Debellut, Frederic, Alena Friedrich, Ranju Baral, Clint Pecenka, Emmanuel Mugisha, and Kathleen M. Neuzil. 2024. "The Cost of Typhoid Illness in Low- and Middle-Income Countries, a Scoping Review of the Literature." PLOS ONE 19 (6): e0305692.
Abstract:
Typhoid fever is responsible for a substantial health burden in low- and middle-income countries (LMICs). New means of prevention became available with the prequalification of typhoid conjugate vaccines (TCV) by the World Health Organization (WHO) in 2018. Policymakers require evidence to inform decisions about TCV. The economic burden related to typhoid fever can be considerable, both for healthcare providers and households, and should be accounted for in the decision-making process. We aimed to understand the breadth of the evidence on the cost of typhoid fever by undertaking a scoping review of the published literature. We searched scientific databases with terms referring to typhoid fever cost of illness to identify published studies for the period January 1st 2000 to May 24th 2024. We also conferred with stakeholders engaged in typhoid research to identify studies pending completion or publication. We identified 13 published studies reporting empirical data for 11 countries, most of them located in Asia. The total cost of a typhoid episode ranged from $23 in India to $884 in Indonesia (current 2022 United States Dollar [USD]). Household expenditures related to typhoid fever were characterized as catastrophic in 9 studies. We identified 5 studies pending completion or publication, which will provide evidence for 9 countries, most of them located in Africa. Alignment in study characteristics and methods would increase the usefulness of the evidence generated and facilitate cross-country and regional comparison. The gap in evidence across regions should be mitigated when studies undertaken in African countries are published. There remains a lack of evidence on the cost to treat typhoid in the context of increasing antimicrobial resistance. Decision-makers should consider the available evidence on the economic burden of typhoid, particularly as risk factors related to antimicrobial resistance and climate change increase typhoid risk. Additional studies should address typhoid illness costs, using standardized methods and accounting for the costs of antimicrobial resistance.
en
Friedrich, Alena and Anna Chorniy
Abstract
As diagnoses of attention-deficit/hyperactivity disorder (ADHD) have risen sharply since the early 1990s, estimates of true prevalence differences between boys and girls have narrowed. Recent psychological research points to differences in risk factors and symptom presentation across boys and girls as potential explanations for the gender gap in diagnosis rates. Economists, meanwhile, have documented tangible consequences of ADHD treatment, including reductions in adverse health outcomes and improvements in both cognitive and non-cognitive academic outcomes. Given these stakes, understanding whether ADHD is diagnosed and treated differently by gender is an important open question. This paper asks whether pediatricians' diagnostic decisions vary systematically by patient gender, and uses subsequent educational outcomes to test whether any such variation reflects bias rather than true differences in underlying severity. Answering this question poses a distinctive empirical challenge, since diagnosis depends on meeting a threshold number of symptom criteria, but without a definitive clinical test, true ADHD status is inherently subjective and thus unobservable to the researcher. Because pediatricians are more likely to diagnose the most severe cases, a simple comparison of outcomes across diagnostic status conflates the effect of treatment with underlying severity. To isolate diagnostic bias from true severity, we adapt the examiner leniency design, first popularized by Kling (2006), to the setting of pediatric ADHD diagnosis. (Findings here)
Friedrich, Alena
Abstract
While traditional sorting models tie redistributive school funding reforms to changes in the implicit price of school quality, few papers empirically test how such policy shifts affect population changes on a national level. Although most basic models predict that enrollments should increase alongside increases in per-pupil revenues in the lowest-income districts, my results suggest a causal link between funding equalization and larger disparities in student populations between districts. Holding baseline income levels constant and using reform-induced changes to instrument for per-pupil revenues, a two-stage least squares analysis suggests that these changes are driven primarily by mobility of white, relatively wealthy residents of low-income districts into higher-income neighborhoods. Analysis of heterogeneity in treatment effects by reform type suggests that these demographic changes were only significant in states that sought to redistribute revenues without increasing baseline average funding levels. Although the potential mechanisms driving these changes are not entirely clear, the results of this paper motivate further investigation of the unintended consequences of school funding reforms and highlight the importance of considering household mobility when implementing redistributive policies.
Diagnostic Criteria Changes and Gender Differences in ADHD Prevalence
Friedrich, Alena
Abstract
In May of 2013, the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) updated the ADHD criteria, raising the required age of symptom onset from before age 7 to before age 12. Because this revision allows symptoms that first become apparent later in childhood to satisfy the onset criterion, it motivates particular investigation into the update's impact on diagnosed ADHD prevalence by gender: girls are often diagnosed later than boys, as their earlier symptoms more frequently go unnoticed. By comparing the pre- and post-2013 change in gender-specific diagnosis rates between 8--11 year olds and children under 7, this paper attempts to identify the impact of the DSM change on diagnosis and treatment rates in this population. If a significant effect is found, this variation could also be used to investigate the change's impact on outcomes for the treatment group.