An Empirical Study of Chronic Diseases in the United States: a Visual Analytics Approach to Public Health

International Journal of Environmental Research and Public Health

Article An Empirical Study of Chronic Diseases in the United States: A Visual Analytics Approach to Public Health

Wullianallur Raghupathi 1 and Viju Raghupathi 2,*

1 Gabelli School of Business, Fordham University, New York, NY 10023, USA; [email protected] 2 Koppelman School of Business, Brooklyn College of the City University of New York, Brooklyn, NY 11210, USA * Correspondence: [email protected]; Tel.: +1-(718)-951-5000

Received: 12 January 2018; Accepted: 27 February 2018; Published: 1 March 2018

Abstract: In this research we explore the current state of chronic diseases in the United States, using data from the Centers for Disease Control and Prevention and applying visualization and descriptive analytics techniques. Five main categories of variables are studied, namely chronic disease conditions, behavioral health, mental health, demographics, and overarching conditions. These are analyzed in the context of regions and states within the U.S. to discover possible correlations between variables in several categories. There are widespread variations in the prevalence of diverse chronic diseases, the number of hospitalizations for speciﬁc diseases, and the diagnosis and mortality rates for different states. Identifying such correlations is fundamental to developing insights that will help in the creation of targeted management, mitigation, and preventive policies, ultimately minimizing the risks and costs of chronic diseases. As the population ages and individuals suffer from multiple conditions, or comorbidity, it is imperative that the various stakeholders, including the government, non-governmental organizations (NGOs), policy makers, health providers, and society as a whole, address these adverse effects in a timely and efﬁcient manner.

Keywords: behavioral health; chronic disease; comorbidity; overarching condition; population health; preventive health

1. Introduction A chronic condition “is a physical or mental health condition that lasts more than one year and causes functional restrictions or requires ongoing monitoring or treatment” [1,2]. Chronic diseases are among the most prevalent and costly health conditions in the United States. Nearly half (approximately 45%, or 133 million) of all Americans suffer from at least one chronic disease [3–5], and the number is growing. Chronic diseases—including, cancer, diabetes, hypertension, stroke, heart disease, respiratory diseases, arthritis, obesity, and oral diseases—can lead to hospitalization, long-term disability, reduced quality of life, and death [6,7]. In fact, persistent conditions are the nation’s leading cause of death and disability [6]. Globally, chronic diseases have affected the health and quality of life of many citizens [8,9]. In addition, chronic diseases have been a major driver of health care costs while also impacting workforce patterns, including, of course, absenteeism. According to the Centers for Disease Control, in the U.S. alone, chronic diseases account for nearly 75 percent of aggregate healthcare spending, or an estimated $5300 per person annually. In terms of public insurance, treatment of chronic diseases comprises an even larger proportion of spending: 96 cents per dollar for Medicare and 83 cents per dollar for Medicaid [4,10–12]. Thus, the understanding, management, and prevention of chronic diseases are important objectives if, as a society, we are to provide better quality healthcare to citizens and improve their overall quality of life.

Int. J. Environ. Res. Public Health 2018, 15, 431; doi:10.3390/ijerph15030431 www.mdpi.com/journal/ijerph Int. J. Environ. Res. Public Health 2018, 15, 431 2 of 24

More than two thirds of all deaths are caused by one or more of these ﬁve chronic diseases: heart disease, cancer, stroke, chronic obstructive pulmonary disease, and diabetes. Additional statistics are quite stark [5,13]: chronic diseases are responsible for seven out of 10 deaths in the U.S., killing more than 1.7 million Americans each year; and more than 75% of the $2 trillion spent on public and private healthcare in 2005 went toward chronic diseases [5]. What makes treating chronic conditions (and efforts to manage population health) particularly challenging is that chronic conditions often do not exist in isolation. In fact, today one in four U.S. adults have two or more chronic conditions [5], while more than half of older adults have three or more chronic conditions. And the likelihood of these types of comorbidities occurring goes up as we age [5]. Given America’s current demographics, wherein 10,000 Americans will turn 65 each day from now through the end of 2029 [5], it is reasonable to expect that the overall number of patients with comorbidities will increase greatly. Trends show an overall increase in chronic diseases. Currently, the top ten health problems in America (not all of them chronic) are heart disease, cancer, stroke, respiratory disease, injuries, diabetes, Alzheimer’s disease, inﬂuenza and pneumonia, kidney disease, and septicemia [14–18]. The nation’s aging population, coupled with existing risk factors (tobacco use, poor nutrition, lack of physical activity) and medical advances that extend longevity (if not also improve overall health), have led to the conclusion that these problems are only going to magnify if not effectively addressed now [19]. A recent Milken Institute analysis determined that treatment of the seven most common chronic diseases coupled with productivity losses will cost the U.S. economy more than $1 trillion dollars annually. Furthermore, compared with other developed nations, the U.S. has ranked poorly on cost and outcomes. This is predominantly because of our inability to effectively manage chronic disease. And yet the same Milken analysis estimates that modest reductions in unhealthy behaviors could prevent or delay 40 million cases of chronic illness per year [11]. If we learn how to effectively manage chronic conditions, thus avoiding hospitalizations and serious complications, the healthcare system can improve quality of life for patients and greatly reduce the ballooning cost burden we all share [10]. The success of population health and chronic disease management efforts hinges on a few key elements: identifying those at risk, having access to the right data about this population, creating actionable insights about patients, and coaching them toward healthier choices. Methods such as data-driven visual analytics help experts analyze large amounts of data and gain insights for making informed decisions regarding chronic diseases [10,20]. According to the U.S.-based Institute of Medicine and the National Research, the vision for 21st century healthcare includes increased attention to cognitive support in decision making [21]. This encompasses computer-based tools and techniques that aid comprehension and cognition. Visualization techniques offer cognitive support by offering mental models of the information through a visual interface [22]. They combine statistical methods and models with advanced interactive visualization methods to help mask the underlying complexity of large health data sets and make evidence-based decisions [23]. Chronic diseases are characterized by high prevalence among populations, rising complication rates, and increased incidence of people with multiple chronic conditions, to name a few. In this scenario, visualization can represent association between preventive measures and disease control, summary health dimensions across diverse patient populations and, timeline of disease prevalence across regions/populations, to offer actionable insights for effective population management and national development [24]. Additionally, visual techniques offer the ability to analyze data at multiple levels and dimensions starting from population to subpopulation to the individual [25]. This paper addresses the challenge of understanding large amounts of data related to chronic diseases by applying visual analytics techniques and producing descriptive analytics. Our overall goal is to gain insight into the data and make policy recommendations. Given that large segments of the U.S. population suffer from one or more chronic disease conditions, a data-driven approach to the analysis of the data has the potential to reveal patterns of association, correlation, and causality. We therefore studied the variables extracted from a highly reliable source, the Centers for Disease Control. Data for variables pertaining to several categories, Int. J. Environ. Res. Public Health 2018, 15, 431 3 of 24 namely chronic condition (“condition” is used interchangeably with “disease”), behavioral health, mental health, preventive health, demographics, overarching conditions, and location for several years (typically 2012 to 2014). We analyzed relationships within each category and across categories to obtain multi-dimensional views and insight into the data. The analytics provide insights and implications that suggest ways for the healthcare system to better manage population health. This paper is organized as follows: Section1 offers an introduction to the research, Section2 discusses the methodology, Section3 presents and discusses the visual charts and results, Section4 contains the scope and limitations of the research, Section5 describes the policy implications and future research, and Section6 presents our conclusions.

2. Materials and Methods This study analyzes the characteristics of chronic diseases in the U.S. and explores the relationships between demographics, behavior habits, and other health conditions and chronic diseases, thereby revealing information for public health practice at the state-specific level. In this data-driven study we use visual analytics [26], conducting primarily descriptive analytics [20] to obtain a panoramic insight into the chronic diseases data set pulled from the Centers for Disease Control and Prevention web site. The discipline of visual analytics aims to provide researchers and policymakers with better and more effective ways to understand and analyze large data sets, while also enabling them to act upon their findings in real time. Visual analytics integrates the analytic capabilities of the computer and the abilities of human analysts, thus inviting novel discoveries and empowering individuals to take control of the analytical process. It sheds light on unexpected and hidden insights, which may lead to beneficial and profitable innovation [27,28]. Driving visual analytics is the aim of turning information overload into opportunity; just as information visualization has changed our view on databases, the goal of visual analytics is to make our way of processing data and information transparent and accessible for analytic discourse. The visualization of these processes provides the means for examining the actual processes and not just the results. Visual analytics applies such technology as business intelligence (BI) tools to combine human analytical skill with computing power. Clearly, this research is highly interdisciplinary, involving such areas as visualization, data mining, data management, data fusion, statistics, and cognitive science, among others. One key understanding of visual analytics is that the integration of these diverse areas is a scientific discipline in its own right [29,30]. Historically, automatic analysis techniques, such as statistics and data mining, were developed independently of visualization and interaction techniques. One of the most important steps in the direction of visual analytics research was the need to move from confirmatory data analysis (using charts and other visual representations to present results) to exploratory data analysis (interacting with the data), first introduced to the statistics research community by John W. Tukey in his book, Exploratory Data Analysis [31]. With improvements in graphical user interfaces and interaction devices, the research community devoted its efforts to information visualization [27]. Eventually, this community recognized the potential of integrating the user’s perspective into the knowledge discovery and data mining process through effective and efficient visualization techniques, interaction capabilities, and knowledge transfer. This led to visual data exploration and visual data mining [29] and widened considerably the scope of applications of visualization, statistics, and data mining—the three pillars of analytics. In visual analytics is defined as “the science of analytical reasoning facilitated by interactive human-machine interfaces” [29]. A more current definition says “visual analytics combines automated analysis techniques with interactive visualizations for an effective understanding reasoning and decision-making on the basis of very large and complex data sets” (both reported in [27]). In their book Illuminating the Path, Thomas and Cook define visual analytics as the science of analytical reasoning facilitated by interactive visual interfaces. One application of visualization is descriptive analytics, the most commonly used and most well understood type of analytics. It was the earliest to be introduced and the easiest by far to Int. J. Environ. Res. Public Health 2018, 15, 431 4 of 24 implement and understand in that it describes data “as is” without complex calculations. Descriptive analytics is more data-driven than other models. Most health data analyses start with descriptive analytics, using data to understand past and current health patterns and trends and to make informed decisions [20]. The models in descriptive analytics categorize, characterize, aggregate, and classify data, converting it into information for understanding and analyzing business decisions, outcomes, and quality. Such data summaries can be in the form of meaningful charts and reports, and responses to queries using SQL. Descriptive analytics uses a significant amount of visualization. One could, for example, obtain standard and customized reports and drill down into the data, running queries to better understand, say, the sales of a product [20]. Descriptive analytics helps answer such questions as: How many patients with diabetes also have obesity? Which of the chronic diseases are more prevalent in different regions of the country? What behavioral habits are correlated to the chronic diseases? Which groups of patients suffer from more than one chronic condition? Is there an association between health insurance (and lack thereof) and chronic diseases? What are cost trade-offs between chronic disease prevention and management? What are typical patient profiles for various chronic diseases? This study concentrates on chronic condition indicators and related demographics, behavior habits, preventive health, and oral health factors. As mentioned, the data source for this study is the Center for Disease Control and Prevention (CDC) [32]. The CDC’s Division of Population Health offers a crosscutting set of 124 indicators that were developed by consensus. Those indicators are integrated from multiple resources, with the help of the Chronic Disease Indicator web site, which serves as a gateway to additional information and data sources. In this research we downloaded secondary data for the United States from the CDC dataset, for the years 2012 to 2014. The data is for states, territories, and large metropolitan areas in the U.S., including the 50 states and District of Columbia, Guam, Puerto Rico, and the U.S. Virgin Islands. Data cleaning, integration, and transformation were conducted on the raw data set. The main categories of variables included—chronic condition, mental health, behavior habits, preventative health, and demographics. In addition, overarching conditions and location were also studied. Table1 summarizes the categories and variables.

Table 1. Chronic diseases and related indicators.

Category Sub-Category Variables (Measure) Deﬁnition Prevalence of diagnosed diabetes among adults aged ≥18 Diabetes Diabetes (%) years—2012–2014 Hospital diabetes (number) Hospitalization with diabetes as diagnosis; 2010 and 2013 Mortality diabetes Mortality rate due to diabetes listed as cause of death, 2010–2014 (per 100,000) Arthritis Arthritis (%) Prevalence of arthritis among adults aged ≥18 years; 2013–2014 Fair or poor health— Prevalence of fair or poor health among adults aged ≥18 years with arthritis (%) arthritis—2013–2014 Prevalence of Arthritis among adults aged ≥18 years who are Obesity—arthritis (%) obese—2013–2014 Current asthma prevalence among adults aged ≥18 years, Asthma Asthma (%) Chronic through 2012–2014 condition Mortality—asthma Asthma mortality rate through 2010–2014 (case per 100,000) Hospital—asthma Hospitalizations for Asthma (case per 100,000) Prevalence of chronic kidney disease among adults Chronic Kidney Disease Kidney (%) aged ≥18 years—2012–2014 Mortality—kidney Mortality with end stage renal disease, through 2010 to 2014 (case per 100,000) Chronic Obstructive Prevalence of chronic obstructive pulmonary disease among adults Pulmonary (%) Pulmonary Disease aged ≥18 years, through 2012 to 2014 Hospital—pulmonary Hospitalization for chronic obstructive pulmonary disease as any (case per 100,000) diagnosis of 2010 and 2013 Mortality—pulmonary Mortality with chronic obstructive pulmonary disease as underlying (case per 100,000) cause among adults aged ≥45 years, through 2010 and 2014. Int. J. Environ. Res. Public Health 2018, 15, 431 5 of 24

Table 1. Cont.

Category Sub-Category Variables (Measure) Definition The crude prevalence rate of at least 14 recent mentally unhealthy Mental health Mental—women (%) days among women aged 18–44 years, through 2012 to 2014 The crude prevalence rate of Postpartum depressive Mental health Postpartum (%) symptoms in 2011 The aged-adjusted mean of recently mentally unhealthy days among Mental (number) adults aged ≥18 years, through 2012 to 2014 Binge drinking prevalence among adults aged ≥18 years, Alcohol Binge drink (%) through 2012 to 2014 Heavy drink (%) Heavy drinking among adults aged ≥18 years, through 2012 to 2014 Nutrition, Physical No leisure-time physical activity among adults aged ≥18 years, Behavioral Activity, and Weight Physical activity (%) through 2012 to 2014 Habits Status Current smokeless tobacco use among adults aged ≥18 years, Tobacco—smokeless (%) through 2012 to 2014 Tobacco (%) Current smoking among adults aged ≥18 years, through 2012 to 2014 Obesity (%) Obesity among adults aged ≥18 years, through 2012 to 2014 Pneumococcal Pneumococcal vaccination among noninstitutionalized adults aged Pneumonia—smoke (%) vaccination 18–64 years who smoke, through 2012 to 2014 Pneumococcal vaccination among noninstitutionalized adults Pneumonia—heart (%) aged 18–64 years with a history of coronary heart disease, through 2012 to 2014 Pneumococcal vaccination among noninstitutionalized adults Pneumonia—asthma (%) aged 18–64 years with asthma, through 2012 to 2014 Pneumococcal vaccination among noninstitutionalized adults Preventive Pneumonia—diabetes (%) health aged 18–64 years with diagnosed diabetes, through 2012 to 2014 Influenza vaccination among noninstitutionalized adults Immunization Influenza—asthma (%) aged 18–64 years with asthma, through 2012 to 2014; Influenza vaccination among noninstitutionalized adults Influenza—diabetes (%) aged 18–64 years with diagnosed diabetes, through 2012 to 2014 Influenza vaccination among noninstitutionalized adults Influenza—heart (%) aged 18–64 years with a history of coronary heart disease or stroke Influenza vaccination among noninstitutionalized adults Influenza (%) aged ≥18 years, through 2012 to 2014 Quit attempts in the past year among current smokers, Smoke Quit (number) through 2012 to 2014 Gender Gender (character) Male and female Demographics Ethnicity Race (character) Race Location State location Location (character) 50 states and District of Columbia, Guam, Puerto Rico, Virgin Islands Current lack of health insurance among adults aged 18–64 years, Overarching Conditions Insurance (%) Overarching through 2012 to 2014 Conditions Poor—self rate (%) Fair or poor self-rated health status among adults aged ≥18 years Sleep (%) Prevalence of sufficient sleep among adults aged ≥18 years

3. Results We use visualization and descriptive analytics to explore chronic conditions, preventive healthcare, mental health, and overarching conditions, with the objective of deciphering relationships and patterns that emerge from the visualization. We would like to point out that since our sample includes adults aged 18 and over our results are applicable for adults in that age group. Figure1 models the average prevalence of diagnosed diabetes among adults aged ≥18 years in the period 2012 to 2014. Puerto Rico leads the pack, followed by Mississippi. Int. J. Environ. Res. Public Health 2018, 15, 431 6 of 24 Int.Int. J. Environ.J. Environ. Res. Res. Public Public Health Health 2018 2018, ,15 15, ,x x 6 6of of 24 24

Figure 1. Diabetes by state. FigureFigure 1 1.. Diabetes by state.state.

As Figure 2 below shows, Puerto Rico has the highest number of citizens among adults aged As Figure Figure 22 below shows, shows, Puerto Puerto Rico Rico has has the the highest highest number number of citizens of citizens among among adults adults aged ≥18 years, in fair or poor health with arthritis for the period 2013 to 2014. Puerto Rico is followed by aged≥18 years,≥18 in years, fair or in poor fair orhealth poor with health arthritis with for arthritis the period for the2013 period to 2014. 2013 Puerto to 2014. Rico Puertois followed Rico by is Tennessee and Mississippi. followedTennessee by and Tennessee Mississippi. and Mississippi.

Figure 2. Arthritis by state. Figure 2.2. Arthritis by state state.. The current asthma prevalence among adults aged ≥18 years for the period 2012 to 2014 is indicated in Figure 3. West Virginia has a higher prevalence of the condition compared to other states. The currentcurrent asthma prevalence among adults aged ≥1818 years years for for the the period period 2012 to 2014 is indicated in Figure 3 3.. WestWest VirginiaVirginia has has aa higherhigher prevalenceprevalence ofof thethe conditioncondition comparedcompared toto other states.

Int. J. Environ. Res. Public Health 2018, 15, 431 7 of 24 Int. J. Environ. Res. Public Health 2018, 15, x 7 of 24 Int. J. Environ. Res. Public Health 2018, 15, x 7 of 24

Figure 3 3.. Asthma by state state.. Figure 3. Asthma by state. With regard to end-stage renal disease, Figure 4 shows that the condition is dispersed widely With regard to end-stage renal disease, Figure4 shows that the condition is dispersed widely amongWith various regard areas. to end -stage renal disease, Figure 4 shows that the condition is dispersed widely among various areas.

Figure 4. End-stage renal disease by region. Figure 4 4.. EndEnd-stage-stage renal disease by region region.. The average value for hospitalization for chronic obstructive pulmonary disease for all The average value for hospitalization for chronic obstructive pulmonary disease for all diagnosesThe average between value 2010 for and hospitalization 2013 is shown for in chronic Figure obstructive 5. Kentucky pulmonary and West disease Virginia for allhave diagnoses higher diagnoses between 2010 and 2013 is shown in Figure 5. Kentucky and West Virginia have higher hospitalizationsbetween 2010 and compared 2013 is shown to other in Figurestates.5 Most. Kentucky of the andareas West are Virginiabelow 45 have cases higher per 100,000. hospitalizations hospitalizations compared to other states. Most of the areas are below 45 cases per 100,000. compared to other states. Most of the areas are below 45 cases per 100,000.

Int. J. Environ. Res. Public Health 2018, 15, 431 8 of 24 Int. J. Environ. Res. Public Health 2018, 15, x 8 of 24

FigureFigure 5. 5.ChronicChronic obstructive obstructive pulmonary disease disease by by state. state.

In exploringIn exploring chronic chronic conditions conditions by location by location in the in U.S. the, U.S., we see we that see some that some conditions, conditions, such as diabetes,such asarthritis, diabetes, and arthritis, obstructive and obstructive pulmonary pulmonary diseases, diseases, are more are moreprevalent prevalent in eastern in eastern states, states, while others,while such others, as asthma, such as occur asthma, more occur often more in oftennortheastern in northeastern states. states.For diabetes, For diabetes, listed listed as a cause as a cause of death for theof deathyears for2010 the to years 2014, 2010 the to states 2014, of the Oklahoma states of Oklahoma and West and Virginia West Virginia had the had relatively the relatively high highaverage thresholdaverage of thresholdover 100 of (age over adjusted 100 (age adjustedrate per rate 100,000). per 100,000). In the In case the caseof asthma, of asthma, West West Virginia Virginia has the the highest prevalence of the condition (among adults), while Maryland, Massachusetts, and New York highest prevalence of the condition (among adults), while Maryland, Massachusetts, and New York had the highest number of hospitalizations. With regard to chronic obstructive pulmonary disease, had the highest number of hospitalizations. With regard to chronic obstructive pulmonary disease, Kentucky and West Virginia had the most hospitalizations compared to other states. The majority Kentuckyof states and are West indeed Virginia below had 45 cases the permost 100,000. hospitalizations With respect compared to arthritis to among other adults,states. aThe majority majority of of statesstates are indeed average below 25%,45 cases with theper exception100,000. ofWith West respect Virginia, to which arthritis averaged among 34.15%. adults, In summary,a majority of statesWest average Virginia below ranks 25%, high with in prevalence the exception for most of West chronic Virginia, conditions, which such averaged as diabetes, 34.15%. asthma, In chronic summary, Westpulmonary Virginia ranks disease, high and in arthritis prevalence when for compared most chronic to all other conditions, states for such the periodas diabetes, 2000 to asthma, 2014. chronic pulmonaryWe disease, looked at and the distributionarthritis when of chronic compared conditions to all byother gender states and for race the to period identify 2000 relevant to 2014. trends andWe looked patterns at (Figures the distribution6 and7). Chronic of chronic conditions conditions differ byby gender. gender Women and race tend to toidentify have signiﬁcantly relevant trends and patternshigher cases (Figures per 100,000 6 and of7). hospitalizations Chronic conditions for asthma. differ Whereas by gender. men Women tend to havetenda to higher have mortalitysignificantly higherrate cases from per chronic 100,000 obstructive of hospitalizations pulmonary for disease, asthma. diabetes, Whereas chronic men kidney, tend to and have other a higher conditions, mortality as shown in Figure6. rate from chronic obstructive pulmonary disease, diabetes, chronic kidney, and other conditions, as We also examined all chronic conditions by race (Figure7) and found that non-Hispanic Blacks shown in Figure 6. have higher mortality rate for pulmonary disease and asthma and a higher hospitalization from diabetes. They are followed by Paciﬁc Islander and American Indians. All categories of arthritis are fairly evenly distributed among Black, non-Hispanic, Multiracial, Whites, and other. Females have a higher hospitalization rate for asthma (per 100,000), while in terms of mortality rate for chronic obstructive pulmonary disease, diabetes, and chronic kidney disease, males have the higher hospitalization rate. Again, American Indian or Alaskan Natives have higher mortality rate for chronic obstructive pulmonary disease, diabetes, and kidney disease. They’re followed by Blacks and non-Hispanics.

Int. J. Environ. Res. Public Health 2018, 15, x 8 of 24

Figure 5. Chronic obstructive pulmonary disease by state.

In exploring chronic conditions by location in the U.S., we see that some conditions, such as diabetes, arthritis, and obstructive pulmonary diseases, are more prevalent in eastern states, while others, such as asthma, occur more often in northeastern states. For diabetes, listed as a cause of death for the years 2010 to 2014, the states of Oklahoma and West Virginia had the relatively high average threshold of over 100 (age adjusted rate per 100,000). In the case of asthma, West Virginia has the highest prevalence of the condition (among adults), while Maryland, Massachusetts, and New York had the highest number of hospitalizations. With regard to chronic obstructive pulmonary disease, Kentucky and West Virginia had the most hospitalizations compared to other states. The majority of states are indeed below 45 cases per 100,000. With respect to arthritis among adults, a majority of states average below 25%, with the exception of West Virginia, which averaged 34.15%. In summary, West Virginia ranks high in prevalence for most chronic conditions, such as diabetes, asthma, chronic pulmonary disease, and arthritis when compared to all other states for the period 2000 to 2014. We looked at the distribution of chronic conditions by gender and race to identify relevant trends and patterns (Figures 6 and 7). Chronic conditions differ by gender. Women tend to have significantly higher cases per 100,000 of hospitalizations for asthma. Whereas men tend to have a higher mortality rate from chronic obstructive pulmonary disease, diabetes, chronic kidney, and other conditions, as shownInt. in J. Environ. Figure Res. 6. Public Health 2018, 15, 431 9 of 24

Int. J. Environ. Res. Public Health 2018, 15, x 9 of 24

Figure 6. Chronic condition by gender.

We also examined all chronic conditions by race (Figure 7) and found that non-Hispanic Blacks have higher mortality rate for pulmonary disease and asthma and a higher hospitalization from diabetes. They are followed by Pacific Islander and American Indians. All categories of arthritis are Figure 6. Chronic condition by gender. fairly evenly distributed among Black, non-Hispanic, Multiracial, Whites, and other.

FigureFigure 7. 7. ChronicChronic conditions conditions byby race. race.

3.1.Females Mental have Health a higher by Gender hospitalization and Race rate for asthma (per 100,000), while in terms of mortality rate for chronicMental obstructive health is an pulmonary important disease, aspect ofdiabetes, national and healthcare chronic kidney impacting disease, chronic males diseases. have the higherWe hospitalization analyzed mental rate. health Again, by gender American (Figure Indian8) and or by Alaskan race (Figure Natives9). When have wehigh examineer mortality how rate for chronicmany days obstructive an individual pulmonary feels “mentally disease, unhealthy” diabetes,for and the kidney years 2012 disease. to 2014, They’re women followed are more by likely Blacks and nonto have-Hispanics. more unhealthy days than men, as shown in Figure8.

3.1. Mental Health by Gender and Race Mental health is an important aspect of national healthcare impacting chronic diseases. We analyzed mental health by gender (Figure 8) and by race (Figure 9). When we examine how many days an individual feels “mentally unhealthy" for the years 2012 to 2014, women are more likely to have more unhealthy days than men, as shown in Figure 8.

Int. J. Environ. Res. Public Health 2018, 15, 431 10 of 24 Int. J. Environ. Res. Public Health 2018, 15, x 10 of 24

Int. J. Environ. Res. Public Health 2018, 15, x 10 of 24

Figure 8. Mental health by gender. Figure 8. Mental health by gender.

Simultaneously, multi-racial, nonFigure-Hispanic 8. Mental women health byin thegender age. group 18 to 44 have a higher crude prevalenceSimultaneously, rate of at multi-racial,least 14 recent non-Hispanic “mentally unhealthy women” days. in the This age group group is followed 18 to 44 by have black a non higher- crudeHispanics. prevalenceSimultaneously, rate multi of at- leastracial, 14 non recent-Hispanic “mentally women unhealthy” in the age group days. 18 This to 44 grouphave a ishigher followed crude by blackprevalence non-Hispanics. rate of at least 14 recent “mentally unhealthy” days. This group is followed by black non- Hispanics.

Figure 9. Mental health by race.

We then studied behavioral habitsFigu rein 9 the. Mental data health set to by gain race insight. into noticeable patterns, if in fact any exist. Figure 9. Mental health by race. We then studied behavioral habits in the data set to gain insight into noticeable patterns, if in fact3.2.We Behavioralany then exist. studied Habits behavioral by Gender and habits Race in the data set to gain insight into noticeable patterns, if in fact any exist.Figure 10 charts behavioral habits by gender. 3.2. Behavioral Habits by Gender and Race 3.2. Behavioral Habits by Gender and Race Figure 10 charts behavioral habits by gender. Figure 10 charts behavioral habits by gender.

Int. J. Environ. Res. Public Health 2018, 15, 431 11 of 24 Int. J. Environ. Res. Public Health 2018, 15, x 11 of 24 Int. J. Environ. Res. Public Health 2018, 15, x 11 of 24

Figure 10. Behavioral habits by gender. Figure 10 10.. BehavioralBehavioral habits by gender gender.. As seen in the chart above, men display higher numbers in the alcohol categories of “binge As seen in the chart above, men display higher numbers in the alcohol categories of “binge drinkingAs seen” and in the“heavy chart drinking above, men”, as display well as higher in “current numbers smokeless in the alcohol” tobacco categories use ofamong “binge adults. drinking” In drinking” and “heavy drinking”, as well as in “current smokeless” tobacco use among adults. In termsand “heavy of engaging drinking”, in “current as well assmoking in “current”, “obesity smokeless””, and tobacco “no leisure use among-time” adults.physical In activity, terms of both engaging men terms of engaging in “current smoking”, “obesity”, and “no leisure-time” physical activity, both men andin “current women smoking”, experience “obesity”, similar and complications, “no leisure-time” that physical highlights activity, the both need men for and positive women experience behavior and women experience similar complications, that highlights the need for positive behavior modification.similar complications, that highlights the need for positive behavior modification. modification. Figure 11 illustrates the analysis of behavioral habits by race. Figure 11 illustrates the analysis of behavioral habits by race.

Figure 11. Behavioral habits by race. Figure 11 11.. Behavioral habits by race race..

Figure 11 reveals that for the behavioral habits of “obesity” and “no leisure-time” physical Figure 11 reveals that for the behavioral habits of “obesity”“obesity” and “no“no leisure-time”leisure-time” physical activity among adults aged 18 and over, the black non-Hispanic and Hispanic races have the highest activity among adults aged 18 and over, over, the black non non-Hispanic-Hispanic and Hispanic races have the highest frequency, while white non-Hispanics have the lowest. By and large, in most behavioral habits, the frequency, while while white white non non-Hispanics-Hispanics have have the the lowest. lowest. By By and and large, large, in most in most behavioral behavioral habits, habits, the other non-Hispanics have the lowest frequency. otherthe other non non-Hispanics-Hispanics have have the lowest the lowest frequency. frequency.

Int. J. Environ. Res. Public Health 2018, 15, 431 12 of 24 Int. J. Environ. Res. Public Health 2018, 15, x 12 of 24

3.3.3.3.Int. Preventive J.Preventive Environ. Res. Health Health Public and Health and ChronicChronic 2018, 15 Conditions, Conditionsx 12 of 24

3.3.We PreventiveWe analyzed analyzed Health the the dataand data Chronic to todetect detect Conditions associations associations between between demographics demographics and and preventive preventive health. health. As FigureAs Figure 12 indicates, 12 indicates, both both men men and and women women appear appear to to engage engage in in preventive health,health, though though women women We analyzed the data to detect associations between demographics and preventive health. As havehave the the edge. edge. With With regard regard to torace, race, Blacks Blacks and and Hispanics Hispanics engage engage less less in in preventive preventive health health overall, overall, as Figure 12 indicates, both men and women appear to engage in preventive health, though women shownas shown in Figure in Figure 13. 13. have the edge. With regard to race, Blacks and Hispanics engage less in preventive health overall, as shown in Figure 13.

Figure 12. Preventive health by gender. Figure 12 12.. Preventive health by gender gender..

Figure 13. Preventive health by race. FigureFigure 1 13.3. PreventivePreventive health byby race.race. While all chronic conditions are debilitating on the economy, for the sake of scope, we selectively While all chronic conditions are debilitating on the economy, for the sake of scope, we selectively analyzeWhile the all influencechronic conditions of a few conditions are debilitating such as on diabetes the economy, and asthma. for the By sake 2034, of the scope, population we select withively analyze the inﬂuence of a few conditions such as diabetes and asthma. By 2034, the population with analyzediabetes the is influence expected of to a increase few conditions by 100% such and asthe diabetes cost expected and asthma. to increase By 2034, by 53% the [3population3]. Figure with14 diabetes is expected to increase by 100% and the cost expected to increase by 53% [33]. Figure 14 diabetesdepicts is the expected association to increasebetween bydiabetes 100% and and pneumococcal the cost expected vaccination to increase for diabetes by 53%. [33]. Figure 14 depicts the association between diabetes and pneumococcal vaccination for diabetes. depicts the association between diabetes and pneumococcal vaccination for diabetes.

Int. J. Environ. Res. Public Health 2018, 15, 431 13 of 24 Int. J. Environ. Res. Public Health 2018, 15, x 13 of 24

Int. J. Environ. Res. Public Health 2018, 15, x 13 of 24

Figure 14.FigureDiagnosed 14. Diagnosed diabetes diabetes ratio by ratio pneumococcal by pneumococcal vaccination vaccination ratio. ratio (#: number). . (#: number).

AsFigure indicated 14. Diagnosed in Figure diabetes 14, ratio there by is pneumococcal a significant vaccination negative ratio relationship. (#: number between). the average As indicated in Figure 14, there is a significant negative relationship between the average pneumococcal vaccination among diabetes patients and the average diagnosed diabetes ratio among pneumococcalAs indicated vaccination in Figure among 14, therediabetes is a patients significant and negative the average relationship diagnosed between diabetes the ratio average among the population (p < 0.0001). As the average pneumococcal vaccination among diabetes patients thepneumococcal population (vaccinationp < 0.0001). among As the diabetes average patients pneumococcal and the average vaccination diagnosed among diabetes diabetes ratio among patients increases, the average diagnosed diabetes ratio decreases (fewer cases of diabetes). Given the the population (p < 0.0001). As the average pneumococcal vaccination among diabetes patients increases,importance the average of asthma diagnosed as another diabetes prevalent ratio chronic decreases condition, (fewer we cases decided of to diabetes). analyze the Given relationship the increases, the average diagnosed diabetes ratio decreases (fewer cases of diabetes). Given the importancebetween of asthma the mortality as another ratio prevalent and influenza chronic vaccinations condition, we for decided asthma to to analyze determine the therelationship efficiency of importance of asthma as another prevalent chronic condition, we decided to analyze the relationship betweenpreventive the mortality measures ratio (Figure and influenza 15). vaccinations for asthma to determine the efficiency of preventivebetween measuresthe mortality (Figure ratio 15 and). influenza vaccinations for asthma to determine the efficiency of preventive measures (Figure 15).

Figure 15. Mortality ratio for asthma and influenza vaccination for asthma.

FigureFigure 15. 15Mortality. Mortality ratioratio forfor asthma and influenza inﬂuenza vaccination vaccination for for asthma asthma..

Int. J. Environ. Res. Public Health 2018, 15, x 14 of 24 Int. J. Environ. Res. Public Health 2018, 15, 431 14 of 24 Figure 15 shows a significant negative association (p < 0.0001): as the rate of influenza vaccination for asthmaFigure increases, 15 shows the a significant mortality negative ratio of associationasthma declines. (p < 0.0001): Analysis as the of rate the of above influenza preventive vaccination health variablesfor asthma shows increases, that resources the mortality and efforts ratio of dedicated asthma declines. to preventive Analysis healthcare of the above offer prompreventiveise. The importancehealth variables of managing shows that chronic resources diseases and efforts is also dedicated highlighted to preventive when we healthcare examine offer the promise.association betweenThe importance behavioral of habits managing and chronicoverarching diseases conditions. is also highlighted when we examine the association between behavioral habits and overarching conditions. 3.4. Behavioral Health and Overarching Conditions 3.4. Behavioral Health and Overarching Conditions Overarching conditions represent situations or factors that directly or indirectly influence the area ofOverarching study. In our conditions research represent we look situations at the influence or factors of that these directly conditions or indirectly on chronic influence diseases, the behavioralarea of study. health, In and our preventive research we health. look The at the overarching influence conditions of these conditions include lack on chronicof health diseases, insurance (%),behavioral self-rated health, health and status preventive (good,health. fair, poor), The overarching and prevalence conditions of sufficient include lacksleep of (%) health for insurance which data (%), self-rated health status (good, fair, poor), and prevalence of sufficient sleep (%) for which data was available. was available. We explored the association of self-assessed health statuses among adults with the behavioral We explored the association of self-assessed health statuses among adults with the behavioral habits of binge drinking and heavy drinking (Figure 16). habits of binge drinking and heavy drinking (Figure 16).

FigureFigure 16 16.. Binge/heavyBinge/heavy drink drink and poor self-ratedself-rated health health status. status.

ThereThere is is a a significant significant negative negative correlation correlation (p < 0.0001) (p between< 0.0001) binge between drinking bingeand self-assessment drinking and selfof-assessment health. That of is health. to say thatThat theis to lower say thethat health the lower self-assessment, the health self the- higherassessment, is the the percentage higher is of the percentagebinge drinking. of binge A drinking. decrease ofA lessdecrease than 1%of less (0.69%) than in 1% self-assessed (0.69%) in self health-assessed is associated health withis associated a 1% withincrease a 1% increase in binge drinking.in binge drinking. Likewise, Likewise, there is a significantthere is a significant negative association negative association (p < 0.0001) between(p < 0.0001) betweenself-assessment self-assessment of health and of health percentage and of pe heavyrcentage drinking. of heavy A decrease drinking. of 1.6% A in self-assesseddecrease of health 1.6% in selfis-assessed associated health with ais 1%associated increase with in heavy a 1% drinking. increase Wein heavy can surmise drinking. that We reduced can surmise self-assessment that reduced of selfhealth-assessment has a stronger of health influence has a on stronger heavy drinking influence than on binge heavy drinking drinking among than adults. binge drinking among adults. Next, we looked at the association between current smoking prevalence and presence of sufficient sleepNext, among we adults looked (Figure at the 17 ). association between current smoking prevalence and presence of sufficient sleep among adults (Figure 17).

Int. J. Environ. Res. Public Health 2018, 15, 431 15 of 24

Int. J. Environ. Res. Public Health 2018, 15, x 15 of 24

FigureFigure 17. Current 17. Current smoking smoking prevalence prevalence by presence of ofsufficient sufﬁcient sleep sleep among among adults adults..

Figure 17 above shows a significant negative association (p < 0.0001) between prevalence of Figurecurrent 17 smoking aboveFigure and shows 17. prevalenceCurrent a significant smoking of sufficient prevalence negative sleep. by presence When association current of sufficient smoking (p < sleep 0.0001) amongprevalence betweenadults decreases. prevalence by of currentless smoking than 1% and (0.38%), prevalence the prevalence of sufficient of sufficient sleep. sleep When increases current by 1%. smoking prevalence decreases by Figure 17 above shows a significant negative association (p < 0.0001) between prevalence of less than 1%The (0.38%), relationship the prevalence between poo ofr self sufficient-rated health sleep status increases and obesity by 1%. is positive (Figure 18). The current smoking and prevalence of sufficient sleep. When current smoking prevalence decreases by Thehigher relationship the prevalence between of fair or poor poor self-ratedself-rated health, health the status higher is and the obesityprevalence is of positive obesity. When (Figure 18). poorless than self- rated1% (0.38%), health theincreases prevalence by 1%, of thesufficient prevalence sleep of increases obesity increasesby 1%. by 0.468779%. The higherThe the relationship prevalence between of fair poo or poorr self-rated self-rated health health,status and the obesity higher is positive is theprevalence (Figure 18). The of obesity. When poorhigher self-rated the prevalence health of increasesfair or poor by self 1%,-rated the health, prevalence the higher of obesityis the prevalence increases of byobesity. 0.468779%. When poor self-rated health increases by 1%, the prevalence of obesity increases by 0.468779%.

Figure 18. Obesity by poor self-rated health status.

FigureFigure 18. 18Obesity. Obesity by by poorpoor self self-rated-rated health health status status..

Similarly, poor self-rated health has a positive association with current smoking, as indicated in Figure 19. As the prevalence of poor self-rated health increases by 1%, the prevalence of current smoking increases by 0.30425%. Int. J. Environ. Res. Public Health 2018, 15, x 16 of 24

Int. J. Environ. Res. Public Health 2018, 15, x 16 of 24 Int. J. Environ.Similarly, Res. Public poor Health self2018-rated, 15 health, 431 has a positive association with current smoking, as indicated in 16 of 24 Figure 19. As the prevalence of poor self-rated health increases by 1%, the prevalence of current smokingSimilarly, increases poor by self 0.30425-rated% health. has a positive association with current smoking, as indicated in Figure 19. As the prevalence of poor self-rated health increases by 1%, the prevalence of current smoking increases by 0.30425%.

FigureFigure 19. 19Smoking. Smoking by self self-rated-rated health health status status..

Figure 19. Smoking by self-rated health status. 3.5. Chronic3.5. Chronic Conditions Conditions and and Overarching Overarching Conditions Conditions In the analysis of various chronic conditions, there are significant clusters of conditions among In3.5. the Chronic analysis Conditions of various and Overarching chronic conditions, Conditions there are signiﬁcant clusters of conditions among men and women, such as the prevalence of asthma, with the women tending to have a higher menprevalence andIn women, the analysisof asthma such of than asvarious the men. prevalencechronic Regarding conditions, ofsuch asthma, chronic there arewithconditions significant the womenas diabetes,clusters tending of there conditions is to a significant have among a higher prevalencepositivemen and of relationship asthma women, than such (p men. as< 0.001) the Regarding prevalence between such lack of asthma, ofchronic health with conditions insurance the women andas diabetes, tending prevalence to there have of diagnosedis a a hig signiﬁcanther positivediabetesprevalence relationship (Figure of asthma 20). (p < 0.001)than men. between Regarding lack ofsuch health chronic insurance conditions and as prevalence diabetes, there of diagnosed is a significant diabetes (Figurepositive 20). relationship (p < 0.001) between lack of health insurance and prevalence of diagnosed diabetes (Figure 20).

Figure 20. Current lack of health insurance by diagnosed diabetes.

We notice in Figure 20 that the distribution of lack of health insurance is sparse compared to that of diagnosed diabetes among adults aged 18 and older. Likewise, for chronic kidney disease (Figure 21) there is a signiﬁcant positive relationship (p < 0.0001) with lack of health insurance. Int. J. Environ. Res. Public Health 2018, 15, x 17 of 24

Int. J. Environ. Res. Public Health 2018, 15, x 17 of 24 Figure 20. Current lack of health insurance by diagnosed diabetes. Figure 20. Current lack of health insurance by diagnosed diabetes. We notice in Figure 20 that the distribution of lack of health insurance is sparse compared to that Int. J. Environ. Res. Public Health 2018, 15, 431 17 of 24 of diagnosedWe notice diabetes in Figure among 20 that adults the ageddistribution 18 and older.of lack Likewise, of health forinsurance chronic iskidney sparse disease compared (Figure to that 21) thereof diagnosed is a significant diabetes positive among relationshipadults aged 18 (p and < 0.0001) older. with Likewise, lack of for health chronic insurance. kidney disease (Figure 21) there is a significant positive relationship (p < 0.0001) with lack of health insurance.

FigureFigure 21. 21Lack. Lack of of insuranceinsurance by by chronic chronic kidney kidney disease disease.. Figure 21. Lack of insurance by chronic kidney disease. The relationship between lack of insurance and hospitalization for chronic pulmonary disease is The relationship between lack of insurance and hospitalization for chronic pulmonary disease positiveThe and relationship significant between (p < 0.0001), lack of asinsurance shown inand Figure hospitalization 22. An increase for chronic in the pulmonary lack of insurance disease is is positive and signiﬁcant (p < 0.0001), as shown in Figure 22. An increase in the lack of insurance is associatedpositive and with significant an increase (p

Figure 22. Lack of insurance by pulmonary disease.

3.6. Association between Chronic Conditions We analyzed for any associations between different chronic conditions. It is important to incorporate gender as a factor in the association and prevalence of chronic diseases, so as to Int. J. Environ. Res. Public Health 2018, 15, x 18 of 24

Figure 22. Lack of insurance by pulmonary disease. Int. J. Environ. Res. Public Health 2018, 15, 431 18 of 24 3.6. Association between Chronic Conditions We analyzed for any associations between different chronic conditions. It is important to developincorporate customized gender plans as a for factor diagnoses in the associatio and treatments.n and prevalence A linear of chronic trend modeldiseases, was so as developed to develop for the relationshipcustomized between plans asthma for diagnoses and diabetes and treatments. (Figure 23 A). linear trend model was developed for the relationship between asthma and diabetes (Figure 23).

FigureFigure 23. 23.Asthma Asthma by by diabetes diabetes..

The model in Figure 23 shows a significant negative relationship (p < 0.01) between asthma and Thediabetes model. We in can Figure see gender23 shows clusters a signiﬁcant for the prevalence negative of relationship asthma. Women (p

FigureFigure 24. 24.Diabetes Diabetes by kidney kidney disease disease..

Figure 24 shows a moderate, positive association (p < 0.01) between prevalence of kidney disease and diabetes. As the prevalence of diagnosed diabetes increases by 1%, the prevalence of chronic kidney disease increases by 0.09%. There are no obvious differences in gender here. The association between diabetes and chronic pulmonary disease is shown in Figure 25, and that between arthritis and asthma is shown in Figure 26.

Figure 25. Diabetes by obstructive pulmonary disease.

In Figure 25, we find a significant positive association between diabetes and chronic pulmonary disease (p < 0.001).

Int. J. Environ. Res. Public Health 2018, 15, x 19 of 24

Int. J. Environ. Res. Public Health 2018, 15, 431 19 of 24

Figure 24. Diabetes by kidney disease. Figure 24 shows a moderate, positive association (p < 0.01) between prevalence of kidney disease and diabetes.Figure 24 shows As the a prevalence moderate, positive of diagnosed association diabetes (p < increases 0.01) between by 1%, prevalence the prevalence of kidney of chronicdisease kidneyand diabetes. disease As increases the prevalence by 0.09%. of There diagnosed are no diabetes obvious increases differences by in 1%, gender the here.prevalence of chronic kidneyThe disease association increases between by 0.09 diabetes%. There and are chronic no obvious pulmonary differences disease in is gender shown here. in Figure 25, and that betweenThe arthritisassociation and between asthma diabetes is shown and in Figure chronic 26 pulmonary. disease is shown in Figure 25, and that between arthritis and asthma is shown in Figure 26.

Figure 25.25. Diabetes by obstructive pulmonary disease.disease.

In Figure 2525,, we ﬁndfind a significant signiﬁcant positive association between diabetes and chronic pulmonary disease (p < 0.001). Int. J. Environ. Res. Public Health 2018, 15, x 20 of 24

Figure 26.26. Arthritis by asthma asthma..

When it it comes comes to to prevalence prevalence of arthritisof arthritis and and asthma, asthma, there there clearly clearly are clusters are clusters for men for and men women, and aswomen, shown as in shown Figure in26 . Figure There is 26. a positiveThere is association a positive suchassociation that an such increase that of an 1% increase in prevalence of 1% inof arthritisprevalence is associatedof arthritis with is associated a 0.4% increase with a in0.4% prevalence increase ofin asthma.prevalence of asthma. Figure 27 shows the association between arthritis and chronic pulmonarypulmonary disease.disease.

Figure 27. Arthritis by chronic obstructive pulmonary disease.

Although there are no defined clusters for men and women with regard to the prevalence of arthritis and chronic obstructive pulmonary disease, there is a significant positive association (p < 0.0001), as Figure 27 illustrates. An increase of 1% in prevalence of arthritis is associated with an increase of 0.3% in chronic obstructive pulmonary disease.

3.7. Summary of Results The visual analytics figures above offer insight into a representative cross section of the data. They provide a bird’s eye view of the dimensions and correlations of chronic diseases “conditions”, behavioral health, and preventive health condition in the U.S. In addition, associations between mental health and chronic conditions, preventive health and chronic conditions, and among chronic conditions themselves highlight the dynamics of interplay between these categories. This

Int. J. Environ. Res. Public Health 2018, 15, x 20 of 24

Figure 26. Arthritis by asthma.

When it comes to prevalence of arthritis and asthma, there clearly are clusters for men and women, as shown in Figure 26. There is a positive association such that an increase of 1% in Int. J. Environ. Res. Public Health 2018, 15, 431 20 of 24 prevalence of arthritis is associated with a 0.4% increase in prevalence of asthma. Figure 27 shows the association between arthritis and chronic pulmonary disease.

Figure 27 27.. Arthritis by chronic obstructive pulmonary disease.disease.

Although there there are are no no defined deﬁned clusters clusters for for men men and and women women with with regard regard to tothe the prevalence prevalence of ofarthr arthritisitis and and chronic chronic obstructive obstructive pulmonary pulmonary disease, disease, there there is is a a significant signiﬁcant positive association (p < 0.0001), 0.0001), as as Figure Figure 27 27 illustrates. illustrates. An An increase increase of of 1% 1% in inprevalence prevalence of ofarthritis arthritis is associated is associated with with an anincrease increase of 0.3% of 0.3% in chronic in chronic obstructive obstructive pulmonary pulmonary diseas disease.e.

3.7.3.7. Summary of Results The visual analytics figures ﬁgures above offer insight into into a a representative cross cross section section of of the the data. data. They provide a bird’s eye view of the dimensions and correlations of chronic diseases “ “conditions”,conditions”, behavioral health, health, and and preventive preventive health health condition condition in the in U.S. the InU.S. addition, In addition, associations associations between between mental healthmental and health chronic and conditions,chronic conditions, preventive preventive health and health chronic and conditions, chronic conditions, and among and chronic among conditions chronic themselvesconditions themselveshighlight the highlight dynamics the of interplay dynamics between of interplay these categories. between these This understanding categories. This is useful to policymakers in framing appropriate health policies. Preventive healthcare and mental health are both important elements in the management, mitigation, and prevention of chronic conditions. By exploring these in the context of chronic conditions, we offer insight on allocation and prioritization of resources in mitigation and prospective eradication of chronic diseases at a national level. Overarching conditions, including a lack of health insurance, inﬂuence the access to necessary health services, including preventive care. This lack of availability is associated with poor health and the prevalence of chronic diseases. Similarly, self-assessed health status is a good indicator of overall health status, correlating with subsequent health service use, functional status, and mortality [36]. Poor mental health interferes with social functioning as well as health condition and should therefore be monitored in chronic disease mitigation. Experiencing activity limitation due to poor physical or mental health undermines efforts to achieve a healthy lifestyle and therefore should be addressed at individual, state, and national levels.

4. Scope and Limitations Our research has a few limitations. First, our study is cross-sectional and covers only the years 2012 to 2014, the years for which data is available. Second, we included only a limited set of variables (indicators) from the large data repository on the CDC website. A more comprehensive study could draw from other sources and a larger set of variables. Third, as population and public health have emerged as key disciplines in the contemporary health ecosystem, more scalable, macro-level, and drill-down studies would inform greater understanding of chronic diseases. Fourth, one would Int. J. Environ. Res. Public Health 2018, 15, 431 21 of 24 assume that the quality of publicly available data is high and error-free. Lastly, the study is limited to examining associations and correlations and does not investigate causality. Furthermore, we only apply visual analytics and descriptive analytics, which have limitations in and of themselves.

5. Implications This study has analyzed chronic conditions in conjunction with several demographic variables, including gender and race. There are widespread variations in the prevalence of diverse chronic diseases, the number of hospitalizations for specific diseases, and the diagnosis and mortality rates for different states. For some chronic diseases—such as diabetes, arthritis, and obstructive pulmonary —the prevalence in the east is higher than in other regions, while, there is higher prevalence for other conditions, such as asthma, in the northeast. The south and midwest also show their own prevalence of chronic diseases. Likewise, there are variations for hospitalization and mortality rates. In addition, there are gender differences related to chronic conditions. For example, women tend to have higher cases per 100,000 for asthma-related hospitalizations. Men, on the other hand, appear to have higher mortality rates for chronic obstructive pulmonary disease, diabetes, chronic kidney, and others. Also, when we examined chronic conditions by race, we noticed that American Indian or Alaska Natives had higher mortality rates for chronic obstructive pulmonary disease, diabetes, chronic kidney, and so on, followed by Black and non-Hispanic groups. In addition, the study analyzed demographics of mental health, behavior habits, and preventive health. The associations between behavioral health and chronic conditions and between preventive health and chronic conditions were also analyzed. There is a positive relationship between average female coronary heart disease mortality ratio and average female tobacco use ratio. There is a negative relationship between the average pneumococcal vaccination among diabetes patients and the average diagnosed diabetes ratio among the population. Referring to the relationship between behavioral health and overarching conditions, the study found a negative correlation between age-adjusted prevalence percentage of fair or poor self-rated health status among adults aged ≥18 years and binge drinking adults. The current smoking prevalence and sufficiency of sleep among adults is negatively related. The current lack of health insurance is negatively related to both prevalence of current smoking and that of current smokeless tobacco use. The relationship between obesity and poor self-rated health status is positively related. Similarly, current smoking prevalence has a strong, positive correlation with fair or poor self-rated health status. There are different negative or positive correlations between overarching conditions and chronic conditions. For instance, there is a significant positive relationship between the prevalence of a lack of health insurance and that of diagnosed diabetes. But the relationship between prevalence of a lack of health insurance and prevalence of asthma is negatively related. Finally, we conducted analyses of the differences among chronic conditions. There are obvious clusters between men and women for asthma, although women tend to have a higher prevalence of asthma than men. Overall, prevalence of asthma is negatively related to the prevalence of diabetes. There is a moderate, positive correlation between prevalence of kidney and diabetes, which is akin to the positive correlation between the prevalence of chronic obstructive pulmonary disease and diabetes, arthritis and asthma, arthritis and chronic obstructive pulmonary disease, and asthma and chronic obstructive pulmonary.

6. Conclusions The study makes multiple essential contributions to chronic disease analysis at the patient/ physician and the state levels. At the patient level, analysis of chronic conditions and related behavioral factors allows patients to be proactive in managing their conditions as well as modifying behavioral health. In this day and age, patients are eager to assimilate health information from various sources [37,38]. Being informed allows patients to self-monitor and seek appropriate and timely medical care [39,40], contributing to an ultimate care model that is increasingly personalized. Int. J. Environ. Res. Public Health 2018, 15, 431 22 of 24

Similar to patients, physicians too have varying information needs in healthcare that need to be satisfied [41]. To physicians, information on chronic conditions and more importantly, associations between multiple conditions and between categories of healthcare, enable developing personalized treatment plans based on patient-specific profiles that integrate various symptoms with environmental and other health data [42]. Additionally, the array of information increases their ability to guide patients in towards lifestyle medicine (making lifestyle changes in healthy diet, exercise etc.) in the management of chronic diseases [43]. The road from sickness to wellness requires integrated efforts from physicians and patients—physicians can coach and guide the patients but the ultimate cross-over to wellness lies in the patients’ hands. Whereas most studies on chronic diseases focus on specific chronic diseases and are somewhat limited, this study offers comprehensive analysis over multiple categories of chronic diseases at the state-level. By utilizing visual analytics and descriptive analytics, our study offers methods for gaining insight into the relationships between behavior habits, preventative health and demographics, and chronic conditions. Moreover, this study contributes in terms of the methodology of analytics used in the research. It demonstrates the efficacy of data-driven analytics, which can help make informed decisions on chronic diseases. Going forward, more theoretical and empirical research is needed. Additional studies can address the relationship between chronic disease conditions and other indicators, such as economic, financial, and social. While chronic disease management has become the focus in modern medicine as our population ages and medical costs continue to rise, research should focus on preventive and mitigating policies. The benefits of prevention and its potential to reduce costs and improve outcomes have received the attention of insurance companies, health care plans, and the U.S. Congress. Healthcare systems are now incentivized to reduce readmissions and physicians are encouraged to meet evidence-based quality measures to provide the best outcomes for patients with chronic disease states.

Author Contributions: Both the authors contributed equally to the data analysis, design, and development of the manuscript. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. Basu, J.; Avila, R.; Ricciardi, R. Hospital readmission rates in U.S. States: Are readmissions higher where more patients with multiple chronic conditions cluster? Health Serv. Res. 2016, 51, 1135–1151. [PubMed] 2. Buttorff, C.; Ruder, T.; Bauman, M. Multiple Chronic Conditions in the United States. Available online: https://www.rand.org/pubs/tools/TL221.html (accessed on 1 January 2018). 3. American Association of Retired Persons. Chronic Conditions among Older Americans. Available online: https://assets.aarp.org/rgcenter/health/beyond_50_hcr_conditions.pdf (accessed on 1 January 2018). 4. Fried, L. America’s Health and Health Care Depend on Preventing Chronic Disease. Available online: https://www.hufﬁngtonpost.com/entry/americas-health-and-healthcare-depends-on-preventing_ us_58c0649de4b070e55af9eade (accessed on 31 December 2017). 5. Tinker, A. How to Improve Patient Outcomes for Chronic Diseases and Comorbidities. Available online: http://www.healthcatalyst.com/wp-content/uploads/2014/04/How-to-Improve-Patient-Outcomes.pdf (accessed on 30 December 2017). 6. Comlossy, M. Chronic Disease Prevention and Management; National Conference of State Legislatures: Denver, CO, USA, 2013. 7. Centers for Disease Control. The Power of Prevention: Chronic Disease the Public Health Challenge of the 21st Century. Available online: www.cdc.gov/chronicdisease/pdf/2009-Power-of-Prevention.pdf (accessed on 31 December 2017). Int. J. Environ. Res. Public Health 2018, 15, 431 23 of 24

8. O’Grady, M.J.; Capretta, J.C. Health-Care Cost Projections for Diabetes and Other Chronic Diseases. Available online: http://www.civicenterprises.net/MediaLibrary/Docs/HR%20-%20health%20care% 20cost%20projections%20for%20diabetes%20and%20other%20chronic%20diseases.pdf (accessed on 30 December 2017). 9. World Health Organization. Ten Facts about Chronic Disease. Available online: http://www.who.int/ features/factfiles/chp/10_en.html (accessed on 31 December 2017). 10. Becker’s Hospital Review. The Key to Population Health: Know Your Chronic Disease Patients and Coach Them. Available online: https://www.beckershospitalreview.com/healthcare-information-technology/ the-key-to-population-health-know-your-chronic-disease-patients-and-coach-them.html (accessed on 31 December 2017). 11. National Association of Chronic Disease Directors. Why Public Health is Necessary to Improve Healthcare. Available online: http://www.chronicdisease.org/?page=whyweneedph2imphc (accessed on 31 December 2017). 12. Trotter, P.; Lobelo, F.; Heather, A.J. Chronic Disease Is Healthcare’s Rising Risk. Available online: https://www.healthitoutcomes.com/doc/chronic-disease-is-healthcare-s-rising-risk-0001 (accessed on 31 December 2017). 13. The Growing Crisis of Chronic Disease in the United States. Available online: https://www. fightchronicdisease.org/sites/default/files/docs/GrowingCrisisofChronicDiseaseintheUSfactsheet_ 81009.pdf (accessed on 31 December 2017). 14. Anderson, G.; Horvath, J. The growing burden of chronic disease in America. Public Health Rep. 2004, 119, 263–270. [CrossRef][PubMed] 15. Beaton, T. Top 10 Most Expensive Chronic Diseases for Healthcare Payers 2017. Available online: https: //healthpayerintelligence.com/news/top-10-most-expensive-chronic-diseases-for-healthcare-payers (accessed on 1 January 2018). 16. Guidestone. The Cost of Chronic Disease and Obesity 2011. Available online: https://www.guidestone. org/-/media/Insurance/WorksiteWellness/CostOfChronicDiseaseAndObesity.pdf?la=en (accessed on 30 December 2017). 17. Healthcare Information and Management Systems Society. Keeping Communities Healthy: Wellness and Chronic Disease Management Initiatives. Available online: http://www.himssanalytics.org/news/top-pop- health-initiatives-essentials-brief-update (accessed on 1 January 2018). 18. U.S. Department of Labor, Bureau of Labor Statistics. Consumer Expenditure Survey. Available online: http://www.bls.gov/cex/#overview (accessed on 1 January 2018). 19. U.S. Department of Health and Human Services. Multiple Chronic Conditions—A Strategic Framework: Optimum Health and Quality of Life for Individuals with Multiple Chronic Conditions; U.S. Department of Health and Human Services: Washington, DC, USA, 2010. 20. Raghupathi, W.; Raghupathi, V. An overview of health analytics. J. Health Med. Inform. 2013, 4, 1–11. [CrossRef] 21. Committee on Engaging the Computer Science Research Community in Health Care Informatics; Nat’l Research Council. Computational Technology for Effective Healthcare: Immediate Steps and Strategic Directions; Stead, W.W., Lin, H.S., Eds.; National Academies Press: Washington, DC, USA, 2009. 22. Khan, M.; Khan, S.S. Data and information visualization methods, and interactive mechanisms: A survey. Int. J. Comp. Appl. 2011, 34, 1–14. 23. Caban, J.J.; Gotz, D. Visual analytics in healthcare—Opportunities and research challenges. J. Am. Med. Assoc. 2015, 22, 260–262. [CrossRef][PubMed] 24. Gotz, D.; Borland, D. Data-driven healthcare: Challenges and opportunities for interactive visualization. IEEE Comp. Graphic Appl. 2016, 36, 90–96. [CrossRef][PubMed] 25. Harle, C.; Neill, D.; Padman, R. Development and evaluation of an information visualization system for chronic disease risk assessment. IEEE Intell. Syst. 2012, 27, 81–85. [CrossRef] 26. Sun, G.-D.; Wu, Y.-C.; Liang, R.-H.; Liu, S.-X. A survey of visual analytics techniques and applications: State-of-the-art research and future challenges. J. Comput. Sci. Technol. 2013, 28, 852–867. [CrossRef] 27. Keim, D.; Kohlhammer, J.; Ellis, G.; Mansman, F. Solving Problems with Visual Analytics. Available online: http://www.vismaster.eu/wp-content/uploads/2010/11/VisMaster-book-lowres.pdf (accessed on 1 January 2018). Int. J. Environ. Res. Public Health 2018, 15, 431 24 of 24

28. Thomas, J.; Cook, K. Illuminating the Path: Research and Department Agenda for Visual Analytics; United States Department of Homeland Security: Washington, DC, USA, 2005. 29. Keim, D.A. Visual exploration of large data sets. Commun. ACM 2001, 44, 38–44. [CrossRef] 30. Wong, P.C.; Thomas, J. Visual analytics—Guest editors’ introduction. IEEE Trans. Comput. Graphics Appl. 2004, 24, 20–21. [CrossRef] 31. Tukey, J.W. Exploratory Data Analysis; Addison-Wesley: Boston, MA, USA, 1977. 32. Centers for Disease Control and Prevention. Leading Indicators for Chronic Diseases and Risk Factors. Available online: https://chronicdata.cdc.gov/ (accessed on 30 December 2017). 33. Bodenheimer, T.; Chen, E.; Bennett, H.D. Confronting the growing burden of chronic disease: Can the U.S. health care workforce do the job? Health Aff. 2016, 28, 64–74. 34. Keselman, A.; Heller, N. Estrogen signaling modulates allergic inﬂammation and contributes to sex differences in asthma. Front. Immunol. 2015, 6, 568. [CrossRef][PubMed] 35. Zein, J.G.; Erzurum, S.C. Asthma is different in Women. Curr. Allergy. Asthma Rep. 2015, 15, 28. [CrossRef] [PubMed] 36. Centers for Disease Control and Prevention. MMWR Recommendations and Reports. Available online: https://www.cdc.gov/mmwr/preview/mmwrhtml/rr5311a6.htm (accessed on 7 February 2018). 37. Amante, D.J.; Hogan, T.P.; Pagoto, S.L.; English, T.M.; Lapane, K.L. Access to care and use of the Internet to search for health information: Results from the U.S. National Health Interview Survey. J. Med. Internet Res. 2015, 17, 106. [CrossRef][PubMed] 38. Eysenbach, G.; Köhler, C. How do consumers search for and appraise health information on the world wide web? Qualitative study using focus groups, usability tests, and in-depth interviews. Br. Med. J. 2002, 324, 573–577. 39. Bodenheimer, T.; Wagner, E.H.; Grumbach, K. Improving primary care for patients with chronic illness. J. Am. Med. Assoc. 2002, 288, 1775–1779. [CrossRef] 40. Himes, B.E.; Weitzman, E.R. Innovations in health information technologies for chronic pulmonary diseases16. Respir. Res. 2016, 17, 1–7. [CrossRef][PubMed] 41. Kourouthanassis, P.E.; Mikalef, P.; Ioannidou, M.; Pateli, A. Exploring the online satisfaction gap of Medical doctors: An expectation-conﬁrmation investigation of information needs. Springer 2014, 820, 217–228. 42. Mikalef, P.; Kourouthanassis, P.E.; Pateli, A.G. Online information search behaviour of physicians. Health Inf. Lib. J. 2017, 34, 58–73. [CrossRef][PubMed] 43. Hayes, C.; West, C.; Egger, G. Chapter 22—Rethinking Chronic Pain in a Lifestyle Medicine Context; Academic Press: Cambridge, MA, USA, 2017; pp. 339–353.

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).