Community Health Interventions: A review of research relevant to Allina Health Systems Backyard Initiative

21 December 2009

J. Michael Oakes, PhD McKnight Presidential Fellow Associate Professor Division of Epidemiology & Community Health Minnesota Population Center University of Minnesota oakes007@umn.edu

EXECUTIVE SUMMARY` This report summarizes scientific research relevant to Allina Health Systems Backyard Initiative (BYI). The BYI is an effort to galvanize a coalition of community residents and community-based, government, education and healthcare organizations towards improving the health and healthcare of residents in Allinas backyard, defined as persons residing within approximately one mile of Allinas corporate headquarters, Abbott Northwestern Hospital and the Phillips Eye Institute. Motivated by the recognition that a new model of disease prevention and healthcare is needed, Allinas BYI represents a novel attempt to extend traditional healthcare beyond the walls of clinics and hospitals and into a geographically bounded community. Announced in May 2008, the BYI began with Allina engaging community stakeholders in structured and informal conversation. Many issues and action ideas were discussed and debated. The work resulted in the following four BYI focus areas/interventions being identified: (1) engaging communities/building bridges, (2) primary and secondary prevention, (3) improving care access and (4) early childhood education. The goal of this document is to summarize the scientific support for the focus areas and to provide a scientific rationale for the BYI efforts. This review is centered on peer-reviewed research papers and published summaries that address the health benefits associated with the BYIs focus areas/intervention. Special attention is given to experimental studies. Recall that in this context experimental studies are those in which a health intervention is randomly assigned to persons or groups. Such studies are especially important for community health initiatives because they help researchers disentangle the impacts of the interventions under investigation from the background characteristics and natural health trajectories of residents. While not without limitations, experiments may accordingly be viewed as the relative gold standard of scientific evidence. The research presented here was selected from works identified in electronic databases, bibliographies of certain key papers and books, and through professional networks. Extensive effort was devoted to identifying the most careful and neutral reviews and key summaries from the tens of thousands of potentially relevant works. Note well that the issue of health system

cost is not considered here. Costs are a separate and perhaps even more complicated matter than the effects of interventions.

Engaging Community/Building Bridges No experimental research was found that addressed the health impact of improved community based organization activity. Related research addressing community activation to prevent youth alcohol abuse was found to demonstrate negligible to modest impacts. A great deal of non-experimental research on the impact of community-based participatory efforts suggest modest health benefits are possible, but this hypothesis, while promising, remains to be fully tested. In any event, community engagement is probably the right thing to do.

Primary and Secondary Prevention In terms of health (actually, disease) screening, there is mixed evidence that early screening for certain cancers can yield preferred outcomes. Routine screening can detect cancers of the breast, colon, rectum, cervix, prostate, oral cavity, and skin at early stages. Yet for most of these cancers early detection has not proven to reduce mortality. What is more, cancer remains rare for those less than 60 years of age. It follows that the overall health benefit for a single community is muted. Evidence of the beneficial effect of primary screening for blood pressure and obesity is mixed. Efforts to improve community physical activity and/or diets have not been very successful. Secondary screening after an event, such as a heart attack suggests stronger potential beneficial impacts. Too often overlooked, dental and oral health is important to overall health. Many dentists and hygienists provide scaling and polishing for patients at a regular interval, even if those patients are thought to be at low risk for developing periodontal disease. There is debate, however, over the clinical and cost effectiveness of routine scaling and polishing and the optimal frequency at which it should be provided. The evidence for preventive dentistry and dental screening for youth is slightly stronger, suggesting it is worthwhile.

ii

Improved Care Access While it may be surprising, the fact is that it is difficult to estimate the health gains associated with the provision of health insurance, especially for otherwise healthy people. It is clear that health insurance increases the amount of health care consumed but it is unclear the degree to which such healthcare consumption actually improves health. Many studies document that the insured tend to have better health outcomes than the uninsured. But the magnitude of the causal link between health insurance and better health has not been definitively established. The reason for this is that ones insurance coverage is determined by many of the same factors that determine health status to begin with; that is, socioeconomic status. Absent a randomized trial wherein some needing health insurance are provided it while insurance is withheld from others, it is difficult to disentangle these effects. Not only are such experiments extremely expensive but there are obvious ethical challenges too. The only randomized experiment addressing the health effects of health insurance is the famous RAND Health Insurance experiment conducted in the 1970s. The results from this trial showed mixed but probably beneficial effects. On the other hand, the provision of health care for the less healthy or already ill seems both necessary and beneficial. Indeed, while evidence is fragmented and incomplete, it seems clear that access to care and greater continuity of care for the chronically ill is associated with less use of hospitals and emergency departments. And while imperfect, there is good evidence to suggest having health insurance is healthy. Recent attention to the impact of medical home or related changes to conventional primary care models has yet to yield sufficient scientific data, but appears promising. End of life care remains an important and difficult issue.

Early Childhood Education Throughout history the best predictor of good health outcomes is ones socioeconomic status, often measured by educational attainment. Indeed, the strong relationship between education and health is the foundation for research into the social determinants of health. There is virtual consensus that early life educational interventions are necessary to mitigate the effects of disadvantage, although research addressing how

iii

early childhood education affects life chances is vast and complicated. Credible research finds that for the otherwise disadvantaged, improved early learning confers value on acquired skills, which leads to self-reinforcing motivation to learn more and early mastery of a range of cognitive, social and emotional competencies makes learning at later ages more efficient and therefore easier and more likely to continue. Environments that do not stimulate the young fail to cultivate these skills and place children at an early disadvantage. Yet as opposed to academic achievement and some employment gains, research addressing the effects of early childhood education on health outcomes later is life is difficult to conduct and there is scant direct evidence. Nevertheless, circumstantial evidence suggests sustained high-quality early education confers critical advantages and subsequent positive health effects. Again, this should not be surprising given the strong and sustained relationship between socioeconomic status (e.g., educational attainment) and health. Research on home visiting interventions is mixed. In terms of the BYI, an important study evaluated the impact of a home visiting program to reduce parental risk factors for child abuse. Unfortunately, the program did not prevent child abuse or promote use of nonviolent discipline. It had a modest impact in preventing neglect. Home visitors often failed to recognize parental risks and seldom linked families with community resources. On the other hand, there is relatively good evidence that home visiting by nurses can improve birth outcomes among the disadvantaged.

Conclusion It is important to stress that there is no research that estimates or even considers the combined effect the four BYI initiatives would have on a given communitys health. This should not be surprising, the BYI is a novel comprehensive community health improvement effort. Further, as implied above, a credible evaluation of the BYI initiative would require enormous resources: ideally, twenty or more communities would be randomized to the treatment or control conditions, and residents would be followed and measured for many years to come. Difficult decisions about following persons moving into and out of target communities would have to be made. Furthermore, decisions about which health measure (e.g., mortality, cardiovascular health, asthma, anxiety) would be

iv

required. None of this is within the scope of the action-oriented BYI. Consequently, stakeholders must look at the research that considers each of the BYI interventions/components independently. Overall, there is only modest direct but strong indirect scientific evidence to support the selection and implementation of Allinas four BYI interventions. Taken individually, the interventions might be expected to improve the short and long-term health of certain community members. Taken together, the BYI interventions should be expected to modestly and meaningfully improve the health of community residents, especially those at higher risk for disease.

TABLE OF CONTENTS

Background ......................................................................................................................... 2 I. Engaging Community/Building Bridges ......................................................................... 5 II. Primary Care and Prevention ......................................................................................... 7 III. The Provision Of Health Insurance............................................................................. 12 IV. Early Childhood Education......................................................................................... 16 V. Home visiting............................................................................................................... 20 VI. Conclusion .................................................................................................................. 22 Works Cited ...................................................................................................................... 25

Background It is widely known that the United States spends more on health care than another other nation and yet it ranks relatively poorly with respect to overall measures of health status such as longevity. Among others, Schroeder (Schroeder 2007) argues that the paradox is explained by between-country differences in the social determinants of health, such as tobacco policy and access to quality education, instead of the availability of hightech medicine. Shroeder stresses the fact that while inadequate healthcare accounts for only approximately 10% of premature deaths it receives the lions share of attention and resources. It is ironic that in this age of genomics, proteomics and very high-tech medicine, the key determinant of health remains socioeconomic status (SES), which is typically measured by educational attainment, income and sometimes occupational prestige. As a general rule, those of higher SES enjoy better health than those of lower SES. Social epidemiologists have long shown this relationship to hold over place and time, and to be graded; that is, for every increment of SES improvement health improves incrementally too. In other words, the relationship between SES and health is not a step wherein only those at the lowest level have inferior health, but rather a linear slope with declines along the way. Recently, the relationship between SES and health has been appreciated by leading medical scholars (Isaacs and Schroeder 2004; Woolf 2009). The implication is that in order to improve health of populations, policymakers and/or interventionists must either (a) improve healthcare for the disadvantaged and/or (b) improve the SES of the disadvantaged (Oakes and Kaufman 2006). The relationship between the health of individuals and the health of communities and larger aggregates, such as states and nations, remains a central question for researchers working on the social determinants of health. The relationship, of course, is complicated because while individuals are affected by larger social, political and macroeconomic forces, they also contribute to them (Macintyre, Ellaway and Cummins 2002; Oakes 2008). What is clear is that impoverished communities pose severe structural obstacles to human development and good health (Bowles, Durlauf and Hoff 2006; Bowles, Gintis and Groves 2005; Brook et al. 1983; Goering and Feins 2003). Further, the relationship between exposures and health over the life course is quite

complex (Pollitt et al. 2008; Turrell et al. 2007). Questions such as What is the effect of a parent smoking during a childs infancy on that same childs risk of lung cancer later in life? are very difficult to answer because of intervening trends and factors. Nevertheless, there is increasing consensus that in order to improve the health of individuals we must improve the health the communities (i.e., their contexts) too. This means that there is increasing consensus that the healthcare system must address the social determinants of health at both the individual and community level. Allina Health Systems Backyard Initiative (BYI) is an effort to galvanize a coalition of community residents and community-based, government, education and healthcare organizations towards improving the health and healthcare of residents in Allinas backyard, defined as persons residing within approximately one mile of Allinas corporate headquarters, Abbott Northwestern Hospital and the Phillips Eye Institute. Motivated by the recognition that a new model of disease prevention and healthcare is needed, Allinas BYI represents a courageous and novel attempt to extend traditional healthcare beyond the walls of clinics and hospitals and into the community. Announced in May 2008, the BYI began with Allina engaging community stakeholders in structured and informal conversation. Many issues and action ideas were discussed and debated. This work resulted in the following four BYI focus areas/interventions being identified: engaging communities/building bridges, primary and secondary prevention, improving access and starting early. Notice, these interventions aim to improve health by (1) improving healthcare access for the disadvantaged and (2) increasing a childs SES through early educational interventions. They are also focused simultaneously on (a) individuals and (b) the whole community. The goal of this document is to summarize the scientific support for the focus areas and to provide a scientific rationale for the BYI efforts. This review is centered on peer-reviewed research papers and published summaries that address the health benefits associated with the BYIs focus areas/intervention only. Still, it is fair to state that this document is strikingly, if not foolishly, ambitious. There are literally thousands of studies, papers and texts on each and every aspect of the interventions discussed here. It is obviously impossible to know everything about each area much less summarize each nuance. That stated, it is nevertheless important to compile the information needed into a

single document so that both area experts and non-experts can benefit from what is presented and, if they so chose, dig deeper by leveraging the source documents provided. Special attention is given to experimental studies. Recall that in this context experimental studies are those in which a health intervention is randomly assigned to persons or groups. While not without detractors (see Imbens and Wooldridge 2009; Nathan and Hollister Jr 2008; Sanson-Fisher et al. 2007), such studies are especially important for community health initiatives because they help researchers disentangle the impacts of the interventions under investigation from the background characteristics and natural health trajectories of residents (Hannan 2006; Oakes 2004). Experiments may accordingly be viewed as the gold standard of scientific evidence (Burtless 1995; Cook 2002). The research presented here was selected from works identified in electronic databases, bibliographies of certain key papers and books, and through professional networks. Extensive effort was devoted to identifying the most careful and neutral reviews and key summaries from the tens of thousands of potentially relevant works. It is worth emphasizing that the issue of health system cost is not considered here. Costs are a separate and perhaps even more complicated matter than the effects of interventions. It will be helpful, especially for the less familiar, to understand that the track record social interventions aiming to improve welfare of any sort is poor. In fact, the late sociologist and distinguished program evaluator, Peter H. Rossi, stated with regret that after three decades of research we must appreciate the net measurable effect of any social intervention should be expected to be nil (Rossi 1987). This Rossis Rule of program evaluation remains true today. In other words, it has proven extremely difficult to improve social conditions so as to improve the welfare/health of program/intervention participants.

I. Engaging Community/Building Bridges Many disadvantaged communities appear to benefit substantially from the wellintentioned work of community-based, not-for-profit organizations. From mental health counseling to job placement services to addiction groups to food pantries, such organizations are often the last thing between community struggle and total despair. Against a mountain of anecdotal evidence of community-based organization effectiveness, it is notable that no experimental or especially rigorous research was found that systematically addressed the health impact of community based organization activity. This does not mean such organizations are ineffective. Instead, the lack of evidence means more research is needed to more accurately estimate the degree of their effectiveness. The few experimental studies examining whether coalitions influence external outcomes, such as reductions in community-wide rates of mortality, morbidity, injury, or risky health behaviors, have produced mixed findings (Zakocs and Edwards 2006). In short, severe methodological obstacles prevent unambiguous understanding (Atienza and King 2002). Still, Zakocs and Edwards (Zakocs and Edwards 2006) review community coalition effectiveness and concluded that the following factors are important to success: formalization of rules/procedures, leadership style, member participation, membership diversity, agency collaboration, and group cohesion. A modest body rigorous research has addressed community-centered or community-based public health interventions. These fall within the framework of community-randomized trials (Hannan 2006). In the 1980s and early 1990s a series of community trials aimed to improve cardiovascular (ie, heart) health (Luepker et al. 2000; Susser 1995). In short, researchers aimed to improve the diets, physical activity and medical responses of communities. For a variety of substantive and scientific reasons, these studies revealed little impact on target communities. Wagner et al (Wagner et al. 2000) report a community-based initiative designed to promote improved health by changing community norms, environmental conditions, and individual behavior in 11 western communities. With the exception of two intervention communities - a largely Hispanic community and a Native American

reservation - researchers found little evidence of positive changes in the outcomes targeted by the 11 intervention communities. The programs that demonstrated positive outcomes targeted dietary behavior and adolescent substance abuse. In a related vein, Wagenaar and colleagues (Wagenaar et al. 2000) conducted a novel experiment in which he endeavored to activate community action in a random set of communities so as to show the effects on youth alcohol use and abuse. Unlike many other interventions, Wagenaars intervention aimed to directly motivate community members to work together toward the common goal -- a collective action problem. While not definitive or especially strong, these results hold promise for similar work to activate community members themselves. A related but clearly distinct approach to community improvement is called community-based participatory research (CBPR). According to Lantz and colleagues (Lantz et al. 2006), CBPR is an approach to research that consciously blurs the line between researchers and the researched, or makes research subjects more than mere objects of research. CBPR is a collaborative approach to research that engages partners from a community in all phases of the research process, with a shared goal of producing knowledge that will be translated into action or positive social change for the community. As Lantz and colleagues reveal, in the realm of public health, CBPR efforts often focus on improving community health status and/or reducing social disparities in health. Much of the published literature regarding CBPR involves examples of intervention research in which a participatory approach was used to identify a community need or problem, to design an intervention, programmatic or policy response, to evaluate the intervention, and to make positive community change based on the research results (Lantz et al. 2006). Examples of intervention research using a CBPR approach include HIV Testing and Counseling for Latina Women in Los Angeles Seattle Partners adult vaccine intervention, the Sierra Stanford Partnership in Northern and the Center for Urban Epidemiological Studies policy research to promote reintegration of drug users leaving jail in New York City (Lantz et al. 2006). While clearly gaining popularity, a key question remains: Does CBPR work? What is the evidence that a participatory approach to public health research is effective and worthwhile? Since, according to Lantz and colleagues, CBPR is an approach to

research rather than an intervention in and of itself, this is a challenging question to attempt to answer. Some could argue that a better question is whether or not CBPR produces research results that are more likely to meet the long-term goals of creating interventions that address important community issues, identifying the mechanisms by which health disparities are created and perpetuated, and enhancing community capacity to identify and address salient issues on a long-term basis. A growing empirical literature suggests that this is indeed the case. In a recent evidence-based review of the CBPR literature related to health sponsored by the Agency for Healthcare Research and Quality (Viswanathan et al. 2004), researchers found evidence of enhanced research quality in 11 of the 12 completed intervention studies reviewed. This included documented evidence of enhanced participant recruitment in 8 studies, improved research methods in 4 studies, improved variable measurement in 3 studies, and improved intervention outcomes in 2 studies. This literature review also concluded that that there was very little evidence of diminished research quality resulting from CBPR was reported (Viswanathan et al. 2004). An additional conclusion was that 47 of the 60 CBPR studies analyzed for the review reported evidence of enhanced community capacity as an outcome of the CBPR project, with 9 studies also documenting increased capacity among researchers. It thus seems fair to state that there is value added from using a participatory approach in health-related research. But until this belief can be experimentally (or at least better) tested, the impact of CBPR on actual community health outcomes remains uncertain.

II. Primary Care and Prevention Since the advent of modern medicine, circa 1920, the public has largely associated the observable increase in length and quality of life with physicians and medicine. There can be no question that there is truth to this, especially when it comes to treatment of illness and trauma care. But the evidence supporting the conclusion for primary prevention (ie, preventing disease in the first place) is less clear. Among others, Thomas McKeown and Robert Fogel are distinguished pillars of skepticism when it comes to the historical role of medicine and physicians in lengthening life (Fogel 1995; McKeown 1976; McKeown and Brown 1955). McKeown ties progress

to the rise of public health infrastructure, such as sanitary plumbing. Taking the long view, Fogel, an economic historian, attributes improvements to human health to advances in nutrition and public health accomplishments not medical care. While most of this work is focused on England and Europe more generally, McKinlay and McKinlay (McKinlay and McKinlay 1977) contributed similar findings for America. The important work Bunker and colleagues (Bunker, Frazier and Mosteller 1994) estimated that just 16% of the life expectancy gain in the twentieth century was due to the beneficial results of medical care. Furthermore, many assumed that when the British adopted national health care in 1948 inequalities in health and life expectancy would dissipate due to better access to medical care. Yet the publication of the so-called Black Report in 1980 showed no such gains occurred. Socioeconomic status (or social class) still predicted health with disturbing precision. Hundreds of more recent contemporary studies come to nearly identical conclusions.

Family physician There is no dispute that physicians and the medical system more generally are often essential to save the life of a trauma victim or for treatment of the chronically ill, whether the illness is diabetes, asthma or other ailments. On the other hand, there is considerable uncertainty in the impact of routine physician care on disease prevention. Goodwin and colleagues (Goodwin et al. 2001) provide a useful overview of the relationship between primary care and health. They write that the potential of primary care practice settings to prevent disease and morbidity through health habit counseling, screening for asymptomatic disease, and immunizations has been incompletely met. These authors persuasively argue that, among other things, 1. Low rates of preventive services stem, in part, from the competing demands and opportunities of other important primary care responsibilities. 2. Previous attempts to increase rates of preventive service delivery have often resulted in modest improvement in the delivery rates of a limited range of preventive services.

3. Interventions that target a broad range of screening and health habit counseling approaches have rarely been evaluated, raising concerns that existing interventions may improve the rate of delivery of some services at the expense of others. Hsiao and Boult (Hsiao and Boult 2008) take the ideas further when they note the commonly held belief is that health care quality affects primary care outcomes, but then state that the evidence for this belief is fragmented and incomplete. What does appear true is that (1) greater continuity of care is associated with less use of hospitals and emergency departments, (2) greater continuity of care is also associated with lower health care costs, and (3) effective communication may be associated with better health status. Saultz and Lochner (Saultz and Lochner 2005) offer a critical review of the literature regarding the relationships between interpersonal continuity of care and the outcomes and cost of health care. The conclude that although the available literature reflects persistent methodologic problems, it is likely that a significant association exists between interpersonal continuity and improved preventive care and reduced hospitalization. Future research in this area must address more specific and measurable outcomes and more direct costs and should seek to define and measure interpersonal continuity more explicitly. In sum, at this point it appears safe to say that consistent and high-quality primary care appears to help prevent disease and minimize the effects of problems once they occur. But research on the effects of scaling-up quality primary care for a whole community is lacking, and thus we cannot yet argue from evidence that such an approach is beneficial. Such conclusions must come from common sense and the weight of available circumstantial evidence.

Cancer screening Cancer is typically viewed as one of the most dreaded diseases of modern times. It is estimated that nearly 1.5 million US men and women will be diagnosed with and

562,340 men and women are expected to die of cancer of all sites in 2009. According to the Centers for Disease Control and Prevention (CDC), the leading incident cancers among males and females in the US (all races, 2005 data) are Prostate (142.2 per 100,000), female breast (117.7 per 100,000), Lung (67.7 per 100,000) and colon/rectum (48.3 per 100,000). Death rates are similar but are, in order of rank, lung, prostate, breast and colon. Because cancer in children is rare, most screening is directed at adults, often aged 40 years or more. Prevention, early detection, and treatments, if not cures, for various cancers has been a priority of medical science for decades. Prevention and early detection of cancer are often comingled such that the latter implies the former. While not technically correct, the point is understandable as late stage cancers are often difficult to treat. For decades the prevailing wisdom has been to encourage cancer screening through PSA tests, mammography, and colonoscopy. Accordingly, such cancer screening has risen dramatically. Unfortunately, the desired benefits of breast and prostate screening have proven elusive. Indeed, many medical scholars now question the recommendations. As recently as their 2009 policy paper (Smith, Cokkinides and Brawley 2009), the American Cancer Society recommends routine mammography for otherwise healthy women beginning at age 40. Yet in their subsequent and pivotal 2009 paper (Esserman, Shieh and Thompson 2009), Esserman and colleagues point to the observable failure of breast cancer screening to reduce mortality while simultaneously increasing the risks associated with early detection of slow-growing (ie, not life-threatening) cancers. Coming to similar conclusions is the National Breast Cancer Coalition (NBCC), who writes on their website (2 November 2009) that there is insufficient evidence to recommend for or against screening mammography in any age group of women. Similar conclusions have emerged for prostate cancer screening: at this time, there are insufficient data to recommend for or against prostate cancer testing in men at average risk of developing the disease. Until recently the American Cancer Society recommended that at age 50, the PSA blood test and the digital rectal examination should be offered to met at average risk (Smith, Cokkinides and Brawley 2009). But two large clinical trials designed to determine the efficacy of PSA testing (Andriole et al. 2009;

10

Schroder et al. 2009) suggest screening may not be as helpful as hoped. The risks associated with a false positive seem to outweigh the benefits of routine screening. Early screening for lung cancer has never had the same appeal as breast and prostate screening (Smith, Cokkinides and Brawley 2009). Chest x-ray, analysis of cells in sputum, and fiberoptic examination of the bronchial passages have shown limited effectiveness in improving survival. Much of this is due to the anatomy of the lungs themselves. Newer tests, such as CT scans and molecular markers in sputum, have produced promising results in detecting lung cancers at earlier, more operable stages when survival is better. However, there are considerable risks associated with lung biopsy and surgery that must be considered when evaluating the risks and benefits of screening. In conclusion, there seems to be little evidence to suggest that screening for lung cancer has positive effects on health outcomes. Despite recent declines in both incidence and mortality, colorectal cancer remains the second most common cause of death from cancer in the United States (Zauber et al. 2008). Largely because the organs anatomy lends itself to easier identification and (surgical) treatment of pre-cancerous and cancerous cells, screening for colon cancer remains effective in the prevention of deadly colon cancers. Research supports screening for and early treatment of colon cancer in persons aged 50 and beyond (Zauber et al. 2008). The risks of disease seems to clearly outweigh the risks of screening (eg, colonoscopy) and treatment (Burke and Inadomi 2009). Thus, among the big four cancers (prostate, breast, lung, colon), only screening for colon cancer in community members aged 50 years or beyond is scientifically justified at this time. There remains considerable debate about the benefits of mamography for women aged 40 or more.

Dental care The Surgeon General has declared that oral health is essential to the general health and well-being of all Americans. Dental caries (ie, cavities) is a common chronic disease that causes pain and disability across all age groups. If left untreated, dental caries can lead to pain and infection, tooth loss, and edentulism (total tooth loss).

11

It is hard to argue with the fact that in term of community prevention, water fluoridation is king. Fluoride's benefits for teeth were discovered in the 1930s and community water fluoridation began in 1945. CDC counts fluoridated water among its greatest public health achievements of the 20th century (CDC 1999). According to Dye et al (Dye et al. 2007), although dental caries has declined significantly among school-aged children since the early 1970s, dental caries has remained the most prevalent chronic disease of childhood. Although significant improvements in oral health for most Americans have been made over the past four decades, oral health disparities remain across some population groups. Research suggests a clear gradient between oral/dental health and socioeconomic status. In fact, approximately 45% of impoverished 20-44 year old are have untreated caries compared to 20% of non-poor persons in the same age group (CDC 2007). Similar ratios are observed for all age groups. What about the benefits of routine dental care? In their recent review of the benefits of routine dental care, (Beirne, Worthington and Clarkson 2007) conclude that research is inconclusive and that there is a need for well conducted trials in this area which include a sufficient number of patients to detect a true impact of routine dental care if any, and that are of significant duration (5 years or more). Once again, and not surprising, evidence suggests dental care for oral trauma and treatment for existing problems is beneficial. Seeing a dentist for a toothache is a good idea. On the other hand, evidence for routine care in community settings, while probably helpful, is not yet conclusive. More rigorous clinical trials are needed.

III. The Provision Of Health Insurance Access to adequate health care vis--vis health insurance remains a central political question of our time. The presumed deleterious effect not having adequate health insurance is based on two important causal factors: that (1) having health insurance is critically important to receiving medical care and (2) that medical care has a positive effect on health status. Evidence for both factors is vast but surprisingly inconclusive. Furthermore, direct evidence of the impact of health insurance on a whole communitys

12

health does not exist. No research addressing the provision of health insurance to a particular community was found. Hadley (Hadley 2003a) provides a remarkably comprehensive and careful review of the literature. In terms of the broad impact of inadequate health insurance on health, the IOM reviewed research and estimated the lack of health insurance (among the poor) caused an excess 18,000 deaths per year. Further, the IOM concluded that the uninsured are much more likely than persons with insurance to go without needed care. One nationally representative survey cited found that uninsured people were less than half as likely as those with insurance to receive needed care, as judged by physicians, for a serious medical condition (IOM 2001). Those without insurance also receive fewer preventive services and less regular care for chronic conditions than people with insurance (IOM 2001). Uninsured people with chronic diseases are less likely to receive appropriate care to manage their health conditions than are those who have health insurance. The impact is that for the five disease conditions that the Committee examined (diabetes, cardiovascular disease, end stage renal disease, HIV infection, and mental illness), uninsured patients had worse clinical outcomes than insured patients (IOM 2002). It may thus seem somewhat odd then that when Brown, Bindman, and Lurie (Brown, Bindman and Lurie 1998) reviewed the literature published between 1966 and 1996, they found the assumption of lack of insurance yielding poor health was not be supported by rigorous research. Why? Because while research shows that health insurance increases amount of health care consumed it is not clear if health care consumption actually improves health. More broadly, the fundamental problem with studying the relationship between health insurance and health is that the insured differ than the uninsured in many ways (SES, health status, race, education, etc.). Since insurance coverage is determined by many of the same factors that determine health status, it is difficult to disentangle these effects. When we ask why is an uninsured person sick we must consider whether it is because they lack insurance or because they are poor. Randomized experiments are needed here but due to ethical and financial reasons they are hard to do in practice. The only randomized experiment on this subject is the RAND Health Insurance Experiment (Brook et al. 1983; Keeler 1985; Newhouse 1982; Newhouse 1993). The

13

simple description of this study is that between 1974 and 1982, a total of 3,956 people between the ages of 14 and 61 who were free of disability that precluded work were randomly assigned to a set of insurance plans for three or five years. The overall results revealed that the more people had to pay for medical care the less of it they used (adults sharing costs of care made 1/3 fewer ambulatory visits and were hospitalized 1/3 less often than those with free care). Importantly, the reduced service use under the cost sharing plans had little or no adverse effect on health for the average person. On the other hand, health among the sick poor (defined as the disadvantaged 6% of the population) was adversely affected by lack of insurance. For those with poor vision and for low income individuals with high blood pressure, free care brought and improvements (vision better by 0.2 Snellen lines, diastolic blood pressure lower by 3mmHg). But free care had no effect on major health habits associated with heart disease and Cancer (smoking, weight, cholesterol levels). Another experiment-like study is going on in Massachusetts, which now has (virtually) universal coverage for citizens. It is not yet clear if health is improving for the newly insured in Massachusetts. Long (Long 2008) reports that in the first year after implementation the proportion of uninsured dropped from 13 to 7 percent, with greatest gains among lower income and younger adults and racial minorities. But the impact on health has not yet been determined. While not focused on the US population, cross national experiments are nevertheless informative. Consider that universal health coverage in Taiwan began in 1995 and Wen (Wen, Tsai and Chung 2008) attempted to assess the role of national health insurance in improving life expectancy and reducing health disparities there. This study found that life expectancy improved for the lower ranked classes after implementation of universal health insurance coverage. However, the magnitude of the reduced disparity was small and large health disparity gaps remained. Wen concluded that relying on health insurance alone to reduce health disparities is not realistic and other measures will need to be taken to reduce health disparities. Canada had similar results. Universal health coverage in Canada was rolled out in stages between 1962 and 1972. Hanratty (Hanratty 1996) examined the effect of universal health coverage on infant health outcomes. He found a 4% decline in the infant mortality rate when using a panel of

14

counties from 1960 to 1975. When using a universal sample of live births from 1960 to 1974, the incidence of low birth weight decreased by an average 1.3% for the entire population and by 8.9% for single parents following the introduction of Canadian National Health Insurance. Of course several quasi-experimental studies have been conducted that analyze a change in policy which created some variation in health coverage. The principal limitations of these studies are that they struggle to disentangle effects of insurance from other factors, results often apply only to specific populations, and some of them have small sample sizes. Studies of the termination of health insurance benefits in medically indigent adults (Lurie et al. 1984; Lurie et al. 1986) and in a low income veterans population (Fihn and Wicher 1988) found evidence of deterioration in health; specifically hypertension was in poor control after termination of benefits. Other studies examine public health insurance in pregnant women, children, and the elderly. These studies showed mixed results. Again, it is clear that health care consumption increases with public insurance (Currie and Gruber 1996; Currie and Gruber 1997) but the evidence on degree of improvement of health outcomes is mixed. There are numerous (thousands, actually) of studies which rely on cross-sectional data or longitudinal data to examine health care consumption, or the effects of health insurance on health. Most of these studies find that having health insurance improves health. Many of them compare health outcomes for insured and uninsured individuals. While some of these studies do have interesting results, but as noted above there may be many (unobservable) differences between those who are insured and those who are not. This means that causal effects of health insurance on health can not really be evaluated using these studies. In addition, some of these studies use health care consumption as a measure of improved health; this is a problem because health care consumption does not necessarily improve health. See (Hadley 2003b) for more. In sum, there is conflicting evidence to date that health insurance improves health in the general population but there is evidence that health insurance improves the health of specific population subgroups. There is a surprising lack of experimental studies examining the impact of health insurance on health in a given community. In the end, this paper endorses Hadleys (Hadley 2003a) articulate conclusion:

15

This review finds that there is a substantial body of research supporting the hypotheses that having health insurance improves health and that better health leads to higher labor force participation and higher income. However, none of these studies are definitive; nor are their findings universally consistent. While all of the studies reviewed, including those whose findings are consistent with the above hypotheses, suffer from methodological flaws of varying degrees, one general observation emerges: there is a substantial degree of qualitative consistency across the studies that support the underlying conceptual model of the relationship between health insurance and health. (page 60S)

IV. Early Childhood Education The general rule in health research is that the higher a persons or communitys socioeconomic status the better their/its health (Rogers, Hummer and Nam 2000). Because educational attainment is arguably the most important component of socioeconomic status (Oakes and Rossi 2003), it follows that the higher ones educational attainment the better their health. Substantial research confirms this relationship. Since education is cumulative, the roots of educational success lie in early life and early education. The question at hand, then, is what determines educational success in early life and what interventions improve it? It is important to emphasize that until recently there has been little attention paid to the effects of schooling be it pre-school or college on health outcomes. Instead, most school-effects research has focused on the impact of this or that program on IQ scores, graduation rates, employment opportunities and so forth. Links from early education, especially, to health outcomes later in life are few and far between. Further, there is virtually no literature on the effect of school segregation on health, in particular youth risk behavior and only one article was found on school racial segregation and school violence. An exception is Tarlov (Tarlov 2008), who stresses the importance of the recognition that the production of child development is related to the production of health. He argues that initiatives that provide high-quality early childhood education from

16

birth to five years are likely to yield high health status both at the time of the initiative and later in life. There is virtual consensus that early life educational interventions are necessary to mitigate the effects of disadvantage, although research addressing how early childhood education affects life chances is long and complicated (Gormley Jr 2007). Heckman (Heckman 2006) reviews evidence on the effects of early environments on child development and achievement. He writes that early learning confers value on acquired skills, which leads to self-reinforcing motivation to learn more and early mastery of a range of cognitive, social and emotional competencies makes learning at later ages more efficient and therefore easier and more likely to continue. Early family environments are major predictors of cognitive and noncognitive abilities. Environments that do not stimulate the young fail to cultivate these skills and place children at an early disadvantage. Children who fall behind may never catch up. The track record for rehabilitation later in life, be it for criminal behavior or literacy, is remarkably poor. Cognitive skills are important, but so too are noncognitive skills such as motivation and perseverance. Heckman writes that Investing in disadvantaged young children is a rare public policy initiative that promotes fairness and social justice and at the same time promotes productivity in the economy and in society at large. Early interventions targeted toward disadvantaged children have much higher returns than later interventions (1902). The rigorous (i.e., experimental) evidence for the benefits of early childhood educational interventions is relatively slim. A key reason is that ethical dilemmas abound and effects often take years or even decades to observe. Still, several important studies exist and merit summary review. The Perry Preschool Program was a 2-year experimental intervention for disadvantaged African American children initially 3-4 years of age and from low SES families. Based in Ypsilanti, MI, the program began in 1962 and went to 1965. The intervention was a morning program followed by accompanying afternoon visits by a teacher to the childs home. A total of 58 young students were in this treated group (65 were in the comparison group). By age ten, IQ scores of the treated children were no better than those children in the comparison group, yet Perry Program children had higher achievement scores since they were more motivated to learn, presumably by the program

17

itself. By age 40, the treated children has higher high-school graduation rates, higher incomes, higher percentage of home ownership, few arrests and so forth. It is for this reason that this program has received so much attention. Of course it is not clear whether the school curriculum or the home visits are responsible for the gains. And several severe methodological concerns remain unanswered (Olsen 2003). Another high-profile effort was the Abecedarian of North Carolina, which was an intensive center-based preschool program that also targeted disadvantaged children, starting at age 4 months. In 1972, 112 children were randomized to the special program or a comparison group. Children in the treatment group received childcare 6-8 hours per day, 5 days per week. Additionally, they received nutritional supplements, social work services and medical care. Importantly, it seems that this effort permanently increased the IQ of children. Follow-up survey research found lower levels of smoking (39% v 55%), which is obviously very important to health outcomes (Olsen 2003). Another program receiving attention is the Chicago Child-Parent Center (CPC) program, which is more recent and larger in scale and less intensive (and less expensive) that the Perry or Abecedarian programs. The CPC program provided educational and family support to children aged 3 to 9. The program ran for 2.5 hours per day, 5 days per week during the school year and 6 weeks during the summer. The curriculum emphasized language and math skills. Compared to a (non-randomized) comparison group, CPC program children had better school and labor market outcomes, they were also less likely to be victims of child abuse or neglect or to engage in criminal activity (Olsen 2003). While the non-experimental design of this programs evaluation is concerning, the results remain promising. Perhaps the best known early childhood educational intervention is Head Start, which is a large scale program that began in 1965 as part of the War on Poverty. It was designed to improve the poor childs opportunities and achievements in order to end the pattern of poverty. Its seven major objectives were to (1) improve the childs physical health, (2) help the childs social and emotional development, (3) improve the childs mental processes, (4) establish patterns and expectations of success, (5) increase the childs ability to relate positively to family members, (6) develop in the child and family

18

a responsible attitude toward society, and (7) increase the sense of dignity and self-worth of the child and his family. (Olsen 2003) Research on the impacts of Head Start have shown mixed results. According to a 1997 Government Accounting Office (GAO) report, from over 600 published research articles, only a few were credibly informative. GAO concluded that the body of research on current Head Start is insufficient to draw conclusions about the impact of the national program (Olsen 2003). This has now changed. Head Start was recently evaluated through an experimental design (Puma et al. 2005). Approximately 5,000 newly entering 3- and 4-year-old children applying for Head Start were randomly assigned to either a Head Start group that had access to Head Start program services or to a non-Head Start group that could enroll in available community non-Head Start services, selected by their parents. Data collection began in fall 2002 and was continued through 2006, following children through the spring of their 1st-grade year. Preliminary results show there were small to moderate statistically significant positive impacts for both 3- and 4-year-old children on several measures across four of the six cognitive constructs, including pre-reading, pre-writing, vocabulary, and parent reports of childrens literacy skills. But no significant impacts were found for the constructs oral comprehension and phonological awareness or early mathematics skills for either age group. For 3-year-olds, there were small to moderate statistically significant impacts in both constructs, higher parent reports of childrens access to health care and reportedly better health status for children enrolled in Head Start. For children who entered the program as 4-year-olds, there are moderate statistically significant impacts on access to health care, but no significant impacts for health status. I Because it considers broader and longer term outcomes, a recent study by Ludwig and Miller (Ludwig and Miller 2007) on the impacts of Head Start merits careful consideration. In this novel regression-discontinuity design study, researchers examined the community benefits and impacts of Head Start interventions, which began in the early and mid 1960s. The researchers note that Head Start is more than a daycare program; it is a comprehensive bundle of child and family intervention components designed to give children a true head start. Components beyond preschool include parent involvement and counseling, nutrition education, social services, mental health services, and (physical)

19

health services. Outcomes suggest that educational attainment (eg, high school graduation and college attendance) was directly improved by Head Start. What is more, related evidence suggests that community-level child mortality declined over the twenty years after program initiation. The remarkable upshot is that this relatively strong study shows that comprehensive community early childhood education not only impacts the success of target children but the long term health of communities in which they grow.

V. Home visiting It seems natural to assume home visitation interventions would increase the health and welfare of target persons, families and communities. Akin to a house call by a physician, social work and other public health nursing home visits have a long record of use. There are some positive results in this line of inquiry, especially as regards pregnancy outcomes. Outcomes such as child abuse appear more recalcitrant. Parker and colleagues (Parker et al. 2008) describe a study called the Community Action Against Asthma (CAAA). This is a community-based participatory research intervention in Detroit, MI that sought to improve childrens asthma-related health by reducing household environmental triggers for asthma. After randomization to an intervention or control group, 298 households with a child aged 7 to 11 with persistent asthma symptoms participated. The intervention consisted of a planned minimum of nine household visits over a 1-year period by community environmental specialists. The aim was to work with the family in making environmental changes in the home to reduce the childs exposure to multiple common asthma triggers. The intervention was effective in increasing some of the measures of lung function, reducing the frequency of cough that wont go away and coughing with exercise, reducing the proportion of children requiring unscheduled medical visits and reporting inadequate use of asthma controller medication, reducing caregiver report of depressive symptoms, reducing concentrations of dog allergen in the dust, and increasing some behaviors related to reducing indoor environmental triggers. David Olds and colleagues (Kitzman et al. 2000; Olds et al. 1998; Olds et al. 2004) have conducted a series of randomized field experiments to assess the impact of home nursing visits. Evidence of impacts on a childs anti-social behavior and a womans

20

birth outcomes is mixed, as the magnitude of observed impacts were small. On the other hand, one of their more prominent studies (Olds et al. 2004) addressed an urban, primarily black sample and examined the effects of prenatal and infancy home visits by nurses on mothers' fertility and economic self-sufficiency and the academic and behavioral adjustment of their children as the children finished kindergarten. A variety of outcomes measures were examined including women's number and timing of subsequent pregnancies, months of employment, use of welfare, food stamps, and Medicaid, educational achievement, behavioral problems attributable to the use of substances, rates of marriage and cohabitation, and duration of relationships with partners and their children's behavior problems, responses to story stems, intellectual functioning, receptive language, and academic achievement. Results were promising. Compared to those in the comparison group, women visited by nurses had fewer subsequent pregnancies and births (1.16 vs 1.38 pregnancies and 1.08 vs 1.28 births, respectively), longer intervals between births of the first and second children (34.28 vs 30.23 months), longer relationships with current partners (54.36 vs 45.00 months), and, since the previous follow-up evaluation at 4.5 years, fewer months of using welfare (7.21 vs 8.96 months) and food stamps (9.67 vs 11.50 months). Nurse-visited children were more likely to have been enrolled in formal out-of-home care between 2 and 4.5 years of age (82.0% vs 74.9%). Children visited by nurses demonstrated higher intellectual functioning and receptive vocabulary scores (scores of 92.34 vs 90.24 and 84.32 vs 82.13, respectively) and fewer behavior problems in the borderline or clinical range (1.8% vs 5.4%). On the other hand, there were no statistically significant program effects on women's education, duration of employment, rates of marriage, being in a partnered relationship, living with the father of the child, or domestic violence, current partner's educational level, or behavioral problems attributable to the use of alcohol or drugs. Overall, one must be cautious given some methodological shortcomings. But the work of Olds and colleagues shows home visits are promising. Similarly, Lee and colleagues (Lee et al. 2009) recently reported results of a study that assessed the effectiveness of a prenatal home-visitation program in reducing adverse birth outcomes among socially disadvantaged pregnant women and adolescents. Here disadvantaged pregnant women and adolescents were randomized to either an intervention group that received bi-weekly home-visitation services (n=236) or to a

21

control group (n=265). Home visitors encouraged healthy prenatal behavior, offered social support, and provided a linkage to medical and other community services. Services were tailored to individual needs. The risk of delivering an LBW baby was significantly lower for the HFNY group (5.1%) than for the control group (9.8%). The risk was further reduced for mothers who were exposed to HFNY at a gestational age of 24 weeks These authors conclude that a prenatal home-visitation program with focus on social support, health education, and access to services holds promise for reducing LBW deliveries among at-risk women and adolescents. On the other hand, in their recent review, Howard and Brooks-Gunn (Howard and Brooks-Gunn 2009) review evaluations of nine home-visiting programs to prevent child abuse: the Nurse-Family Partnership, Hawaii Healthy Start, Healthy Families America, the Comprehensive Child Development Program, Early Head Start, the Infant Health and Development Program, the Early Start Program in New Zealand, a demonstration program in Queensland, Australia, and a program for depressed mothers of infants in the Netherlands. They examine outcomes related to parenting and child well-being, including abuse and neglect. Howard and Brooks-Gunn conclude that, overall, researchers have found little evidence that home-visiting programs directly prevent child abuse and neglect. But home visits can impart positive benefits to families by way of influencing maternal parenting practices, the quality of the childs home environment, and childrens development. And improved parenting skills, say the authors, would likely be associated with improved child well-being and corresponding decreases in maltreatment over time. Howard and Brooks-Gunn also report that the programs have their greatest benefits for low-income, first-time adolescent mothers.

VI. Conclusion The Allina Back Yard Initiative (BYI) aims to improve the health of a geographically bounded neighborhood/community area. Extensive stakeholder discussions led researchers and practitioners to focus improvements on four areas: (1) engaging communities/building bridges, (2) primary and secondary prevention, (3) improving access and (4) starting early. This report aimed to summarize the scientific

22

support for these efforts. As they generally yield more defensible findings, appropriate attention was devoted to experimental studies. In this era of high-tech medicine, multi-million dollar research efforts, and sophisticated social science, it is remarkable that no research on multiple simultaneous efforts to improve a communitys health was found. Above all else, this paper demonstrates the novelty of the BYI. Whereas nearly every other effort to improve health (1) restricted itself to one aspect of community health, such as infant mortality, or (2) one disease, such as diabetes, the BYI aims to improve community health in the broad sense. No research on such a comprehensive effort to improve a given small geographic area was found. The principal finding of this paper is that existing scientific research does not directly support the BYI plan to improve community health. But, there is relatively strong indirect evidence that the effort will prove successful. For example, there is evidence that, separately, community-based collaborations, some screening and preventive medicine, improved access to medical care, and early education improvements will increase overall community health. In other words, the BYI is both groundbreaking and, based on reasonable inference from existing science, likely to succeed in improving the target areas health. It must be mentioned that several prominent commentators argue for efforts much like the BYI. First, Acevedo-Garcia and colleagues make a strong case that we must move beyond merely documenting differences and deficiencies in health and begin to addressing what can be done to improve it (Acevedo-Garcia et al. 2008).Such a what can we actually do is refreshing and dovetails with Rossis idea of implementing politically feasible programs that can be shown to improve lives (Rossi 1980). In a series of commentaries, Woolf persuasively argues that prevention of disease is far superior than treating disease and that in terms of prevention the best health policy is social policy (Woolf 2009; Woolf 2008). Any careful read of the vast literature will force one to come to the same conclusions. Furthermore, Lantz and colleagues clearly articulate that the medicalization of health improvements fails to recognize social structural effects, such as poverty, education and fundamental living conditions (Lantz, Lichtenstein and Pollack 2007). Such fundamental causes (Link and Phelan 1995) lie at the heart of modern

23

social epidemiology. Finally, and quite remarkably, when Williams and colleagues independently reviewed the literature and considered how best they might improve community health they arrived at conclusions quite similar to this paper and the BYI initiative (Williams et al. 2008). In the face of insufficient and ambiguous research, such an independent validation of the BYI is reassuring.

24

Works Cited Acevedo-Garcia, D., T. L. Osypuk, N. McArdle, and D. R. Williams. 2008. "Toward a policy-relevant analysis of geographic and racial/ethnic disparities in child health." Health Affairs 27:321. Andriole, G. L., E. D. Crawford, R. L. Grubb, 3rd, S. S. Buys, D. Chia, T. R. Church, M. N. Fouad, E. P. Gelmann, P. A. Kvale, D. J. Reding, J. L. Weissfeld, L. A. Yokochi, B. O'Brien, J. D. Clapp, J. M. Rathmell, T. L. Riley, R. B. Hayes, B. S. Kramer, G. Izmirlian, A. B. Miller, P. F. Pinsky, P. C. Prorok, J. K. Gohagan, and C. D. Berg. 2009. "Mortality results from a randomized prostate-cancer screening trial." N Engl J Med 360:1310-9. Atienza, A. A., and A. C. King. 2002. "Community-based health intervention trials: an overview of methodological issues." Epidemiologic reviews 24:72-79. Beirne, Paul V., Helen V. Worthington, and Jan E. Clarkson. 2007. "Routine scale and polish for peridontal health in adults." Pp. 1-145 in Cochrane Database of Systematic Reviews: John Wiley & Sons, Ltd. Bowles, Samuel, Steven N. Durlauf, and Karla Hoff (Eds.). 2006. Poverty Traps. New York: Russell Sage Foundation. Bowles, Samuel, Herbert Gintis, and Melissa Osborne Groves (Eds.). 2005. Unequal Chances: Family background and economic success. New york: Russell Sage. Brook, RH, JE Ware, WH Rogers, EB Keeler, AR Davies, and et al. 1983. "Does free care improve adults's health? Results from a randomized controlled trial." New Engl J Med 309:1426-1434. Brown, M. E., A. B. Bindman, and N. Lurie. 1998. "Monitoring the consequences of uninsurance: A review of methodologies." Medical Care Research and Review 55:177. Bunker, J. P., H. S. Frazier, and F. Mosteller. 1994. "Improving health: measuring effects of medical care." Milbank Q 72:225-58. Burke, C. A., and J. M. Inadomi. 2009. "American college of gastroenterology guidelines for colorectal cancer screening 2008." Am J Gastroenterol 104:739-750.

25

Burtless, G. 1995. "The case for randomized field trials in economic and policy research." The Journal of Economic Perspectives 9:63-84. CDC. 1999. "Achievements in Public Health, 1900-1999." Morbitity and Mortality Weekly 48:933-940. . 2007. " Percentage of Persons with Untreated Dental Caries,* by Age Group and Poverty Status --- National Health and Nutrition Examination Survey (NHANES), United States, 2001--2004." Morbitity and Mortality Weekly 56:889. Cook, T. D. 2002. "Randomized experiments in educational policy research: A critical examination of the reasons the educational evaluation community has offered for not doing them." Educational Evaluation and Policy Analysis 24:175. Currie, J, and J Gruber. 1996. "Saving babies: the efficacy and cost of recent changes in the Medicaid eligibility of pregnant women." J. Polit. Econ 104:1263-1296. . 1997. "The technology of birth: health insurance, medical interventions and infant health." Natl Bur Econ Res. Dye, B. A., S. Tan, V. Smith, B. G. Lewis, L. K. Barker, G. Thornton-Evans, P. I. Eke, E. D. Beltran-Aguilar, A. M. Horowitz, and C. H. Li. 2007. "Trends in oral health status: United States, 1988-1994 and 1999-2004." Vital and health statistics. Series 11, Data from the national health survey:1. Esserman, L., Y. Shieh, and I. Thompson. 2009. "Rethinking screening for breast cancer and prostate cancer." JAMA 302:1685-92. Fihn, SD, and JB Wicher. 1988. "Withdrawing routine outpatient medical services: effects on access and health." J Gen Intern Med July/Aug:356-362. Fogel, R. W. 1995. "The contribution of improved nutrition to the decline of mortality rates in Europe and America." The state of humanity:6171. Goering, John, and Judith D. Feins (Eds.). 2003. Choosing a Better Life: Evaluating the Moving to Opportunity Study. Washington, DC: Urban Institute Press. Goodwin, Meredith A., Stephen J. Zyzanski, Sue Zronek, Mary Ruhe, Sharon M. Weyer, Nancy Konrad, Diane Esola, and Kurt C. Stange. 2001. "A Clinical Trial of Tailored Office Systems for Preventive Service Delivery: The Study to Enhance Prevention by Understanding Practice (STEP-UP)." American Journal of Preventive Medicine 21:20-28.

26

Gormley Jr, W. T. 2007. "Early childhood care and education: Lessons and puzzles." Journal of Policy Analysis and Management 26:633. Hadley, J. 2003a. "Sicker and poorer-The consequences of being uninsured: A review of the research on the relationship between health insurance, medical care use, health, work, and income." Medical Care Research and Review 60:3. . 2003b. "Sicker and poorer - the consequences of being uninsured: a review of the research on the relationship between health insurance, medical care use, health, work, and income." Med Care Res Rev 60:3S-75. Hannan, Peter J. 2006. "Experimental social epidemiology: Controlled community trials." Pp. 335-364 in Methods in Social Epidemiology, edited by J. Michael Oakes and Jay S. Kaufman. San Francisco: Jossey-Bass / Wiley. Hanratty, M. 1996. "Canadian national health insurance and infant health." Am Econ Rev 86:276-284. Heckman, James J. 2006. "Skill formation and the economics of investing in disadvantaged children." Science 312:1900-1902. Howard, K. S., and J. Brooks-Gunn. 2009. "The role of home-visiting programs in preventing child abuse and neglect." The Future of Children, 19 (2):77-104. Hsiao, Chun-Ju, and Chad Boult. 2008. "Effects of Quality on Outcomes in Primary Care: A Review of the Literature." American Journal of Medical Quality 23:302310. Imbens, G. W., and J. M. Wooldridge. 2009. "Recent developments in the econometrics of program evaluation." Journal of Economic Literature 47:5-86. IOM. 2001. Coverage Matters: Insurance and Health Care. Washington, DC: National Academies Press. . 2002. Care Without Coverage. Washington, DC: National Academies Press. Isaacs, S. L., and S. A. Schroeder. 2004. "Class - the ignored determinant of the nation's health." N Engl J Med 351:1137-42. Keeler, E. 1985. "How free care reduced hypertension in the health insurance experiment." JAMA 254:1926-1931. Kitzman, H., D. L. Olds, K. Sidora, C. R. Henderson, Jr., C. Hanks, R. Cole, D. W. Luckey, J. Bondy, K. Cole, and J. Glazner. 2000. "Enduring effects of nurse home

27

visitation on maternal life course: a 3-year follow-up of a randomized trial." JAMA 283:1983-9. Lantz, P. M., R. L. Lichtenstein, and H. A. Pollack. 2007. "Health policy approaches to population health: the limits of medicalization." Health Affairs 26:1253. Lantz, Paula M., Barbara A. Israel, Amy J. Shultz, and Angela Reyes. 2006. "Community-based Participatory Research: Rationale and relevance for Social epidemiology." Pp. 233-260 in Methods in Social Epidemiology, edited by J. Michael Oakes and Jay S. Kaufman. San Francisco: Jossey-Bass Lee, E., S. D. Mitchell-Herzfeld, A. A. Lowenfels, R. Greene, V. Dorabawila, and K. A. DuMont. 2009. "Reducing Low Birth Weight Through Home Visitation A Randomized Controlled Trial." American Journal of Preventive Medicine 36:154160. Link, B. G., and J. Phelan. 1995. "Social conditions as fundamental causes of disease." J Health Soc Behav Supplement:80-94. Long, Sharon. 2008. "Who gained the most under health reform in Massachusetts." Washington, DC: Urban Institute. Ludwig, Jens, and Douglas L. Miller. 2007. "Does Head Start Improve Children's Life Chances? Evidence from a Regression Discontinuity Design." Quartetly Journal of Economics 122:159-208. Luepker, R. V., J. M. Raczynski, S. Osganian, R. J. Goldberg, J. R. Finnegan, Jr., J. R. Hedges, D. C. Goff, Jr., M. S. Eisenberg, J. G. Zapka, H. A. Feldman, D. R. Labarthe, P. G. McGovern, C. E. Cornell, M. A. Proschan, and D. G. SimonsMorton. 2000. "Effect of a community intervention on patient delay and emergency medical service use in acute coronary heart disease: The Rapid Early Action for Coronary Treatment (REACT) Trial." Jama 284:60-7. Lurie, N, NB Ward, MF Shapiro, and RH Brook. 1984. "Termination from Medi-Cal Does it affect health?" N Engl J Med 311:480-484. . 1986. "Termination of Medi-Cal benefits. A follow-up study one year later." N Engl J Med 314:1266-1268.

28

Macintyre, Sally, Anne Ellaway, and Steven Cummins. 2002. "Place effects on health: how can we conceptualise, operationalise and measure them?" Soc Sci Med 55:125-39. McKeown, T. 1976. The modern rise of population: E. Arnold. McKeown, T., and R. G. Brown. 1955. "Medical evidence related to English population changes in the eighteenth century." Population Studies:119-141. McKinlay, J. B., and S. M. McKinlay. 1977. "The questionable contribution of medical measures to the decline of mortality in the United States in the twentieth century." Milbank Mem Fund Q Health Soc 55:405-28. Nathan, R. P., and R. G. Hollister Jr. 2008. "The role of random assignment in social policy research." Journal of Policy Analysis and Management 27:401. Newhouse, J.P. 1982. "A Summary of the Rand Health Insurance Study.": NYAS. Newhouse, Joseph P. 1993. Free for all?: Lessons from the Rand Health Insurance Experiment. Oakes, J. M. 2004. "The (mis)estimation of neighborhood effects: causal inference for a practicable social epidemiology." Soc Sci Med 58:1929-52. . 2008. "Invited commentary: rescuing Robinson Crusoe." Am J Epidemiol 168:9-12. Oakes, J. M., and P. H. Rossi. 2003. "The measurement of SES in health research: current practice and steps toward a new approach." Soc Sci Med 56:769-84. Oakes, J. Michael, and Jay S. Kaufman. 2006. "Introduction." Pp. 3-20 in Methods in Social Epidemiology, edited by J. Michael Oakes and Jay S. Kaufman. San Francisco: Jossey-Bass Olds, D., C. R. Henderson, Jr., R. Cole, J. Eckenrode, H. Kitzman, D. Luckey, L. Pettitt, K. Sidora, P. Morris, and J. Powers. 1998. "Long-term effects of nurse home visitation on children's criminal and antisocial behavior: 15-year follow-up of a randomized controlled trial." JAMA 280:1238-44. Olds, D. L., H. Kitzman, R. Cole, J. A. Robinson, K. Sidora, D. W. Luckey, C. R. Henderson Jr, C. Hanks, J. Bondy, and J. Holmberg. 2004. "Effects of nurse home-visiting on maternal life course and child development: age 6 follow-up results of a randomized trial." Pediatrics 114:1550.

29

Olsen, Darcy A. 2003. "Understanding the Promise of Universal Preschool." Pp. 329-352 in Early Childhood Programs for a New Century, edited by Arthur J. Reynolds, Margaret C. Wang, and Herbert J. Walberg. Washington, DC: CWLA Press. Parker, E.A., B.A. Israel, T.G. Robins, G. Mentz, X. Lin, W. Brakefield-Caldwell, and et al. 2008. "Evaluation of Community Action Against Asthma: a community health worker intervention to improve childrens asthma-related health by reducing household environmental triggers for asthma. ." Health Education & Behavior 35:376-395. Pollitt, R. A., J. S. Kaufman, K. M. Rose, A. V. Diez-Roux, D. Zeng, and G. Heiss. 2008. "Cumulative life course and adult socioeconomic status and markers of inflammation in adulthood." J Epidemiol Community Health 62:484-91. Puma, Michael, Stephen Bell, Ronna Cook, Camilla Heid, and Michael Lopez. 2005. "Head Start Impact Study: First Year Findings." Washington, DC: U.S. Department of Health and Human Services, Administration for Children and Families. Rogers, Richard G., Robert A. Hummer, and Charles B. Nam. 2000. Living and Dying in the USA: Behavioral, Health, and Social Differentials of Adult Mortality. New York: Academic Press. Rossi, Peter H. 1980. "The Presidential Address: The Challenge and Opportunities of Applied Social Research." American Sociological Review 45:889-904. Rossi, Peter H. 1987. "The iron law of evaluation and other metallic rules." Pp. 3-20 in Research in Social Problems and Public Policy, edited by Joann L. Miller and Michael Lewis. Greenwich, CT: JAI Press. Sanson-Fisher, R. W., B. Bonevski, L. W. Green, and C. D'Este. 2007. "Limitations of the randomized controlled trial in evaluating population-based health interventions." American Journal of Preventive Medicine 33:155-161. Saultz, John W., and Jennifer Lochner. 2005. "Interpersonal Continuity of Care and Care Outcomes: A Critical Review." Annals of Family Medicine 3:159-166. Schroder, F. H., J. Hugosson, M. J. Roobol, T. L. Tammela, S. Ciatto, V. Nelen, M. Kwiatkowski, M. Lujan, H. Lilja, M. Zappa, L. J. Denis, F. Recker, A. Berenguer, L. Maattanen, C. H. Bangma, G. Aus, A. Villers, X. Rebillard, T. van der Kwast,

30

B. G. Blijenberg, S. M. Moss, H. J. de Koning, and A. Auvinen. 2009. "Screening and prostate-cancer mortality in a randomized European study." N Engl J Med 360:1320-8. Schroeder, S. A. 2007. "Shattuck Lecture. We can do better--improving the health of the American people." N Engl J Med 357:1221-8. Smith, R. A., V. Cokkinides, and O. W. Brawley. 2009. "Cancer screening in the United States, 2009: a review of current American Cancer Society guidelines and issues in cancer screening." CA: A Cancer Journal for Clinicians 59:27. Susser, Mervyn. 1995. "Editorial: The Tribulations of Trials--Intervention in Communities." American Journal of Public Health 85:156-158. Tarlov, Alvin R. 2008. "The coproduction of human development and health." Pp. 41-64 in Investing in Early Childhood Development, edited by Alvin R. Tarlov and Michaelle Precourt Debbink. New York: Palgrave. Turrell, G., J. W. Lynch, C. Leite, T. Raghunathan, and G. A. Kaplan. 2007. "Socioeconomic disadvantage in childhood and across the life course and allcause mortality and physical function in adulthood: evidence from the Alameda County Study." J Epidemiol Community Health 61:723-30. Viswanathan, M., A. Ammerman, E. Eng, G. Gartlehner, K. N. Lohr, D. Griffith, S. Rhodes, C. Samuel-Hodge, S. Maty, and L. Lux. 2004. "Community-based participatory research: assessing the evidence." Evidence report/technology assessment 99:290-02. Wagenaar, A. C., D. M. Murray, J. P. Gehan, M. Wolfson, J. L. Forster, T. L. Toomey, C. L. Perry, and R. Jones-Webb. 2000. "Communities mobilizing for change on alcohol: outcomes from a randomized community trial." J Stud Alcohol 61:85-94. Wagner, E. H., T. M. Wickizer, A. Cheadle, B. M. Psaty, T. D. Koepsell, P. Diehr, S. J. Curry, M. Von Korff, C. Anderman, and W. L. Beery. 2000. "The Kaiser Family Foundation Community Health Promotion Grants Program: findings from an outcome evaluation." Health Services Research 35:561. Wen, CP, SP Tsai, and WS Chung. 2008. "A 10-year experience with universal health insurance in Taiwan: measuring changes in health and health disparity." Annals of Internal Medicine 148:258-267.

31

Williams, D. R., M. V. Costa, A. O. Odunlami, and S. A. Mohammed. 2008. "Moving Upstream: How Interventions That Address the Social Determinants of Health Can Improve Health and Reduce Disparities." Journal of Public Health Management and Practice 14:S8. Woolf, S. H. 2009. "Social policy as health policy." JAMA 301:1166. Woolf, Steven H. 2008. "The Power of Prevention and What It Requires." JAMA 299:2437-2439. Zakocs, R. C., and E. M. Edwards. 2006. "What explains community coalition effectiveness? A review of the literature." American Journal of Preventive Medicine 30:351-361. Zauber, A. G., I. Lansdorp-Vogelaar, A. B. Knudsen, J. Wilschut, M. van Ballegooijen, and K. M. Kuntz. 2008. "Evaluating test strategies for colorectal cancer screening: a decision analysis for the US Preventive Services Task Force." Annals of internal medicine 149:659.

32