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.
Volume 30, Issue 4
There’s a lot of mental health research in the latest issue of Health Economics, which is great for me. An analysis of some real-world data attempts to identify the impacts of psychotherapy for schizophrenia. The authors used hospital data and employed an instrumental variable approach to identify changes in suicide attempts, readmissions, and labour market outcomes associated with hospitals’ varying propensities to refer patients to psychotherapy. It’s of some concern that the authors weren’t able to find any clear benefits from psychotherapy. Another study reports on a survey focussing on the emotional wellbeing and mental health of child labourers in India. The survey included 947 children and collected data on various indicators such as happiness, stress, and fear. As you might expect, these kids have got it particularly rough.
Other studies in mental health evaluate policy changes and market structures. One study reports on a difference-in-differences analysis to evaluate the introduction of mental health parity laws for insurance coverage across US states. The author focuses on the impact on pregnant women and newborns. The introduction of the laws was associated with a drop in the number of women in labour receiving a mental health diagnosis and a decrease in the probability of adverse birth outcomes. Also from the US is a study showing that improved access to mental health care can reduce disability insurance enrollment. A study from the Netherlands looked at market power and price setting in mental health care. In 2014, a new policy required that providers negotiate prices with each insurer. The analysis shows that providers with greater market power negotiated significantly higher prices.
A Health Economics Letter pursues an interesting idea; how might health status and economic preferences be associated? By ‘economic preferences’, we mean stuff that might influence all of a person’s choices and decisions, such as time preferences and risk preferences. The researchers use international survey data to show that people with a limiting physical or mental health condition tend to attach less value to future outcomes and are more risk-averse.
Other topics covered in this issue include the costs associated with pollution, DRG-based reimbursement for sepsis conditions, the impact of market entry of for-profit providers in the English NHS, the impact of the Spanish Flu epidemic on trust, and a study on social accountability programmes. There’s also a study showing that unobserved factors associated with paternal socioeconomic status partially explain an increased likelihood of obesity.
Volume 24, Issue 3
This issue of Value in Health includes a themed section on nutrition economics. Since reading How Not To Die – a Secret Santa gift from a colleague – I’ve become more interested in nutrition research. The section starts with an introductory article highlighting the value of collaboration across disciplines. The themed section includes a systematic review of economic evaluations of personalised nutrition. Of the 49 studies included, most focussed on diabetes or obesity and reported interventions as cost-effective relative to the control. A simulation study models the potential impact of the removal of food subsidies in the Canary Islands. It baffles me that any government would subsidise butter, cheese, and sugar, and this study demonstrates why. Based on price elasticity estimates, the authors estimated that around 4% of the population’s obesity is attributable to these subsidies. Finally, a longitudinal study of children in China explores the role of public health insurance. A difference-in-differences analysis demonstrates that introducing a new insurance scheme did not benefit children’s nutrition but rather led to greater health service use with out-of-pocket expenditures, reducing disposable income and calorific intake.
I was interested to read about a discrete choice experiment eliciting people’s preferences for digital mental health interventions. In particular, the researchers sought to understand why people might be less likely to use them compared with face-to-face interventions. Critical attributes included the potential to engage with a psychotherapist, the intervention’s effectiveness having been proven, and low price. The upshot is that it may make more sense to pursue a blended care model, incorporating face-to-face interaction with digital interventions.
This issue includes several applied evaluative studies. A discrete event simulation was used to evaluate the extension of screening intervals for abdominal aortic aneurysm (AAA) in the NHS in the UK. The findings suggest that extended intervals are likely to be cost-effective, at least for small AAAs. There’s also a cost-effectiveness analysis of the drug cemiplimab for advanced cutaneous squamous cell carcinoma, and an evaluation of a wellness programme with a commentary.
A few studies seek to inform HTA and reimbursement policy. There’s a comparison of added therapeutic value estimates by French and German HTA agencies and a study estimating the added value of ultra-expensive drugs in the US. There’s also a study estimating the impact of budget cap policies, outlining a set of principles for their implementation that seek to prevent harm to public health. A review of the use of a societal perspective in economic evaluations of rare diseases suggests that a change in perspective is not likely to change recommendations.
Several papers in this issue explore questions relating to health outcome measurement, including a Rasch analysis of an ICHOM standard set for cleft lip and palate, and a study using ‘goal attainment scaling’ in neurogenic bladder and bowel dysfunction.
There’s a review of the psychometric performance of preference-based measures for children, including 76 studies on four different instruments. In general, the evidence is quite limited, consisting of many small studies with narrow scope. While there’s now a substantial body of literature on the EQ-5D-Y, it mostly relates to the validity of the dimensions rather than index values used to estimate QALYs. Speaking of QALYs, this issue also includes a comparison of DALYs and QALYs in the evaluation of breast cancer screening; the two are not interchangeable.
Finally, there’s a nonsensical editorial criticising the EQ-5D, in which the authors claim it is useless because it does not generate ratio scale values, only ordinal values. The authors have failed to recognise the existence of interval scale properties, which EQ-5D index values do have. We wrote about it briefly in our Drop Dead paper. There is a very real sense in which an improvement from 0.2 to 0.4 is equivalent in value to an improvement from 0.4 to 0.6, relying on trade-offs with either time or risk, depending on your methodology.