Every Monday our authors provide a round-up of the latest peer-reviewed journal publications. We cover all issues of major health economics journals as well as some other notable releases. If you’d like to write one of our weekly journal round-ups, get in touch.
American Journal of Health Economics
Volume 7, Number 1
We’ve featured numerous studies in the round-ups recently that concern the impact of parental (especially maternal) health and circumstances on children. But could a child’s education affect their parents’ health? That’s obviously a tricky question; healthy parents are more likely to have well-educated kids. But a study in this issue suggests that the reverse applies too. The analysis relies on longitudinal data, using geographic proximity to schools as an instrumental variable. I half-hoped to see some moral hazard in parental behaviour; why look after yourself if you have a well-educated kid who might do it instead? Contrary to this, the main finding is that an extra year of schooling for kids in China leads to a 7.7 percentage point increase in their parents’ likelihood of being in good health. The authors suggest that the mechanisms may include better management of chronic conditions, reduced smoking, and more preventive care.
Analyses of flu vaccination behaviour seem apposite during the COVID pandemic, and one in this issue shows what you might expect. People in the US who get the flu, having not had the jab, are later more likely to get vaccinated. People who do get the jab but then get the flu are subsequently less likely to get vaccinated.
Another study reports on an evaluation of a programme to reduce 30-day hospital readmissions within Medicare. The programme being evaluated is the Chicago Southland Coalition for Transition Care, which involved social workers. The authors use a difference-in-differences model and find that, overall, readmissions were reduced by about 14%. The associated cost saving was approximate to the cost of running the programme.
Finally, the market for illegal opioids may be stimulated by the reformulation of prescription opioids, which may have occurred in the US with OxyContin. A study in this issue compares states’ trends in overdoses, modelling their exposure to OxyContin reformulation using nonmedical OxyContin use rates before reformulation as an indicator. The findings suggest that reformulation is associated with more fatal overdoses from fentanyl and cocaine.
Cost Effectiveness and Resource Allocation
Volume 19, February 2021
Several applied cost-effectiveness analyses were published in CERA last month, for pneumonia case management in Nigeria, adjuvant therapies for breast cancer in China, and text messaging to support smoking cessation. There’s also a study using the EUnetHTA Core Model to evaluate a technology in South Africa. The authors use an ultrasound technology for their case study and find that the Core Model provided a useful framework, but that pragmatic adaptations to the context were needed. Last month also included a study on access to WHO ‘key access’ antibiotics in Pakistan, showing that private sector pharmacies often do not stock the drugs and that some are unaffordable.
The study that interested me most used the global burden of disease study to estimate cost-per-DALY thresholds. The authors construct a panel for 176 countries between 2000 and 2016, including data on health expenditure, GDP per capita, the Human Development Index (HDI), and age-standardised DALY rates. The estimated cost-per-DALY averted is summarised as a proportion of GDP per capita, ranging from 0.34 for low HDI countries to 1.46 for very high HDI countries. It’s interesting to present these figures, not least to aid comparison between different countries. But to imply – as the authors do – that these figures represent anything approximating opportunity cost, or that they should be used at the national level to prioritise resource allocation, seems pretty foolish. It should come as no surprise that health expenditures and DALYs are correlated across countries and years. But the authors seem to make no attempt to demonstrate causality. And no thought is given to the fact that the quality of the DALY estimates may well be associated with the level of health expenditure.
PharmacoEconomics
Volume 39, Issue 2
NICE’s methods review has provided nourishment for researchers in recent months. The latest issue of PharmacoEconomics bears some of the fruit. One of the items on the table is discount rates. An opinion piece in this issue supports the notion that NICE’s discount rate should be reduced to 1.5%. But the authors echo NICE in identifying a link between lowering a discount rate and lowering the cost-effectiveness threshold. That’s fine in theory, but it makes little sense to formally link the two when nobody is confident about the basis for either. This issue’s ‘leading article’ comes from a similar perspective, discussing NICE’s approach to modifiers or ‘equity weights’ in HTA. The authors claim to ‘demonstrate’ that NICE’s current approach reduces (equity-weighted) population health. Of course, the article does nothing of the sort because that would require a whole load of evidence of the kind that we simply don’t have. Rather, the authors set out a series of simple theoretical expositions that might be used to simulate the consequences of alternative approaches to equity weighting, relative to alternative objectives. The use of some unrealistic assumptions (mostly about prices and opportunity costs) render the authors’ primary assertions about the current state of affairs untenable. Nevertheless, it’s a useful exercise. There is some convincing logic in the authors’ support for a ‘net benefit’ approach, whereby outcome weighting is incorporated within a decision analysis rather than applied to ICERs.
Keeping with HTA methods is a study on the presentation of uncertainty in cost-effectiveness analysis. The author explains that cost-effectiveness acceptability curves may hide important information about uncertainty, particularly the relative uncertainty in the costs and outcomes associated with alternative programmes. A CEAC may, for example, show that an intervention with highly uncertain effects is more likely to be cost-effective than an intervention with highly certain effects, and a risk-averse decision-maker would be none-the-wiser. The proposed strategy facilitates a ranking of alternatives according to their risk-adjusted net monetary benefit. It gives us a new acronym in CERAC – the cost-effectiveness risk-aversion curve – plotting the net benefit-to-risk ratio against alternative threshold values. I’m surprised this hasn’t appeared in the literature sooner. I recall a teacher criticising the CEAC on these grounds during my MSc in 2010. Also on HTA methods is a study exploring the use of a social accounting matrix in understanding indirect non-medical costs in an economic evaluation that adopts a societal perspective. The researchers use a case study of rotavirus vaccination in the Netherlands to demonstrate how we might expect this to give different results to the current research standards.
The issue also includes a couple of applied studies, with a NICE review group article on an appraisal of lymphoma treatments and a budget impact model for an ovarian cancer drug. There are a couple of reviews, one of cost-effectiveness analyses of osteoporosis drugs and another of utility values for people with heart failure. On a similar note, there’s an analysis of utility values from a database of multiple sclerosis patients.
Credits
