The
NEW ENGLA ND JOURNAL
of
MEDICINE
Perspective november 14, 2013
The National Children’s Study — A Proposed Plan Alan E. Guttmacher, M.D., Steven Hirschfeld, M.D., Ph.D., and Francis S. Collins, M.D., Ph.D.
T
he National Children’s Study (NCS) is a longitudinal, observational, birth-cohort study that will seek to identify the foundations of healthy adulthood by examining the effects of a broad
range of environmental influences and biologic factors on children’s health, growth, and development. Its authorizing legislation, the Children’s Health Act of 2000, directs the NCS to “(1) incorporate behavioral, emotional, educational, and contextual consequences to enable a complete assessment of the physical, chemical, biological, and psychosocial environmental influences on children’s well-being; (2) gather data on environmental influences and outcomes on diverse populations of children, which may include the consideration of prenatal exposures; and (3) consider health disparities among children, which may include the consideration of prenatal exposures.”1 The current NCS plan is to enroll, at or before birth, 100,000 children who are representative
of the U.S. population and to gather, through 21 years of age, data regarding biologic makeup, environmental exposures, and growth, development, and health, along with environmental and biologic samples.2 Other U.S.-based longitudinal studies provide information about children’s health, growth, or development. None of them, however, are as large as the NCS, entail collecting such detailed biologic and environmental data and samples, or include longitudinal phenotyping from before birth through the age of 21. To maximize data interoperability, the NCS is being coordinated with similar studies in other countries (including France, Japan, Britain, and Canada), but none of those studies will examine the same populations or environmental factors as the NCS.
With its coordinated longi tudinal biologic, environmentalexposure, and phenotypic data and samples, the NCS will provide an important resource for understanding children’s growth, development, and health. Early in the design process, a core set of scientific hypotheses was developed to define the study. However, since the study’s developers believe that its scope should be limited only by scientific creativity and not by current consensus priorities, exemplar hypotheses (see box) now inform, rather than define, the study’s design. The NCS is committed to broad, rapid sharing of all data and samples, while respecting participants’ privacy and confidentiality. No individuals or institutions that gather data and samples will have prioritized claims to them. Electronic data will be available to all qualified researchers through controlledaccess mechanisms, in keeping with current National Institutes of Health practices. Because bio-
n engl j med 369;20 nejm.org november 14, 2013
The New England Journal of Medicine Downloaded from nejm.org on August 12, 2015. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.
1873
PERS PE C T IV E
The National Children’s Study
Five Exemplar Hypotheses. Dietary protein intake influences deciduous-tooth composition Kitchen-dust analysis predicts onset and severity of reactive airway disease Pesticide concentrations in urine correlate with body-mass index in an age-dependent manner Cytokine profiles of cord blood predict incidence of ear infections by the age of 2 Paternal bedtime reading strengthens executive function in preschool children
logic and environmental samples are exhaustible, there will be an application process for obtaining them. To maximize their use, the NCS will share promptly with the entire research community the results of all analyses performed. Rather than classifying participants into predetermined disease categories, the NCS will collect primary observations and event data to enable researchers to apply their own criteria in defining cases. For example, rather than creating a category of children with reactive airway disease, NCS researchers will accurately capture medical histories and information on participants’ experiences and respiratory symptoms, as well as biospecimens, genetic analyses, and environmental samples. Researchers will be able to use these data in conjunction with the case definitions and classifications that they deem most appropriate for analyses. The NCS has been using a pilot study, or Vanguard Study, to fieldtest both scientific and logistic elements before scaling up for the Main Study. Forty Vanguard Study sites were launched between 2009 and mid-2011. When the contracts for the Vanguard Study sites became eligible for recompetition in 2012, the number of contractors was reduced substantially, to improve scientific comparability and increase efficiency. Thus far, the Vanguard Study has focused primarily on testing recruitment and enrollment strategies, but researchers will go on to collect exposure, 1874
biologic, and phenotypic data and samples from nearly 5000 children until they reach 21 years of age. The Vanguard Study field-tested three recruitment strategies: household-based, direct outreach, and provider-based. Provider-based recruitment proved the most efficient, in terms of the number of women contacted per newborn enrolled and the cost per newborn enrolled (see Table 1 in the Supplementary Appendix, available with the full text of this article at NEJM.org). The proposed Main Study sample consists of 100,000 liveborn children, approximately 90,000 of whom would be born to members of a national probability sample of women enrolled at various stages of pregnancy. The probability sample would have two strata: 45,000 women recruited at hospitals and birthing centers (the primary sampling units), and 45,000 pregnant women seen by prenatal care providers who refer patients to the same hospitals and birthing centers. The remaining 10,000 children would come from recruitment of populations of particular scientific interest, such as a cohort recruited before conception by targeting nulliparous women, women without access to health care, populations affected by health disparities, or those with exposures to scientifically relevant geographic, economic, or industrial conditions. Enrollment in the Main Study is projected to take 4 years. Since many exposures and outcomes of
interest occur early in childhood, early retention is crucial. With 99% annual retention, the sample would retain 80,000 participants to 21 years of age; with 97% annual retention, 54,000 would be retained throughout. Data collection will focus on key periods of growth and development. In women enrolled sufficiently early, data would be collected before 20 weeks of gestation and once more during pregnancy. Data collection would be most intensive in early childhood, with 7 of 13 early-childhood visits being face-to-face encounters, permitting collection of biospecimens and environmental samples. The other 6 visits would involve remote data collection, typically by telephone. Subsequent data-collection interactions are planned every 2 years through the age of 21. The plan is thus to collect data at 21 points prenatally and postnatally, but that may change on the basis of experience in the Vanguard Study, scientific opportunities, logistic factors, and available resources. The timing of visits will be allowed to vary within a window of several weeks around a particular age, to improve compliance and capture data across a range of ages. A core questionnaire will be administered at every visit, with supplemental modules for specific participants or subpopulations depending on age, developmental stage, specific exposures, and other variables. This approach should permit collection of pertinent information while reducing costs and the burden on participants. The NCS will also use such methods as diaries, interviews, and abstracting of medical records to collect data regarding medical events and health care interactions, as well as regarding noise exposure, geographic movements,
n engl j med 369;20 nejm.org november 14, 2013
The New England Journal of Medicine Downloaded from nejm.org on August 12, 2015. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.
PE R S PE C T IV E
The National Children’s Study
social interactions and networks, behavior, cultural characteristics, and lifestyle. In addition, phy sical measurements, interviews, questionnaires, and images will be used to assess social behavior, neurodevelopment, physical growth, and family mental health and dynamics. Although the Vanguard Study will determine the precise environmental assessments to be used, the current plan involves approaches including collection of biospecimens (e.g., blood, urine, saliva, skin swabs, cord blood and samples, placental tissue, and breast milk) and environmental samples (e.g., airborne particulate matter, dust, soil, and water), as well as community-based environmental data, to assess environmental exposures. The study will use observations of the interior and exterior of residences and neigh-
borhoods to identify sources of environmental contaminants and neighborhood characteristics. Questionnaires regarding household occupations, work commutes, lifestyle, hobbies, and daily routines will also be deployed. The final design for the initial stages of the Main Study is anticipated within a few months after the release of an analysis of the proposed Main Study design, due next summer, from the Institute of Medicine and the National Research Council. Initial contracts should be awarded in the first half of 2015, and the study launched several months later. Because of the importance of the fetal and early-childhood periods, the NCS should provide complex new information within a year or two after launch. Of course, as data and samples accumulate, the study’s value and impact should
grow commensurately. We expect its longitudinal collection of linked environmental, biologic, and phenotypic data and samples to provide important insights into health, growth, and development, not only of U.S. children but of people of all ages and countries. Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. From the Eunice Kennedy Shriver National Institute of Child Health and Human Development (A.E.G., S.H.) and the Office of the Director, National Institutes of Health (F.S.C.) — both in Bethesda, MD. 1. H.R. 4365. An act to amend the Public Health Service Act with respect to children’s health. January 24, 2000 (http://www.gpo .gov/fdsys/pkg/BILLS-106hr4365enr/pdf/ BILLS-106hr4365enr.pdf). 2. The National Children’s Study: Institute of Medicine workshop steering committee briefing document. October 16, 2012 (http:// www.nationalchildrensstudy.gov/research/ workshops/Pages/IOM-SC-white-paper -october-2012pdf.pdf). DOI: 10.1056/NEJMp1311150 Copyright © 2013 Massachusetts Medical Society.
Public Reporting, Consumerism, and Patient Empowerment Robert S. Huckman, Ph.D., and Mark A. Kelley, M.D.
S
everal forces in the United States — including the Affordable Care Act (ACA) of 2010 — have promoted greater public reporting of health care outcomes. By many accounts, this reporting is largely ignored by consumers (see graph),1 perhaps because the information is hard to find or difficult to understand. We propose another potential explanation — namely, that the public spotlight is not aimed at information that most patients value. Current public reports typically compare health care providers in terms of quality or cost to help consumers decide where or from whom to seek care. For example, patients in New York and Pennsylvania can view the cardiac-surgery outcomes for spe-
cific surgeons and hospitals. Such reporting assumes that patients have already decided to pursue cardiac surgery and are using this information simply to select the best provider. Unfortunately, this information does little to help patients decide whether they want or need surgery in the first place. Current public reports also tend to assume that patients can accurately interpret quality metrics. For example, what is the difference between a hospital with a 1% complication rate and another with a 2% rate? One perspective is that the first facility is twice as good as the second. An alternative view is that the absolute risk of a complication is so low at both institutions that choosing between them should hinge
on other factors, such as convenience, cost, and reputation. Patients may favor this latter interpretation more often than we imagine. Even some patients with education beyond high school have difficulty understanding basic statistics,2 so it’s not surprising that many of them view public reports as unhelpful. Rather than choosing between providers of a specific procedure, perhaps patients are seeking an answer to a more fundamental and personal question: “Is the proposed treatment or procedure the best option given my condition, my financial status, and my social or family situation?” Presenting and explaining this information to patients — as the first step in what is termed shared
n engl j med 369;20 nejm.org november 14, 2013
The New England Journal of Medicine Downloaded from nejm.org on August 12, 2015. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.
1875
NEW ENGLA ND JOURNAL
of
MEDICINE
Perspective november 14, 2013
The National Children’s Study — A Proposed Plan Alan E. Guttmacher, M.D., Steven Hirschfeld, M.D., Ph.D., and Francis S. Collins, M.D., Ph.D.
T
he National Children’s Study (NCS) is a longitudinal, observational, birth-cohort study that will seek to identify the foundations of healthy adulthood by examining the effects of a broad
range of environmental influences and biologic factors on children’s health, growth, and development. Its authorizing legislation, the Children’s Health Act of 2000, directs the NCS to “(1) incorporate behavioral, emotional, educational, and contextual consequences to enable a complete assessment of the physical, chemical, biological, and psychosocial environmental influences on children’s well-being; (2) gather data on environmental influences and outcomes on diverse populations of children, which may include the consideration of prenatal exposures; and (3) consider health disparities among children, which may include the consideration of prenatal exposures.”1 The current NCS plan is to enroll, at or before birth, 100,000 children who are representative
of the U.S. population and to gather, through 21 years of age, data regarding biologic makeup, environmental exposures, and growth, development, and health, along with environmental and biologic samples.2 Other U.S.-based longitudinal studies provide information about children’s health, growth, or development. None of them, however, are as large as the NCS, entail collecting such detailed biologic and environmental data and samples, or include longitudinal phenotyping from before birth through the age of 21. To maximize data interoperability, the NCS is being coordinated with similar studies in other countries (including France, Japan, Britain, and Canada), but none of those studies will examine the same populations or environmental factors as the NCS.
With its coordinated longi tudinal biologic, environmentalexposure, and phenotypic data and samples, the NCS will provide an important resource for understanding children’s growth, development, and health. Early in the design process, a core set of scientific hypotheses was developed to define the study. However, since the study’s developers believe that its scope should be limited only by scientific creativity and not by current consensus priorities, exemplar hypotheses (see box) now inform, rather than define, the study’s design. The NCS is committed to broad, rapid sharing of all data and samples, while respecting participants’ privacy and confidentiality. No individuals or institutions that gather data and samples will have prioritized claims to them. Electronic data will be available to all qualified researchers through controlledaccess mechanisms, in keeping with current National Institutes of Health practices. Because bio-
n engl j med 369;20 nejm.org november 14, 2013
The New England Journal of Medicine Downloaded from nejm.org on August 12, 2015. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.
1873
PERS PE C T IV E
The National Children’s Study
Five Exemplar Hypotheses. Dietary protein intake influences deciduous-tooth composition Kitchen-dust analysis predicts onset and severity of reactive airway disease Pesticide concentrations in urine correlate with body-mass index in an age-dependent manner Cytokine profiles of cord blood predict incidence of ear infections by the age of 2 Paternal bedtime reading strengthens executive function in preschool children
logic and environmental samples are exhaustible, there will be an application process for obtaining them. To maximize their use, the NCS will share promptly with the entire research community the results of all analyses performed. Rather than classifying participants into predetermined disease categories, the NCS will collect primary observations and event data to enable researchers to apply their own criteria in defining cases. For example, rather than creating a category of children with reactive airway disease, NCS researchers will accurately capture medical histories and information on participants’ experiences and respiratory symptoms, as well as biospecimens, genetic analyses, and environmental samples. Researchers will be able to use these data in conjunction with the case definitions and classifications that they deem most appropriate for analyses. The NCS has been using a pilot study, or Vanguard Study, to fieldtest both scientific and logistic elements before scaling up for the Main Study. Forty Vanguard Study sites were launched between 2009 and mid-2011. When the contracts for the Vanguard Study sites became eligible for recompetition in 2012, the number of contractors was reduced substantially, to improve scientific comparability and increase efficiency. Thus far, the Vanguard Study has focused primarily on testing recruitment and enrollment strategies, but researchers will go on to collect exposure, 1874
biologic, and phenotypic data and samples from nearly 5000 children until they reach 21 years of age. The Vanguard Study field-tested three recruitment strategies: household-based, direct outreach, and provider-based. Provider-based recruitment proved the most efficient, in terms of the number of women contacted per newborn enrolled and the cost per newborn enrolled (see Table 1 in the Supplementary Appendix, available with the full text of this article at NEJM.org). The proposed Main Study sample consists of 100,000 liveborn children, approximately 90,000 of whom would be born to members of a national probability sample of women enrolled at various stages of pregnancy. The probability sample would have two strata: 45,000 women recruited at hospitals and birthing centers (the primary sampling units), and 45,000 pregnant women seen by prenatal care providers who refer patients to the same hospitals and birthing centers. The remaining 10,000 children would come from recruitment of populations of particular scientific interest, such as a cohort recruited before conception by targeting nulliparous women, women without access to health care, populations affected by health disparities, or those with exposures to scientifically relevant geographic, economic, or industrial conditions. Enrollment in the Main Study is projected to take 4 years. Since many exposures and outcomes of
interest occur early in childhood, early retention is crucial. With 99% annual retention, the sample would retain 80,000 participants to 21 years of age; with 97% annual retention, 54,000 would be retained throughout. Data collection will focus on key periods of growth and development. In women enrolled sufficiently early, data would be collected before 20 weeks of gestation and once more during pregnancy. Data collection would be most intensive in early childhood, with 7 of 13 early-childhood visits being face-to-face encounters, permitting collection of biospecimens and environmental samples. The other 6 visits would involve remote data collection, typically by telephone. Subsequent data-collection interactions are planned every 2 years through the age of 21. The plan is thus to collect data at 21 points prenatally and postnatally, but that may change on the basis of experience in the Vanguard Study, scientific opportunities, logistic factors, and available resources. The timing of visits will be allowed to vary within a window of several weeks around a particular age, to improve compliance and capture data across a range of ages. A core questionnaire will be administered at every visit, with supplemental modules for specific participants or subpopulations depending on age, developmental stage, specific exposures, and other variables. This approach should permit collection of pertinent information while reducing costs and the burden on participants. The NCS will also use such methods as diaries, interviews, and abstracting of medical records to collect data regarding medical events and health care interactions, as well as regarding noise exposure, geographic movements,
n engl j med 369;20 nejm.org november 14, 2013
The New England Journal of Medicine Downloaded from nejm.org on August 12, 2015. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.
PE R S PE C T IV E
The National Children’s Study
social interactions and networks, behavior, cultural characteristics, and lifestyle. In addition, phy sical measurements, interviews, questionnaires, and images will be used to assess social behavior, neurodevelopment, physical growth, and family mental health and dynamics. Although the Vanguard Study will determine the precise environmental assessments to be used, the current plan involves approaches including collection of biospecimens (e.g., blood, urine, saliva, skin swabs, cord blood and samples, placental tissue, and breast milk) and environmental samples (e.g., airborne particulate matter, dust, soil, and water), as well as community-based environmental data, to assess environmental exposures. The study will use observations of the interior and exterior of residences and neigh-
borhoods to identify sources of environmental contaminants and neighborhood characteristics. Questionnaires regarding household occupations, work commutes, lifestyle, hobbies, and daily routines will also be deployed. The final design for the initial stages of the Main Study is anticipated within a few months after the release of an analysis of the proposed Main Study design, due next summer, from the Institute of Medicine and the National Research Council. Initial contracts should be awarded in the first half of 2015, and the study launched several months later. Because of the importance of the fetal and early-childhood periods, the NCS should provide complex new information within a year or two after launch. Of course, as data and samples accumulate, the study’s value and impact should
grow commensurately. We expect its longitudinal collection of linked environmental, biologic, and phenotypic data and samples to provide important insights into health, growth, and development, not only of U.S. children but of people of all ages and countries. Disclosure forms provided by the authors are available with the full text of this article at NEJM.org. From the Eunice Kennedy Shriver National Institute of Child Health and Human Development (A.E.G., S.H.) and the Office of the Director, National Institutes of Health (F.S.C.) — both in Bethesda, MD. 1. H.R. 4365. An act to amend the Public Health Service Act with respect to children’s health. January 24, 2000 (http://www.gpo .gov/fdsys/pkg/BILLS-106hr4365enr/pdf/ BILLS-106hr4365enr.pdf). 2. The National Children’s Study: Institute of Medicine workshop steering committee briefing document. October 16, 2012 (http:// www.nationalchildrensstudy.gov/research/ workshops/Pages/IOM-SC-white-paper -october-2012pdf.pdf). DOI: 10.1056/NEJMp1311150 Copyright © 2013 Massachusetts Medical Society.
Public Reporting, Consumerism, and Patient Empowerment Robert S. Huckman, Ph.D., and Mark A. Kelley, M.D.
S
everal forces in the United States — including the Affordable Care Act (ACA) of 2010 — have promoted greater public reporting of health care outcomes. By many accounts, this reporting is largely ignored by consumers (see graph),1 perhaps because the information is hard to find or difficult to understand. We propose another potential explanation — namely, that the public spotlight is not aimed at information that most patients value. Current public reports typically compare health care providers in terms of quality or cost to help consumers decide where or from whom to seek care. For example, patients in New York and Pennsylvania can view the cardiac-surgery outcomes for spe-
cific surgeons and hospitals. Such reporting assumes that patients have already decided to pursue cardiac surgery and are using this information simply to select the best provider. Unfortunately, this information does little to help patients decide whether they want or need surgery in the first place. Current public reports also tend to assume that patients can accurately interpret quality metrics. For example, what is the difference between a hospital with a 1% complication rate and another with a 2% rate? One perspective is that the first facility is twice as good as the second. An alternative view is that the absolute risk of a complication is so low at both institutions that choosing between them should hinge
on other factors, such as convenience, cost, and reputation. Patients may favor this latter interpretation more often than we imagine. Even some patients with education beyond high school have difficulty understanding basic statistics,2 so it’s not surprising that many of them view public reports as unhelpful. Rather than choosing between providers of a specific procedure, perhaps patients are seeking an answer to a more fundamental and personal question: “Is the proposed treatment or procedure the best option given my condition, my financial status, and my social or family situation?” Presenting and explaining this information to patients — as the first step in what is termed shared
n engl j med 369;20 nejm.org november 14, 2013
The New England Journal of Medicine Downloaded from nejm.org on August 12, 2015. For personal use only. No other uses without permission. Copyright © 2013 Massachusetts Medical Society. All rights reserved.
1875
