In the final panel session, speakers discussed policy issues that can impact the health–education collaboration to improve educational outcomes. Rochelle Davis of the Healthy Schools Campaign discussed the use of education metrics by the health sector to better align the two sectors to support children’s health. Denise Chrysler of the Network for Public Health Law at the University of Michigan School of Public Health explained how the Family Educational Rights and Privacy Act (FERPA) impacts data sharing by schools. Kelly Hall of Pembroke Policy Consulting described health care payment options that could be leveraged to impact educational outcomes. Kent McGuire of the William and Flora Hewlett Foundation discussed the challenges of measuring both the in-school and out-of-school factors that contribute to educational attainment. The session was moderated by Marc Gourevitch of the New York University Langone Health. (Highlights of this session are presented in Box 5-1.)
The Healthy Schools Campaign
The Healthy Schools Campaign advocates for policies and practices that support schools in creating conditions for student health and school wellness so that all students are in school, healthy, and ready to learn, Davis said. The intent is to leverage the power of school health to address the growing health and academic disparities that exist in the United
States. School health is about supporting the mental and physical needs of students, providing a supportive and healthy environment, and teaching students how to be healthy so they can become productive adults.
The Healthy Schools Campaign takes policy- and systems-level approaches to change, she said, and metrics and data sharing are critical to long-term sustainable change. The campaign works at the local level in schools and communities across Chicago to empower parents, principals, teachers, and others to integrate health and wellness into every aspect of the school experience. Davis said that lessons learned from this work have been applied to address policies and system-level barriers at the district, community, state, and national levels.
Through its work in Chicago, the Healthy Schools Campaign recognized the importance of integrating health metrics into the education accountability system. In 2010, while the Healthy Schools Campaign was working with about 30 schools to meet the requirements of the U.S. Department of Agriculture’s HealthierUS School Challenge, it became clear that schools were only focused on measures that were part of their
publicly available annual school progress report and that they were reluctant to take on any project that did not help them meet the requirements of the progress report. To address this, the Healthy Schools Campaign asked the school district to include a metric about the HealthierUS School Challenge on the progress report, which it did. The response from schools was immediate, Davis said, and within 2 years, more than 300 Chicago schools met the HealthierUS School Challenge. This successful change of practice was the groundwork for future policy work. The Chicago public school system has one of the strongest wellness policies in the country, she said. The school system now has an Office of Student Health and Wellness and a chief health officer.
At the national level, the Healthy Schools Project partnered with Trust for America’s Health and others to bring the concept of incorporating health metrics into education accountability systems to a broader audience. A Metrics Working Group of educators, data experts, and others was co-convened through the National Collaborative on Education and Health. One of the key conclusions of the working group was that chronic absenteeism is a strong health proxy metric to include in education accountability systems.
Chronic absenteeism is designed to track the time that a student is not in the classroom. Unlike traditional attendance data, chronic absenteeism includes excused, unexcused, and suspension absences. As discussed by Basch and others, health barriers are often the cause of chronic absenteeism (e.g., asthma, oral health issues, violence and trauma, food insecurity, mental health, teen pregnancy, vision and hearing deficits). Chronic absenteeism is gaining traction as a metric within the education sector, Davis said, and the U.S. Department of Education’s (ED’s) Office for Civil Rights now collects chronic absenteeism data every other year and makes them publicly available. Starting in 2018, chronic absenteeism is now a requirement on every school progress report, as per the Every Student Succeeds Act (ESSA). In addition, 36 states and the District of Columbia have incorporated chronic absenteeism into their state accountability systems.
The Role of Education Metrics in Advancing the Goals of the Health Sector
Many of the examples discussed thus far at the workshop confirm the importance of metrics and data sharing across sectors, Davis said. The Healthy Schools Campaign developed a better understanding of the role that education metrics could play in advancing the goals of the health sectors through its work supporting 151 state teams of health and
1 At the workshop, it was incorrectly stated that there were 50 state teams.
education state staff to expand Medicaid-funded school health services. As the health sector has increased its focus on prevention, population health, care coordination, and the social determinants of health, the issue of education metrics has generated more attention, she said. To expand its understanding of how the health sector is using education metrics, the Healthy Schools Campaign commissioned in-depth interviews.2
The interviews revealed that education metrics are being used by the health sector. The most commonly used metrics were attendance, high school graduation rates, kindergarten readiness, third-grade reading, and measures of teacher wellness and turnover. These metrics are used more often for surveillance and prevention, Davis reported, and less often to support individual student outcomes. Interviewees reported significant barriers, but there was a strong interest in the topic and a strong understanding of the potential that data alignment between the sectors has for improving outcomes. Davis expanded on three key barriers that were identified in the interviews and shared recommendations that might facilitate action at the state and local levels (see Box 5-2).
Opportunity for Action: Every Student Succeeds Act and Chronic Absenteeism
The implementation of the ESSA presents an important opportunity to bring the health and education sectors into closer alignment through shared metrics, Davis said. Health disparities impact educational disparities, and leveraging the power of school health to support equity is an often overlooked approach, she said. National education policy in the United States is based on three layers of accountability: between the federal government and the states, between the states and the school districts, and between school districts and their communities. She observed that the current federal administration is taking a hands-off approach to trying to achieve equity and that the accountability structure3 is not strong at the moment.
The accountability structure between states and school districts is based on a set of standards that schools need to meet. They may use federal resources to conduct a needs assessment and implement programs and evaluations to help them meet the state’s accountability measures. The new opportunity that presents under the ESSA is that states may go beyond test
2 A copy of the draft report was provided to workshop participants and is available at https://healthyschoolscampaign.org/wp-content/uploads/2018/06/HSC-Education-Metrics-for-Health-Systems-Prelim-Report-DRAFT.pdf (accessed January 27, 2019).
3 Davis was referring to the accountability structure between the federal government and the states.
scores for accountability. This opens up the possibility for states to include health measures in their needs assessment and allows school districts to use federal funding to reduce chronic absenteeism, Davis said. The Healthy Schools Campaign has developed guidance for states to help them think about how to incorporate health into their needs assessment.4 There are several challenges, however. Davis noted that, for the 36 states and the District of Columbia that have included chronic absenteeism in their accountability measures, the metric is weighted relatively low (from 5 percent to 20 percent of a school’s score). In addition, some states are seeking to redefine what is a student, what is a school day, and what counts as being absent in ways that could mask and undercut any of these new metrics.
As mentioned, for the accountability structure between school districts and their communities, chronic absenteeism will be on every publicly reported school progress report in fall 2018. This provides an opportunity for the broader community, including those in the health sector, to understand what is happening in the school and to engage.
In Chicago, noted Davis, the Healthy Schools Campaign is working with the Advocate Health System to develop a new school service delivery model with four schools to meet the shared goal of reducing chronic absenteeism. The Advocate Health System found that the chronically absent students in the schools near their hospitals are the same children who present in their emergency department. This alignment of goals has created a very different framework for the school and the hospital to work together, Davis said.
Chrysler began by contrasting the perspective of an attorney with that of someone in the health sector. First, there has been much discussion about breaking down silos and sharing information within the health sector and across sectors. Attorneys, she said, favor keeping silos intact because they protect information that is identifiable to individuals. The numerous privacy laws in existence are concerned with where the data are coming from, where they are going to, and the purpose of sharing them.
Next, the health sector is looking to build relationships, while attorneys are focused on defining relationships. How the law applies depends on the relationships among the parties, she explained. For example,
4 This guidance was provided to workshop attendees and is available at https://healthyschoolscampaign.org/wp-content/uploads/2017/05/Framework-for-Action-Addressing-Chronic-Absenteeism-through-ESSA-Implementation.pdf (accessed August 22, 2018).
whether the FERPA applies to school-based clinics depends on who is running the clinic and if that organization is under the control of the school. If the local health department runs the clinic and operates it independently of the school district, the FERPA does not apply, she said.
Finally, the way the health sector and attorneys perceive risk is different. Laws are often contradictory, silent, or ambiguous, Chrysler said, and usually require interpretation and professional judgment. If there is any doubt about taking an action, an attorney will advise against the action because they think of risk in terms of protecting against liability. Attorneys see risk in sharing data, while the health sector sees risk in not sharing data, she said.
Chrysler advised those in the health sector to develop an ongoing relationship with their organization’s attorneys. Help them to understand the context of potential action and ask for their help in achieving it, she said.
Family Educational Rights and Privacy Act Overview
The FERPA is a federal law (20 U.S.C. § 1232g; 34 CFR Part 99) that protects the privacy of students’ education records, grants parents and adult students certain rights regarding education records, and applies to schools and educational programs that receive ED funds. The FERPA limits what schools can share with others, but it does not limit what public health or health care providers can disclose to schools, Chrysler said. Education records are anything directly related to the student that is maintained by the school or a party acting on behalf of the school (e.g., employees, contractors). This includes student transcripts, disciplinary records, immunization records, and other medical or health-related records that schools have.
Chrysler explained that the FERPA does not necessarily apply just because services are provided on school premises. School-based health clinics that are not under the control of the school district are not subject to the FERPA (e.g., a public health nurse provides health services to students on school grounds). The FERPA does not apply to university student health centers when information is exchanged among health care professionals specifically for providing treatment to students. She added that the FERPA would apply if student data were provided by a university student health center to a public health department (as public health is not a treatment provider).
Data Disclosure Under the Family Educational Rights and Privacy Act
The FERPA does allow disclosure of de-identified data, data shared with written consent, and data shared under a FERPA exception, Chrysler
said. Under the FERPA, de-identifying records involves removing personally identifiable information, including names, addresses, personal identifiers (such as Social Security or student numbers), and characteristics or other information that would make students’ identities easily traceable. The school must determine whether a reasonable person in the school community, who does not have personal knowledge of the relevant circumstances, could identify the student with reasonable certainty. Chrysler pointed out that the definition of de-identification is statute specific, and the FERPA has a different definition of de-identification than the Health Insurance Portability and Accountability Act does.
Written consent is generally required before a school can disclose personally identifiable information (electronic consent is permitted). Consent is specific, Chrysler explained, and must identify records that may be disclosed, the purpose of the disclosure, and to whom disclosure may be made. There are exceptions, however, and the FERPA allows sharing of data without consent under certain circumstances.
Common Family Educational Rights and Privacy Act Exceptions to Consent
Designated Directory Information
Designated directory information (e.g., a student’s name, address, e-mail, phone number, grade, date of birth) is generally not considered sensitive information. The school must inform parents about directory information and allow them a reasonable amount of time to opt out. This exception allows schools to provide class rosters and updated student contact information. Chrysler advised participants to check the school’s website for its policies and procedures on sharing information, including directory information.
Sharing with Other School Officials
Schools can share information with “school officials” in the same school if they have a legitimate need to know the information. This could include school nurses, teachers, administrators, counselors, or bus drivers, as well as volunteers or contractors, as long as the disclosure is in the educational interest of the student (i.e., the person needs the information to fulfill a professional responsibility to the school). The school must explain this in a required annual notice. Again, Chrysler advised participants to consult the school website to see the annual notice regarding disclosure of information to others.
To Improve Educational Efforts
Schools may disclose student information to public health agencies and others to evaluate and improve health education programs and health accommodations in schools. In this situation, public health is assisting the school with its mission. The data recipient must use and protect information consistent with a data-sharing agreement (i.e., data cannot be used for other purposes). For example, a school may provide asthma information to a public health agency to evaluate and develop efforts to improve students’ attendance and performance.
Health and Safety Emergency
Under this exception, disclosure of protected information must be necessary to protect the health or safety of the student or others and must be related to an actual, impending, or imminent emergency. Disclosure is limited to the period of the emergency and is only to appropriate parties. Decisions are made on a case-by-case basis. Schools must record in the student’s education record the threat that formed the basis for the disclosure and the parties to whom the information was disclosed. A health emergency necessitating disclosure might include disease outbreaks or urgent environmental threats.
Family Educational Rights and Privacy Act Resources
In closing, Chrysler referred participants to the ED website for resources on the FERPA, such as regulations, guidance, model forms, tutorials, and a data-sharing toolkit for communities.5
Health Insurance Coverage and Educational Outcomes
Health insurance coverage is often seen as necessary but insufficient to improving health care access for children, Hall said. Coverage is also important for educational outcomes, she said, and referred the audience to a 2015 study of the impact of Medicaid expansion in the late 1980s and early 1990s on children’s future educational attainment (Cohodes et al., 2016). The study found that Medicaid enrollment of newly eligible chil-
5 See https://www2.ed.gov/policy/gen/guid/fpco/index.html (accessed May 30, 2019) and https://www2.ed.gov/programs/promiseneighborhoods/datasharingtool.pdf (accessed May 30, 2019).
dren was more closely correlated with graduation from high school and college than class size or participation in Head Start.
The number of children who are uninsured in the United States is at an all-time low. The uninsured child rate is about 5 percent, but the distribution is disparate, Hall said. In Texas and Utah, for example, the rates are more than 10 percent. Enrollment is very low among immigrant families, even when the children themselves are eligible, she said. Coverage for parents often makes the difference as to whether children are enrolled, Hall continued. When states opted in to Medicaid expansion under the Patient Protection and Affordable Care Act (ACA) and coverage was expanded for adults (children were already eligible at those income levels), there was increased enrollment in those states of children who were already eligible but not enrolled. The reverse is also true. In Maine, where adult eligibility was cut in 2012, there was a drop of 13 percentage points in child enrollment in Medicaid in the income brackets where children were still eligible but the parents were no longer eligible. In summary, there are educational outcomes benefits to be reaped when schools become engaged in helping students and their parents become enrolled in health insurance.
Other Financial Levers to Support Health–Education Collaboration
As discussed, the coordinated care organizations in Oregon provide financial incentives that pair community-based outcomes, such as educational attainment or readiness, with health care financing. Other examples of leveraging payment to foster behavior change in health care providers come from the Center for Medicare & Medicaid Innovation, established under the ACA. It is clear that hospitals and providers will adapt behavior to the payment model being tested or deployed. For example, research has long demonstrated effective approaches to preventing hospital-acquired conditions and reducing hospital readmission rates, but hospitals were generally not addressing these issues on their own. There was no meaningful shift in attention to these concerns until Medicare stopped paying for hospital-acquired conditions and applied significant penalties to hospitals with excess readmissions. Metrics that hold health care providers accountable, whether by payors, public scorecards, or other methods, have been shown to be more successful than many of the shared incentive approaches, Hall summarized. Accountable care organizations and bundled payments are current popular approaches in health care, but this distributed shared savings approach has not consistently produced the desired outcomes.
A challenge when developing financial incentives for improved children’s health outcomes is that children’s care is, in general, relatively
inexpensive. Except for a small number of very high-cost children, most children already present a low cost to the system, and there is little opportunity to reduce health care spending on children any further, Hall explained. Consequently, there are not sizeable shared savings to be distributed from investing in preventive health for children as there are for preventive health initiatives in higher-cost populations, such as those aged 65 and older, and people with chronic disease. A more blunt approach will be needed, she said. In Medicaid, for example, states could mandate in their contracts with their managed care companies that certain data be collected or publicly reported, or they could include community-based measures that require managed care companies and health care providers to establish community partnerships to achieve set goals.
After more than a decade of educational approaches that have not improved educational outcomes (e.g., competition, governance arrangements, school choice, rooting out and removing “bad” teachers), there is a movement in the field to address the needs and well-being of the whole child, not just the academic part of the child, McGuire said. As mentioned by Davis, the implementation of the ESSA presents an opportunity to do this, and McGuire said that it remains to been seen how states and localities decide to interpret and implement the ESSA.
While this is good news, McGuire said, implementing a robust system of metrics will require substantial effort. Education is focused on attainment. In-school, academic factors that are strong predictors of academic success and attainment include early academic and cognitive skills (e.g., literacy). Academic success is a strong predictor of whether a student will move upstream through the system. Attendance is a function of both time and student engagement, but educational systems consider attendance less in terms of accountability and more as a measure of student behavior. Ninth-grade grade point average is a strong predictor of whether a student graduates or graduates on time. College readiness is also an indicator of attainment, as a high school diploma does not take one very far anymore and completing a post-secondary degree or credential is often needed.
Outside of school, there are family, neighborhood, and community factors that either contribute to or inhibit readiness, engagement, and educational attainment, McGuire said. These include parenting, home
learning, and environmental conditions (e.g., segregation, housing patterns, transportation issues). These elements are difficult to measure, both practically and legally. If there were ever a national indicator system that could reliably measure these elements, the education system could weigh and balance both the in-school and out-of-school factors that could give rise to better outcomes.
There are substantive challenges to moving forward. Most of the current approaches are problem oriented, not solution oriented, McGuire said. They point to the problem but do not offer any guidance on what to do about it. Approaches are needed that describe how to achieve the performance or the outcome that is desired. There has also been a focus on capturing data on what people perceive or think, rather than measuring the behavior that contributes to the problem. Approaches are also needed that provide analytic links among key components (e.g., linking data on health conditions to providing resources at the school or classroom level). McGuire added that people in the system often will not identify a problem if they do not think they can fix it, because doing so could put them at risk. Similarly, some feel that there is no point in developing measures and indicators of factors that cannot be changed or impacted. Finally, if the indicators and measures do not reveal causal linkages to educational attainment, then interest in them wanes.
Attainment is as important to the education sector as better health outcomes are to the health sector, McGuire said. Indicators of attainment are largely conceived and advanced in an accountability context. However, he noted, if educators do not know how to foster educational success, just holding them accountable for doing so will not work.
Overcoming Barriers Resulting from the Family Educational Rights and Privacy Act
Gourevitch observed that the FERPA was put into place for good reasons, but its impact on data sharing for the benefit of student outcomes is becoming increasingly evident, and schools and providers are using work-arounds, such as consents. Chrysler noted that there has been ongoing advocacy for changes to the FERPA, especially from the immunization community. Davis wondered to what extent data-sharing barriers are the result of the actual statute and regulations and to what extent they are the result of the attitude of lawyers who interpret the law and decline requests for data sharing because they do not understand the shared vision. Chrysler responded that it takes a lot of work to map the laws and the streams of data and find ways that the data might be shared. Consent
is always the gold standard for data sharing, she noted, but there can still be problems. It is also important to understand the characteristics of those who do not consent and who are therefore being left out.
Jeffrey Levi of The George Washington University noted the need to remember why current privacy laws are in place and to take great care in making changes and diminishing legal protection. He recalled a time when confidentiality laws were being implemented for AIDS reporting and how local elected officials wanted health departments to report the names of HIV-infected teachers so they could be fired. He reminded participants that children who registered under Deferred Action for Childhood Arrivals were promised by the federal government that there would be absolute confidentiality and no consequences, yet a change of administration has had severe consequences.
Metrics and Measurement
Philip Alberti of the Association of American Medical Colleges observed that there has been discussion of how interventions could potentially benefit health and educational equity, but the issue of measuring inequity so it can be targeted has not been raised. With regard to chronic absenteeism, for example, there are racial biases or inequities in suspensions. He suggested the need to stratify some of these educational metrics by race/ethnicity, sexual orientation, or other de-identified demographic data. He asked how programs are ensuring that interventions are actually addressing disparities. This is a very complicated question, Davis responded. She said that there are concerns that national education policy does not really achieve equity. The first accountability system between the federal government and the state government does not specifically focus on equity. Some states do place a higher value on trying to address equity, and there is the opportunity to focus on equity in the accountability structure between the state and the school district. The ESSA presents new opportunities, she said, but states often do not have the capability to support school districts. She reiterated the point by McGuire that if people do not know how to fix the problem, the correction will not happen. There is also fear that labelling schools as failing reinforces a very negative narrative, she observed. With regard to chronic absenteeism, the strategy has often been to blame parents and criminalize students for lack of attendance. The hope is that, by changing the metric to chronic absenteeism and building in an early warning system, the focus can instead be on supporting families and helping students. Simply changing the metric will not change the narrative, however. Davis added that what is needed is a change of the mindset and commitment of those in leadership, at both the state and school district levels, and in
community engagement. The challenges of racism are much deeper than any policy solution, she said.
Matt Stiefel of Kaiser Permanente observed that measurement in both health and education has been focused on children’s readiness, development, and function on their way to becoming adults and productive members of society. There is not much attention in either field on the child’s quality of life or well-being along the way. McGuire agreed and said he was encouraged by recent conversations about the science of learning, in which these measures of health and well-being are being discussed. Davis also noted the disconnect between metrics and readiness for life. She mentioned a special education teacher who was recognized by her school for record improvements on all of her students’ test scores but who still feels she has not helped her students be ready to succeed in life.
Hanh Cao Yu of The California Endowment emphasized the need to think about the role of data and success metrics and to avoid using that data to pathologize children and their families when studying the root causes of inequities. Metrics need to be applied to the blind spots in systems, she said, and there needs to be attention to the evidence base about what works and what does not for addressing inequities. McGuire agreed and noted that most of the existing measures and the narrative around them places the onus on the children. The conversation tends to be about personal deficits, and it is a struggle to bring the discussion back up to the institution and system levels.
Medicaid and Education
Kaplan recalled the persuasive argument by Woolf that there is a relationship between educational attainment and health outcomes and that investing in education can achieve greater health. He recalled the corresponding argument by Basch that investing in health can lead to better educational outcomes. He noted the data discussed by Hall that states that accepted the Medicaid expansion and enrolled eligible children had better educational outcomes. However, he said that a recent National Bureau of Economic Research report suggests that Medicaid expansion has not improved health outcomes. The question is whether investing in Medicaid to achieve better educational outcomes is a way to support the under-resourced educational system.
Hall disagreed with the analysis that expanding Medicaid does not improve health outcomes. Perhaps measuring life expectancy does not show dramatic change; however, across a range of metrics, Medicaid has led to change. There are important ancillary benefits to children’s educational outcomes from their parents and themselves being insured that have nothing to do with their health, she continued. For example, there
is improved economic security for families that have gained coverage through the Medicaid expansion, including reductions in payday lending, overall family debt, and bankruptcies. Expanding Medicaid might not be the answer to improving graduation rates, she said, but it is a wise investment that can improve health outcomes for children and their parents and increase economic stability, readiness, food security, and housing security. She suggested that managed care companies should work to improve educational outcomes in partnership with the community in order to earn the full capitation rate and that hospital systems should follow the lead of the Cincinnati Children’s Hospital Medical Center and invest in the community from their overall budgets.
Pamela Russo of the Robert Wood Johnson Foundation (RWJF) asked Davis whether a Health Impact Project health impact assessment of the ESSA that was done before it was passed resulted in any change in the bill. She added that RWJF, through the Health Impact Project, is supporting Community Catalyst, which is working with states on ESSA implementation. Specifically, they are working to promote assessment of student substance abuse and referral to treatment, rather than punitive action or reporting to the police. She asked about other state-level interventions that could be of value. Davis replied that the Healthy Schools Campaign was involved with that health impact assessment, which she clarified was done shortly after the law was passed, during the initial phase of the regulatory process. She suggested that the health impact assessment might have had more traction had there not been a change of administration. The resulting document was well done, she said, and provides one of the best and most recent compilations of the impact of health interventions on education.7 It also created a framework for conversation for organizations to engage states on the recommendations. One of the challenges for implementation of the ESSA is that the regulations were overturned by the new administration and a new guidance was issued. The new administration’s requirements regarding state plans are vague, and decision making is happening at the state level, creating opportunities for engagement, Davis said. She noted that RWJF is funding the Healthy Schools Campaign to work in several states to promote needs assessment.
7 The report is available at https://www.pewtrusts.org/-/media/assets/2017/08/hip_the_every_student_succeeds_act_creates_opportunities_to_improve_health_and_education_at_low_performing_schools.pdf (accessed May 30, 2019).
This page intentionally left blank.