Program on U.S. Health Policy
The Program on U.S. Health Policy (USHP), launched by CHW in early 2012, sponsors research and activities addressing aspects of domestic health care and health policy. View our most recent call for faculty research proposals, and explore our active projects below.
Does Pharmaceutical Innovation Improve Outcomes? The Case of Medication for ADHDPIs: Anna Chorniy & Janet Currie, Princeton University; Lyudmyla Sonchak, SUNY Oswego
In 2012 there were five million U.S. children who had been diagnosed with attention-deficit/hyperactivity disorder (ADHD), a 20% increase from 2003. The Medicaid bill for prescription ADHD medications in thirty U.S. states in our data increased from $103 million in 1999 to $666 million in 2009, a 548% increase in real terms. Drug innovation is one of the drivers of increasing costs. While the newer drugs are certainly more expensive than their older counterparts, it is not yet clear if newer medications are more effective in improving patient health and/or whether they reduce spending on other types of medical care, by, for example preventing Emergency room visits. Nor is it clear whether new medications helped fuel the increases in the caseload. In the proposed study, we will first ask whether pharmaceutical innovation has affected the number of Medicaid children receiving ADHD medications, and then we will ask whether it led to improved outcomes among children and teens diagnosed with the condition.
Epidemiology of Antibiotic Use in U.S. Hospitals*PIs: Ramanan Laxminarayan and Simon Levin, Princeton University
Endemic antibiotic-resistant bacteria (ARB) in hospitals are a major challenge to in-patient health and significantly increase the costs of hospital stays. Antibiotic selection pressure is a significant determinant of prevalence of ARB, and previous studies have shown that antibiotics are prescribed heavily and often inappropriately in acute-care hospitals (Craig, Uman et al. 1978, Elhanan, Sarhat et al. 1997, Gorecki, Schein et al. 1999). Unnecessary use of antimicrobials contributes to the transmission of antimicrobial-resistant nosocomial pathogens in part by eliminating normal anaerobic bacterial flora that inhibit overgrowth of pathogenic microorganisms. The overuse and misuse of antimicrobials, combined with the growing challenges posed by antimicrobial-resistant organisms, have led to calls for better antimicrobial therapy in the in-patient setting.
Though various interventions to improve antimicrobial use have been developed, the implementation of these interventions has been hampered by a lack of data on the “epidemiology of antimicrobial use”—that is, how clinicians make decisions to start, stop, and change antimicrobial therapy for in-patients. A deeper understanding of what types of clinical syndromes tend to prompt antimicrobial therapy and what information then prompts changes in that therapy will be critical to ongoing efforts at the US Centers for Disease Control and Prevention (CDC) to improve in-patient antimicrobial use. For example, in the past 40 years no multicenter studies have examined the most common clinical syndromes for which antimicrobials are prescribed in the in-patient setting (Scheckler and Bennett 1970). Knowing which syndromes account for the most frequent use could inform the development of improved diagnostics and treatment algorithms to optimize antimicrobial therapy for these syndromes. Likewise, there is no information from multicenter studies on how frequently antibiotics are stopped when microbiologic studies show no sign of infection. Knowing how commonly this occurs and the factors that are associated with decisions to stop or continue antimicrobials in the setting of negative cultures could inform interventions to help improve these decisions in health care facilities.
In this study, we aim to characterize the general epidemiology of antibiotic use in the selected hospitals. The research will rely on an existing effort, funded by the CDC that uses full chart abstractions to capture detailed information on antimicrobial prescriptions, patient’s clinical information at the time of the prescription, subsequent changes to the antimicrobial therapy, and clinical information at the time antimicrobial prescription changes are made in hospitals. We have collected data from six acute-care facilities that vary with respect to location, size, type, and presence of antibiotic stewardship programs. We propose to use these existing chart-review data, and collect additional data on the performance of antibiotic stewardship programs from the six facilities to characterize the general epidemiology of antibiotic use in the selected hospitals.
Exploring Small Area Variations in Children’s Health Care
PI: Janet Currie, Princeton University
One of the “stylized facts” about American medicine is that some areas spend much more on health care without achieving noticeably better results in terms of reductions in morbidity and mortality. Reining in such “excess” spending is an important goal of health care reform, yet we have little understanding of the reasons these variations exist. We do not even know how common they are in the non-elderly population, since virtually all of this research has been conducted using data on the elderly. It is possible that these patterns are quite different in the elderly and non-elderly since the elderly all have relatively generous health insurance coverage under the Medicare program. This research will first investigate small area variations in health care for children, and in subsequent health outcomes. The research will then investigate reasons that have been suggested for these variations including financial incentives facing providers; differences in patient populations; differences in legal regimes; spillovers from one provider to another in terms of practice style; and learning by individual practitioners.
Impact of Meningcoccocal B Vaccination during the 2013 Princeton University Outbreak*
PI: Nicole E. Basta, University of Minnesota
Meningococcal disease outbreaks caused by the bacterial pathogen Neisseria meningitides are a significant public health threat whenever they occur due to the severity of the disease and the high risk of mortality. Serogroup B outbreaks are on the rise, and how best to control these outbreaks is a key health policy question. We propose to investigate the impact of the introduction of the 4CMenB vaccine, not yet licensed in the US but imported, on the 2013 Princeton University outbreak. By assessing the level of immunity induced by the vaccine against the outbreak strain, we will be able to determine how well protected vaccinated students are compared to students who chose not to be vaccinated and to estimate the extent of protective immunity in the student population. The results of this investigation will provide insight into the impact of the vaccination campaign and provide evidence to inform future policy decisions.
Improving Maternity Care in New Jersey: How Do Hospitals Respond to Policy Mandates to Reduce Cesarean Section Rates and Improve Breastfeeding Rates?
PI: Elizabeth Armstrong, Princeton University
This research proposes to investigate how the 50-some maternity hospitals in New Jersey are responding to maternity care policy mandates to reduce cesarean rates and increasing breastfeeding rates. Cesarean rates and breastfeeding rates are complex, multifactorial outcomes that reflect provider practices, institutional protocols, patient preferences, sociodemographic characteristics of the patient population, and cultural norms. The research project will analyze hospital-level and state-wide data, and conduct interviews with a range of actors from a sample of New Jersey hospitals, to investigate the following questions: How do providers and hospital administrators seek to achieve mandates to reduce cesareans and increase breastfeeding? What do they see as the most important determinants of these outcomes? How do they prioritize institutional responses? And what is the effect of these policy shifts on actual clinical practice as well as patient outcomes?
Long-run Effects of Privatizing Public Health InsurancePI: Ilyana Kuziemko, Princeton University
Since the 1980s, health care costs and income inequality have risen more quickly in the US than in other developed countries. Past work suggests that one policy response to rising health-care costs--privatization of public insurance programs such as Medicaid and Medicare--may have exacerbated health inequality by rationing care to those in poor health. Another important strand of literature documents strong links between health and economic outcomes.
This project asks whether more direct links can be drawn between privatization and later-life inequality. Specifically, I examine whether the privatization of Medicaid—which has been shown to exacerbated health inequalities among infants--affects later-life disparities in elementary- and secondary-school outcomes.
Mapping Access to Hospital Care: Exploratory Research on the Drivers and Consequences of Hospital Closures in New Jersey (1990 to the Present)PI: Alecia McGregor, Princeton University
Over the last 20 years, the United States has seen a rapid increase in hospital closures. Recent national studies show that for vulnerable, inner-city populations, access to hospital-based services has been disproportionately affected by closures. Although closures in low-income cities (e.g. Trenton, Newark, and Plainfield) generated enormous controversy, little is known on the impact of these changes on hospital availability, or on the political drivers of variation in which New Jersey hospitals closed.
I will conduct exploratory research from November 2015 through August 2016, with two objectives: 1) To map the spatial changes in hospital availability in New Jersey at three periods—1990, 2000, and 2010—to generate hypotheses on which populations are most affected, and; 2) To examine the political, economic and health systems-related drivers of closures through an analysis of the Certificate of Need process. Study results will further clarify the drivers and consequences of shifts in New Jersey’s fast-changing landscape of hospital care.
mDiary Study of Adolescent Relationships
We will expand on a successful pilot by testing the scalability of a relationship diary study (mDiary) administered using smartphones. Specifically, we will administer bi-weekly mobile relationship diaries over a 12-month period to 250 youth in middle adolescence that have participated in the Fragile Families and Child Wellbeing Study (FFCWS) since birth. FFCWS is a unique panel study that represents births in large to medium cities, with oversamples of unmarried parents, and is especially useful for studying minority and disadvantaged populations. Pairing the FFCWS and mDiary data offers a unique opportunity to study the childhood and adolescent precursors to healthy and unhealthy partnering behaviors, and consequences for adolescent reproductive and emotional health.
The scientific value of the proposed diary study transcends our specific focus on adolescent relationships because of the growing interest in using mobile technologies for health research and interventions across a range of outcomes.
Making Health Insurance Work: Data-Driven Incentive Alignment
PIs: Mark Braverman and Sylvain Chassang, Princeton University
Capitation schemes, in which plans are paid expected health care costs for the people who join them, are a key component of health care insurance policies, but existing schemes are subject to gaming by health plans seeking to attract patients with lower than expected cost of care. In the past ill designed capitation schemes have contributed to the failure of health insurance exchanges in Texas (1999) and California (2006). This research program, at the intersection of Economics and Computer Science, studies the value of large exhaustive data sets of patient characteristics in resolving this issue. It proposes novel uses for such data that go beyond standard data mining applications, and instead focuses on correct incentive provision.
Modeling the Impact of Pre-Exposure Prophylaxis and Targeted Hepatitis C Case-finding and Treatment on the HIV and HCV epidemics in Newark, NJ*PI: Bryan Grenfell, Princeton University
A key issue for US health policy is prioritizing the various approaches toward addressing the ongoing domestic HIV epidemic. While national prevalence is low on a global scale (<1%), prevalence in certain impoverished urban areas remains alarmingly high. Key at-risk populations in these high-prevalence hotspots may suffer additional morbidity from opportunistic infections such as Hepatitis C and tuberculosis. The research aim of this two-year proposal is to develop and deploy a novel mathematical model to assess the impact of a range of treatment and prevention interventions on the HIV and Hepatitis C (HCV) epidemics in Newark, NJ, one of these high-prevalence hotspots. This work will also be leveraged to provide significant undergraduate training opportunities in this important area of US health. The first stage of the project investigating interventions for HIV mono-infection along the Test-and-Treat pathway has already been piloted, and subsequent major refinements will form the main body of the grant. These extensions involve planning and implementing new models that take into account HCV transmission as well as a possible intervention in the form of pre-exposure prophylaxis (PrEP). Detailed research is necessary to carry out assessment of the impact of PrEP and Hepatitis C treatment. Results of this modeling work will be used to inform economic models to generate cost-effectiveness assessments and to help shape local and national control policy. We will collaborate with researchers at Princeton and the University of Medicine and Dentistry of New Jersey in Newark, NJ.
Policy and Health Implications of Novel Genetic Variants in the Fragile Families and Child Wellbeing Study
A rapidly emerging literature shows that controlling for genetic heterogeneity can improve estimates of the effects of the social and economic environment on health and wellbeing. This new information has already produced novel perspectives on policy outcomes, and has the potential to greatly enhance the efficiency and precision of policy making.
To date, virtually all of this literature has been based on two types of genetic information – single nucleotide polymorphisms (SNPs) and variable number tandem repeats (VNTRs). This project will use genetic data from the Fragile Families and Child Wellbeing Study (approximately 6,000 child and mother DNA samples) to conduct pilot work on an additional source of genetic information – copy number variants (CNVs). Although CNVs are the largest source of human genetic variation, this information has not been incorporated into studies by social scientists. The potential value of the proposed research to the community of scientists who study links between human genetic variation and social and policy-related measures and outcomes is extremely high.