd December 2019

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.

The treatment decision under uncertainty: the effects of health, wealth and the probability of death. Journal of Health Economics Published 16th November 2019

It’s important to understand how people make decisions about treatment. At the end of life, the question can become a matter of whether to have treatment or to let things take their course such that you end up dead. In order to consider this scenario, the author of this paper introduces the probability of death to some existing theoretical models of decision-making under uncertainty.

The diagnostic risk model and the therapeutic risk model can be used to identify risk thresholds that determine decisions about treatment. The diagnostic model relates to the probability that disease is present and the therapeutic model relates to the probability that treatment is successful. The new model described in this paper builds on these models to consider the impact on the decision thresholds of i) initial health state, ii) probability of death, and iii) wealth. The model includes wealth after death, in the form of a bequest. Limited versions of the model are also considered, excluding the bequest and excluding wealth (described as a ‘QALY model’). Both an individual perspective and an aggregate perspective are considered by excluding and including the monetary cost of diagnosis and treatment, to allow for a social insurance type setting.

The comparative statics show a lot of ambiguity, but there are a few things that the model can tell us. The author identifies treatment as having an ‘insurance effect’, by reducing diagnostic risk, a ‘protective effect’, by lowering the probability of death, and a risk-increasing effect associated with therapeutic risk. A higher probability of death increases the propensity for treatment in both the no-bequest model and the QALY model, because of the protective effect of treatment. In the bequest model, the impact is ambiguous, because treatment costs reduce the bequest. In the full model, wealthier individuals will choose to undergo treatment at a lower probability of success because of a higher marginal utility for survival, but the effect becomes ambiguous if the marginal utility of wealth depends on health (which it obviously does).

I am no theoretician, so it can take me a long time to figure these things out in my head. For now, I’m not convinced that it is meaningful to consider death in this way using a one-period life model. In my view, the very definition of death is a loss of time, which plays little or no part in this model. But I think my main bugbear is the idea that anybody’s decision about life saving treatment is partly determined by the amount of money they will leave behind. I find this hard to believe. The author links the finding that a higher probability of death increases treatment propensity to NICE’s end of life premium. Though I’m not convinced that the model has anything to do with NICE’s reasoning on this matter.

Moving toward evidence-based policy: the value of randomization for program and policy implementation. JAMA [PubMed] Published 15th November 2019

Evidence-based policy is a nice idea. We should figure out whether something works before rolling it out. But decision-makers (especially politicians) tend not to think in this way, because doing something is usually seen to be better than doing nothing. The authors of this paper argue that randomisation is the key to understanding whether a particular policy creates value.

Without evidence based on random allocation, it’s difficult to know whether a policy works. This, the authors argue, can undermine the success of effective interventions and allow harmful policies to persist. A variety of positive examples are provided from US healthcare, including trials of Medicare bundled payments. Apparently, such trials increased confidence in the programmes’ effects in a way that post hoc evaluations cannot, though no evidence of this increased confidence is actually provided. Policy evaluation is not always easy, so the authors describe four preconditions for the success of such studies: i) early engagement with policymakers, ii) willingness from policy leaders to support randomisation, iii) timing the evaluation in line with policymakers’ objectives, and iv) designing the evaluation in line with the realities of policy implementation.

These are sensible suggestions, but it is not clear why the authors focus on randomisation. The paper doesn’t do what it says on the tin, i.e. describe the value of randomisation. Rather, it explains the value of pre-specified policy evaluations. Randomisation may or may not deserve special treatment compared with other analytical tools, but this paper provides no explanation for why it should. The authors also suggest that people are becoming more comfortable with randomisation, as large companies employ experimental methods, particularly on the Internet with A/B testing. I think this perception is way off and that most people feel creeped out knowing that the likes of Facebook are experimenting on them without any informed consent. In the authors’ view, it being possible to randomise is a sufficient basis on which to randomise. But, considering the ethics, as well as possible methodological contraindications, it isn’t clear that randomisation should become the default.

A new tool for creating personal and social EQ-5D-5L value sets, including valuing ‘dead’. Social Science & Medicine Published 30th November 2019

Nobody can agree on the best methods for health state valuation. Or, at least, some people have disagreed loud enough to make it seem that way. Novel approaches to health state valuation are therefore welcome. Even more welcome is the development and testing of methods that you can try at home.

This paper describes the PAPRIKA method (Potentially All Pairwise RanKings of all possible Alternatives) of discrete choice experiment, implemented using 1000Minds software. Participants are presented with two health states that are defined in terms of just two dimensions, each lasting for 10 years, and asked to choose between them. Using the magical power of computers, an adaptive process identifies further choices, automatically ranking states using transitivity so that people don’t need to complete unnecessary tasks. In order to identify where ‘dead’ sits on the scale, a binary search procedure asks participants to compare EQ-5D states with being dead. What’s especially cool about this process is that everybody who completes it is able to view their own personal value set. These personal value sets can then be averaged to identify a social value set.

The authors used their tool to develop an EQ-5D-5L value set for New Zealand (which is where the researchers are based). They recruited 5,112 people in an online panel, such that the sample was representative of the general public. Participants answered 20 DCE questions each, on average, and almost half of them said that they found the questions difficult to answer. The NZ value set showed that anxiety/depression was associated with the greatest disutility, though each dimension has a notably similar level of impact at each level. The value set correlates well with numerous existing value sets.

The main limitation of this research seems to be that only levels 1, 3, and 5 of each EQ-5D-5L domain were included. Including levels 2 and 4 would more than double the number of questions that would need to be answered. It is also concerning that more than half of the sample was excluded due to low data quality. But the authors do a pretty good job of convincing us that this is for the best. Adaptive designs of this kind could be the future of health state valuation, especially if they can be implemented online, at low cost. I expect we’ll be seeing plenty more from PAPRIKA.

Credits

Using Discrete Choice Experiments in Health Economics Course

This popular course, offered by the Health Economics Research Unit (HERU) at the University of Aberdeen, Scotland, covers the theoretical and practical issues of discrete choice experiments (DCEs) in health economics. The course takes place annually and in 2018 was fully booked.

The course provides:

  • An introduction to the theoretical basis for the development and application of DCEs in health economics.
  • Step by step guide to the design of DCEs, questionnaire development, data input, data analysis and interpretation of results.
  • An update on methodological issues raised in the application of DCEs in health economics.

Jason Shafrin’s journal round-up for 7th October 2019

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.

Combined impact of future trends on healthcare utilisation of older people: a Delphi study. Health Policy [PubMed] [RePEc] Published October 2019

Governments need to plan for the future. This is particularly important in countries where the government pays for the lion’s share of health care expenditures. Predicting the future, however, is not an easy task. One could use quantitative approaches and simply extrapolate recent trends. One could attempt to consult with political experts to determine what policies are likely to be incurred. Another approach is to use a Delphi Panel to elicit expert opinions on future trends in health care utilization to help predict future health care needs. This approach was the one taken by Ravensbergen and co-authors in an attempt to predict trends in health care utilization among older adults in the Netherlands in 2040.

The Delphi Panel approach was applied in this study as follows. First, individuals received a questionnaire via email. Researchers presented the experts with trends from the Dutch Public Health Foresight Study (Volksgezondheid Toekomst Verkenning) to help ground all experts with the same baseline information. The data and questions largely asked separately about trends for either the old (65–80 years) or the oldest old (>80 years). After the responses from the first questionnaire were received, responses were summarized and provided back to each panelist in an anonymous manner. Panelists were then able to revise their views on a second questionnaire taking into account the feedback by the other panelists. Because the panelists did not meet in person, this approach should be considered a modified Delphi Panel.

The Delphi panel identified three broad trends: increased use of eHealth tools, less support, and change in health status. While the panel thought eHealth was important, experts rarely reached consensus how eHealth would affect healthcare utilization. The experts did find consensus, however, in believing that the the share of adults aged 50-64 will decline relative to the share of individuals aged ≥ 85 years, implying fewer caregivers will be available and more of the oldest old will be living independently (i.e. with less support). Because less informal care will be available, the Delphi believed that the demand for home care and general practitioner services will rise. The respondents also believed that in most cases changes in health status will increase health care utilization of general practitioner and specialist services. There was less agreement about trends in the need for long-term care or mental health services, however.

The Delphi Panel approach may be useful to help governments predict future demand for services. More rigorous approaches, such as betting markets, are likely not feasible since the payouts would take too long to generate much interest. Betting markets could be used to predict shorter-run trends in health care utilization. The risk with betting markets, however, is that some individuals could act strategically to drive up or down predictions to increase or decrease reimbursement for certain sectors.

In short, the Delphi Panel is likely a reasonable, low-cost approach for predicting trends in health care utilization. Future studies, however, should validate how good the predictions are from using this type of method.

The fold-in, fold-out design for DCE choice tasks: application to burden of disease. Medical Decision Making [PubMed] Published 29th May 2019

Discrete choice experiments (DCEs) are a useful way to determine what treatment attributes patients (or providers or caregivers) value. Respondents are presented with multiple treatment options and the options can be compared across a series of attributes. An attribute could be treatment efficacy, safety, dosing, cost, or a host of other attributes. One can use this approach to measure the marginal rate of substitution across attributes. If cost is one of the attributes, one can measure willingness to pay for specific attributes.

One of the key challenges of DCEs, however, is attribute selection. Most treatments differ across a range of attributes. Most published DCEs however have four, five, or at most seven attributes presented. Including more attributes makes comparisons too complicated for most respondents. Thus, researchers are left with a difficult choice: (i) a tractable but overly simplified survey, or (ii) a realistic, but overly complex survey unlikely to be comprehended by respondents.

One solution proposed by Lucas Goossens and co-authors is to use a Fold-in Fold-out (FiFo) approach. In this approach, related attributes may be grouped into domains. For some questions, all attributes within the same domain have the same attribute level (i.e., fold in); in other questions, attributes may vary within the domain (i.e., fold out).

To be concrete, in the Goossens paper, they examine treatments for chronic obstructive pulmonary disorder (COPD). They use 15 attributes divided into three domains plus two stand-alone attributes:

a respiratory symptoms domain (with four attributes: shortness of breath at rest, shortness of breath during physical activity, coughing, and sputum production), a limitations domain (four attributes: limitations in strenuous physical activities, limitations in moderate physical activities, limitations in daily activities, and limitations in social activities), a mental problems domain (five attributes: feeling depressed, fearing that breathing gets worse, worrying, listlessness, and tense feeling), a fatigue attribute, and an exacerbations attribute.

This creative approach simplifies the choice set for respondents, but allows for a large number of attributes. Using the data collected, the authors used a Bayesian mixed logit regression model to conduct the analysis. The utility function underlying this assumed domain-specific parameters, but also included parameters for within-domain attribute weights to vary in the questions where it was folded out.

One key challenge, however, is that the authors found that individuals placed more weight on attributes when their domains were folded out (i.e., attribute levels varied within domain) compared to when their domains were folded in (i.e., attribute levels were the same within the domain). Thus, I would say that if five, six or seven attributes can capture the lion’s share of differences in treatment attributes across treatments, use the standard approach; however, if more attributes are needed, the FiFo approach is an attractive option researchers should consider.

The health and cost burden of antibiotic resistant and susceptible Escherichia coli bacteraemia in the English hospital setting: a national retrospective cohort study. PLoS One [PubMed] Published 10th September 2019

Bacterial infections are bad. The good news is that we have antibiotics to treat them so they no longer are a worry, right? While conventional wisdom may believe that we have many antibiotics to treat these infections, in recent years antibiotic resistance has grown. If antibiotics no longer are effective, what is the cost to society?

One effort to quantify the economic burden of antibiotic resistance by Nichola Naylor and co-authors used national surveillance and administrative data from National Health Service (NHS) hospitals in England. They compared the cost for patients with similar observable characteristics with E. coli bacteraemia compared to those who did not have E. coli bacteraemia. Antibiotic resistance in the study was defined as E. coli bacteraemia using laboratory-based definitions of ‘resistant’ and ‘intermediate’ isolates. The antibiotics to which resistance was considered included ciprofloxacin, third generation cephalosporins (ceftazidime and/or cefotaxime), gentamicin, piperacillin/tazobactam and carbapenems (imipenem and/or meropenem).

The authors use an Aalen-Johansen estimator to measure cumulative incidence of in-hospital mortality and length of stay. Both approaches control for the patient’s age, sex, Elixhauser comorbidity index, and hospital trust type. It does not appear that the authors control for the reason for admission to the hospital nor do they propensity match people with those without antibiotic resistance. Thus, it is likely that significant unobserved heterogeneity across groups remains in the analysis.

Despite these limitations, the authors do have some interesting findings. First, bacterial infections are associated with increased risk of death. In-hospital mortality is 14.3% for individuals infected with E. Coli compared to 1.3% for those not infected. Accounting for covariates, the subdistribution hazard rate (SHR) for in-hospital mortality due to E. coli bacteraemia was 5.88. Second, E. coli bacteraemia was associated with 3.9 excess hospital days compared to patients who were not antibiotic resistance. These extra hospital days cost £1,020 per case of E. coli bacteraemia and the estimated annual cost of E. coli bacteraemia in England was £14.3m. If antibiotic resistance has increased in recent years, these estimates are likely to be conservative.

The issue of antibiotic resistance presents a conundrum for policymakers. If current antibiotics are effective, drug-makers will have no incentive to develop new antibiotics since the new treatments are unlikely to be prescribed. On the other hand, failing to identify new antibiotics in reserve means that as antibiotic resistance grows, there will be few treatment alternatives. To address this issue, the United Kingdom is considering a ‘subscription style‘ approach to pay for new antibiotics to incentivize the development of new treatments.

Nevertheless, the paper by Naylor and co-authors provides a useful data point on the cost of antibiotic resistance.

Credits