Sam Watson’s journal round-up for 11th February 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.

Contest models highlight inherent inefficiencies of scientific funding competitions. PLoS Biology [PubMed] Published 2nd January 2019

If you work in research you will have no doubt thought to yourself at one point that you spend more time applying to do research than actually doing it. You can spend weeks working on (what you believe to be) a strong proposal only for it to fail against other strong bids. That time could have been spent collecting and analysing data. Indeed, the opportunity cost of writing extensive proposals can be very high. The question arises as to whether there is another method of allocating research funding that reduces this waste and inefficiency. This paper compares the proposal competition to a partial lottery. In this lottery system, proposals are short, and among those that meet some qualifying standard those that are funded are selected at random. This system has the benefit of not taking up too much time but has the cost of reducing the average scientific value of the winning proposals. The authors compare the two approaches using an economic model of contests, which takes into account factors like proposal strength, public benefits, benefits to the scientist like reputation and prestige, and scientific value. Ultimately they conclude that, when the number of awards is smaller than the number of proposals worthy of funding, the proposal competition is inescapably inefficient. It means that researchers have to invest heavily to get a good project funded, and even if it is good enough it may still not get funded. The stiffer the competition the more researchers have to work to win the award. And what little evidence there is suggests that the format of the application makes little difference to the amount of time spent by researchers on writing it. The lottery mechanism only requires the researcher to propose something that is good enough to get into the lottery. Far less time would therefore be devoted to writing it and more time spent on actual science. I’m all for it!

Preventability of early versus late hospital readmissions in a national cohort of general medicine patients. Annals of Internal Medicine [PubMed] Published 5th June 2018

Hospital quality is hard to judge. We’ve discussed on this blog before the pitfalls of using measures such as adjusted mortality differences for this purpose. Just because a hospital has higher than expected mortality does not mean those death could have been prevented with higher quality care. More thorough methods assess errors and preventable harm in care. Case note review studies have suggested as little as 5% of deaths might be preventable in England and Wales. Another paper we have covered previously suggests then that the predictive value of standardised mortality ratios for preventable deaths may be less than 10%.

Another commonly used metric is readmission rates. Poor care can mean patients have to return to the hospital. But again, the question remains as to how preventable these readmissions are. Indeed, there may also be substantial differences between those patients who are readmitted shortly after discharge and those for whom it may take a longer time. This article explores the preventability of early and late readmissions in ten hospitals in the US. It uses case note review and a number of reviewers to evaluate preventability. The headline figures are that 36% of early readmissions are considered preventable compared to 23% of late readmissions. Moreover, it was considered that the early readmissions were most likely to have been preventable at the hospital whereas for late readmissions, an outpatient clinic or the home would have had more impact. All in all, another paper which provides evidence to suggest crude, or even adjusted rates, are not good indicators of hospital quality.

Visualisation in Bayesian workflow. Journal of the Royal Statistical Society: Series A (Statistics in Society) [RePEc] Published 15th January 2019

This article stems from a broader programme of work from these authors on good “Bayesian workflow”. That is to say, if we’re taking a Bayesian approach to analysing data, what steps ought we to be taking to ensure our analyses are as robust and reliable as possible? I’ve been following this work for a while as this type of pragmatic advice is invaluable. I’ve often read empirical papers where the authors have chosen, say, a logistic regression model with covariates x, y, and z and reported the outcomes, but at no point ever justified why this particular model might be any good at all for these data or the research objective. The key steps of the workflow include, first, exploratory data analysis to help set up a model, and second, performing model checks before estimating model parameters. This latter step is important: one can generate data from a model and set of prior distributions, and if the data that this model generates looks nothing like what we would expect the real data to look like, then clearly the model is not very good. Following this, we should check whether our inference algorithm is doing its job, for example, are the MCMC chains converging? We can also conduct posterior predictive model checks. These have had their criticisms in the literature for using the same data to both estimate and check the model which could lead to the model generalising poorly to new data. Indeed in a recent paper of my own, posterior predictive checks showed poor fit of a model to my data and that a more complex alternative was better fitting. But other model fit statistics, which penalise numbers of parameters, led to the alternative conclusions. So the simpler model was preferred on the grounds that the more complex model was overfitting the data. So I would argue posterior predictive model checks are a sensible test to perform but must be interpreted carefully as one step among many. Finally, we can compare models using tools like cross-validation.

This article discusses the use of visualisation to aid in this workflow. They use the running example of building a model to estimate exposure to small particulate matter from air pollution across the world. Plots are produced for each of the steps and show just how bad some models can be and how we can refine our model step by step to arrive at a convincing analysis. I agree wholeheartedly with the authors when they write, “Visualization is probably the most important tool in an applied statistician’s toolbox and is an important complement to quantitative statistical procedures.”

Credits

 

Chris Sampson’s journal round-up for 4th February 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.

Patient choice and provider competition – quality enhancing drivers in primary care? Social Science & Medicine Published 29th January 2019

There’s no shortage of studies in economics claiming to identify the impact (or lack of impact) of competition in the market for health care. The evidence has brought us close to a consensus that greater competition might improve quality, so long as providers don’t compete on price. However, many of these studies aren’t able to demonstrate the mechanism through which competition might improve quality, and the causality is therefore speculative. The research reported in this article was an attempt to see whether the supposed mechanisms for quality improvement actually exist. The authors distinguish between the demand-side mechanisms of competition-increasing quality-improving reforms (i.e. changes in patient behaviour) and the supply-side mechanisms (i.e. changes in provider behaviour), asserting that the supply-side has been neglected in the research.

The study is based on primary care in Sweden’s two largest cities, where patients can choose their primary care practice, which could be a private provider. Key is the fact that patients can switch between providers as often as they like, and with fewer barriers to doing so than in the UK. Prospective patients have access to some published quality indicators. With the goal of maximum variation, the researchers recruited 13 primary health care providers for semi-structured interviews with the practice manager and (in most cases) one or more of the practice GPs. The interview protocol included questions about the organisation of patient visits, information received about patients’ choices, market situation, reimbursement, and working conditions. Interview transcripts were coded and a framework established. Two overarching themes were ‘local market conditions’ and ‘feedback from patient choice’.

Most interviewees did not see competitors in the local market as a threat – conversely, providers are encouraged to cooperate on matters such as public health. Where providers did talk about competing, it was in terms of (speed of) access for patients, or in competition to recruit and keep staff. None of the interviewees were automatically informed of patients being removed from their list, and some managers reported difficulties in actually knowing which patients on their list were still genuinely on it. Even where these data were more readily available, nobody had access to information on reasons for patients leaving. Managers saw greater availability of this information as useful for quality improvement, while GPs tended to think it could be useful in ensuring continuity of care. Still, most expressed no desire to expand their market share. Managers reported using marketing efforts in response to greater competition generally, rather than as a response to observed changes within their practice. But most relied on reputation. Some reported becoming more service-minded as a result of choice reforms.

It seems that practices need more information to be able to act on competitive pressures. But, most practices don’t care about it because they don’t want to expand and they face no risk of there being a shortage of patients (in cities, at least). And, even if they did want to act on the information, chances are it would just create an opportunity for them to improve access as a way of cherry-picking younger and healthier people who demand convenience. Primary care providers (in this study, at least) are not income maximisers, but satisficers (they want to break-even), so there isn’t much scope for reforms to encourage providers to compete for new patients. Patient choice reforms may improve quality, but it isn’t clear that this has anything to do with competitive pressure.

Maximising the impact of patient reported outcome assessment for patients and society. BMJ [PubMed] Published 24th January 2019

Patient-reported outcome measures (PROMs) have been touted as a way of improving patient care. Yet, their use around the world is fragmented. In this paper, the authors make some recommendations about how we might use PROMs to improve patient care. The authors summarise some of the benefits of using PROMs and discuss some of the ways that they’ve been used in the UK.

Five key challenges in the use of PROMs are specified: i) appropriate and consistent selection of the best measures; ii) ethical collection and reporting of PROM data; iii) data collection, analysis, reporting, and interpretation; iv) data logistics; and v) a lack of coordination and efficiency. To address these challenges, the authors recommend an ‘integrated’ approach. To achieve this, stakeholder engagement is important and a governance framework needs to be developed. A handy table of current uses is provided.

I can’t argue with what the paper proposes, but it outlines an idealised scenario rather than any firm and actionable recommendations. What the authors don’t discuss is the fact that the use of PROMs in the UK is flailing. The NHS PROMs programme has been scaled back, measures have been dropped from the QOF, the EQ-5D has been dropped from the GP Patient Survey. Perhaps we need bolder recommendations and new ideas to turn the tide.

Check your checklist: the danger of over- and underestimating the quality of economic evaluations. PharmacoEconomics – Open [PubMed] Published 24th January 2019

This paper outlines the problems associated with misusing methodological and reporting checklists. The author argues that the current number of checklists available in the context of economic evaluation and HTA (13, apparently) is ‘overwhelming’. Three key issues are discussed. First, researchers choose the wrong checklist. A previous review found that the Drummond, CHEC, and Philips checklists were regularly used in the wrong context. Second, checklists can be overinterpreted, resulting in incorrect conclusions. A complete checklist does not mean that a study is perfect, and different features are of varying importance in different studies. Third, checklists are misused, with researchers deciding which items are or aren’t relevant to their study, without guidance.

The author suggests that more guidance is needed and that a checklist for selecting the correct checklist could be the way to go. The issue of updating checklists over time – and who ought to be responsible for this – is also raised.

In general, the tendency seems to be to broaden the scope of general checklists and to develop new checklists for specific methodologies, requiring the application of multiple checklists. As methods develop, they become increasingly specialised and heterogeneous. I think there’s little hope for checklists in this context unless they’re pared down and used as a reminder of the more complex guidance that’s needed to specify suitable methods and achieve adequate reporting. ‘Check your checklist’ is a useful refrain, though I reckon ‘chuck your checklist’ can sometimes be a better strategy.

A systematic review of dimensions evaluating patient experience in chronic illness. Health and Quality of Life Outcomes [PubMed] Published 21st January 2019

Back to PROMs and PRE(xperience)Ms. This study sets out to understand what it is that patient-reported measures are being used to capture in the context of chronic illness. The authors conducted a systematic review, screening 2,375 articles and ultimately including 107 articles that investigated the measurement properties of chronic (physical) illness PROMs and PREMs.

29 questionnaires were about (health-related) quality of life, 19 about functional status or symptoms, 20 on feelings and attitudes about illness, 19 assessing attitudes towards health care, and 20 on patient experience. The authors provide some nice radar charts showing the percentage of questionnaires that included each of 12 dimensions: i) physical, ii) functional, iii) social, iv) psychological, v) illness perceptions, vi) behaviours and coping, vii) effects of treatment, viii) expectations and satisfaction, ix) experience of health care, x) beliefs and adherence to treatment, xi) involvement in health care, and xii) patient’s knowledge.

The study supports the idea that a patient’s lived experience of illness and treatment, and adaptation to that, has been judged to be important in addition to quality of life indicators. The authors recommend that no measure should try to capture everything because there are simply too many concepts that could be included. Rather, researchers should specify the domains of interest and clearly define them for instrument development.

Credits

 

Rita Faria’s journal round-up for 28th January 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.

Appraising the value of evidence generation activities: an HIV modelling study. BMJ Global Health [PubMed] Published 7th December 2018

How much should we spend on implementing our health care strategy versus getting more information to devise a better strategy? Should we devolve budgets to regions or administer the budget centrally? These are difficult questions and this new paper by Beth Woods et al has a brilliant stab at answering them.

The paper looks at the HIV prevention and treatment policies in Zambia. It starts by finding the most cost-effective strategy and the corresponding budget in each region, given what is currently known about the prevalence of the infection, the effectiveness of interventions, etc. The idea is that the regions receive a cost-effective budget to implement a cost-effective strategy. The issue is that the cost-effective strategy and budget are devised according to what we currently know. In practice, regions might face a situation on the ground which is different from what was expected. Regions might not have enough budget to implement the strategy or might have some leftover.

What if we spend some of the budget to get more information to make a better decision? This paper considers the value of perfect information given the costs of research. Depending on the size of the budget and the cost of research, it may be worthwhile to divert some funds to get more information. But what if we had more flexibility in the budgetary policy? This paper tests 2 more budgetary options: a national hard budget but with the flexibility to transfer funds from under- to overspending regions, and a regional hard budget with a contingency fund.

The results are remarkable. The best budgetary policy is to have a national budget with the flexibility to reallocate funds across regions. This is a fascinating paper, with implications not only for prioritisation and budget setting in LMICs but also for high-income countries. For example, the 2012 Health and Social Care Act broke down PCTs into smaller CCGs and gave them hard budgets. Some CCGs went into deficit, and there are reports that some interventions have been cut back as a result. There are probably many reasons for the deficit, but this paper shows that hard regional budgets clearly have negative consequences.

Health economics methods for public health resource allocation: a qualitative interview study of decision makers from an English local authority. Health Economics, Policy and Law [PubMed] Published 11th January 2019

Our first paper looked at how to use cost-effectiveness to allocate resources between regions and across health care services and research. Emma Frew and Katie Breheny look at how decisions are actually made in practice, but this time in a local authority in England. Another change of the 2012 Health and Social Care Act was to move public health responsibilities from the NHS to local authorities. Local authorities are now given a ring-fenced budget to implement cost-effective interventions that best match their needs. How do they make decisions? Thanks to this paper, we’re about to find out.

This paper is an enjoyable read and quite an eye-opener. It was startling that health economics evidence was not much used in practice. But the barriers that were cited are not insurmountable. And the suggestions by the interviewees were really useful. There were suggestions about how economic evaluations should consider the local context to get a fair picture of the impact of the intervention to services and to the population, and to move beyond the trial into the real world. Equity was mentioned too, as well as broadening the outcomes beyond health. Fortunately, the health economics community is working on many of these issues.

Lastly, there was a clear message to make economic evidence accessible to lay audiences. This is a topic really close to my heart, and something I’d like to help improve. We have to make our work easy to understand and use. Otherwise, it may stay locked away in papers rather than do what we intended it for. Which is, at least in my view, to help inform decisions and to improve people’s lives.

I found this paper reassuring in that there is clearly a need for economic evidence and a desire to use it. Yes, there are some teething issues, but we’re working in the right direction. In sum, the future for health economics is bright!

Survival extrapolation in cancer immunotherapy: a validation-based case study. Value in Health Published 13th December 2018

Often, the cost-effectiveness of cancer drugs hangs in the method to extrapolate overall survival. This is because many cancer drugs receive their marketing authorisation before most patients in the trial have died. Extrapolation is tested extensively in the sensitivity analysis, and this is the subject of many discussions in NICE appraisal committees. Ultimately, at the point of making the decision, the correct method to extrapolate is a known unknown. Only in hindsight can we know for sure what the best choice was.

Ash Bullement and colleagues take advantage of hindsight to know the best method for extrapolation of a clinical trial of an immunotherapy drug. Survival after treatment with immunotherapy drugs is more difficult to predict because some patients can survive for a very long time, while others have much poorer outcomes. They fitted survival models to the 3-year data cut, which was available at the time of the NICE technology appraisal. Then they compared their predictions to the observed survival in the 5-year data cut and to long-term survival trends from registry data. They found that the piecewise model and a mixture-cure model had the best predictions at 5 years.

This is a relevant paper for those of us who work in the technology appraisal world. I have to admit that I can be sceptical of piecewise and mixture-cure models, but they definitely have a role in our toolbox for survival extrapolation. Ideally, we’d have a study like this for all the technology appraisals hanging on the survival extrapolation so that we can take learnings across cancers and classes of drugs. With time, we would get to know more about what works best for which condition or drug. Ultimately, we may be able to get to a stage where we can look at the extrapolation with less inherent uncertainty.

Credits