Vital Signs: State-Level Variation in Nonfatal and Fatal Cardiovascular Events Targeted for Prevention by Million Hearts 2022

Introduction Despite its preventability, cardiovascular disease remains a leading cause of morbidity, mortality, and health care costs in the United States. This study describes the burden, in 2016, of nonfatal and fatal cardiovascular events targeted for prevention by Million Hearts 2022, a national initiative working to prevent one million cardiovascular events during 2017–2021. Methods Emergency department (ED) visits and hospitalizations were identified using Healthcare Cost and Utilization Project databases, and deaths were identified using National Vital Statistics System data. Age-standardized Million Hearts–preventable event rates and hospitalization costs among adults aged ≥18 years in 2016 are described nationally and across states, as data permit. Expected 2017–2021 event totals and hospitalization costs were estimated assuming 2016 values remain unchanged. Results Nationally, in 2016, 2.2 million hospitalizations (850.9 per 100,000 population) resulting in $32.7 billion in costs, and 415,480 deaths (157.4 per 100,000) occurred. Hospitalization and mortality rates were highest among men (989.6 and 172.3 per 100,000, respectively) and non-Hispanic blacks (211.6 per 100,000, mortality only) and increased with age. However, 805,000 hospitalizations and 75,245 deaths occurred among adults aged 18–64 years. State-level variation occurred in rates of ED visits (from 56.4 [Connecticut] to 274.8 per 100,000 [Kentucky]), hospitalizations (484.0 [Wyoming] to 1670.3 per 100,000 [DC]), and mortality (111.2 [Vermont] to 267.3 per 100,000 [Mississippi]). Approximately 16.3 million events and $173.7 billion in hospitalization costs could occur during 2017–2021 without preventive intervention. Conclusions and Implications for Public Health Practice Million Hearts–preventable events place a considerable health and economic burden on the United States. With coordinated efforts, many of these events could be prevented in every state to achieve the initiative’s goal.


Introduction
Heart disease and stroke are largely preventable (1-3). However, despite decades-long improvement in outcomes, they remain leading causes of morbidity, mortality, and health care costs in the United States (2). Moreover, considerable disparities persist and recent evidence suggests that heart disease and stroke event rates are increasing among certain demographic groups, including adults aged 35-64 years (2,4). In response, CDC and the Centers for Medicare & Medicaid Services launched Million Hearts 2022, a national initiative working to prevent one million heart attacks, strokes, and other acute cardiovascular events during 2017-2021 (1,5).
Million Hearts 2022, in collaboration with multiple federal, state, and nongovernmental partners, supports the implementation of a selected set of evidence-based public health and clinical strategies aimed at keeping adults healthy and optimizing care to prevent cardiovascular events. This includes using strategies that improve the "ABCS" (aspirin when appropriate, blood pressure control, cholesterol management, and smoking cessation) of cardiovascular care; reducing sodium consumption, tobacco use, and physical inactivity; improving care among persons who have had cardiovascular events*; and addressing known disparities in cardiovascular outcomes † (6). Despite their efficacy, implementation of these strategies throughout the country has been inconsistent, which might contribute to the disparities and geographic variation observed in cardiovascular disease (CVD) outcomes (2,4). * Includes use of cardiac rehabilitation among persons who have had a qualifying cardiac event or procedure and improving awareness of the potential negative health sequelae of and avoiding exposure to poor air quality among persons who have had a previous cardiovascular event. † Includes improving hypertension control among blacks/African Americans and improving prevention and management of CVD risk factors among adults aged 35-64 years and persons with mental and/or substance use disorders who use tobacco. US Department of Health and Human Services/Centers for Disease Control and Prevention This study describes the distribution, by demographic characteristics and state, of nonfatal (emergency department [ED] visits and hospitalizations) and fatal cardiovascular events that occurred during 2016 and are being targeted for prevention by Million Hearts 2022. Furthermore, it provides a baseline for states by estimating the number of events and hospitalization costs expected to occur if 2016 rates remain unchanged during 2017-2021. These findings can be used by Million Hearts 2022 partners to understand the recent and potential future event burden if no further intervention occurs, and to focus their use of prevention strategies and assess the potential effect on cardiovascular event totals.

Methods
This study leverages a previously published methodology to identify mutually exclusive nonfatal and fatal cardiovascular events (Million Hearts-preventable events) among adults aged ≥18 years attributed to acute myocardial infarctions, strokes, precursor cardiovascular conditions, and other cardiovascular conditions, by applying specified International Classification of Diseases, Tenth Revision (ICD-10) codes within administrative data (Supplementary Table 1, https://stacks.cdc.gov/ view/cdc/58170) (7). § Where data availability permit, event rates are described at the state and national levels for 2016 (most current available data). At the state level, the Agency for Healthcare Research and Quality's (AHRQ) ¶ Healthcare Cost and Utilization Project (HCUP) State Emergency Department Database (SEDD) was used to describe 2016 treat-and-release ED event rates for 34 states and the District of Columbia (DC)** and the HCUP State Inpatient Database (SID) was used to describe 2016 acute, nonfatal hospitalization event rates for 46 states and DC. † † Data to generate weighted national estimates for 2016 ED events were unavailable at the time of publication; these data § § will be used for national Million Hearts surveillance as they become available. Weighted national estimates and standard errors were determined for 2016 hospitalization events by using HCUP's National Inpatient Sample (NIS), ¶ ¶ a database developed from data collected via the State Inpatient Database. The National Center for Health Statistics' National Vital Statistics System*** Mortality Data were used to describe 2016 mortality rates for the nation and all 50 states and DC. Event rates were stratified by gender and age group (18-44, 45-64, 65-74, and ≥75 years), race and Hispanic origin (mortality only † † † ), event type (acute myocardial infarction, stroke, precursor condition, and other condition), and state. § § § Overall mutually exclusive event rates equal the sum of the treat-and-release ED visits, acute, nonfatal hospitalizations, and deaths and are available for 35 jurisdictions with complete data. Rates were standardized by age to the 2010 U.S. Census. Costs associated with hospitalizations were determined by applying HCUP cost-to-charge ratios ¶ ¶ ¶ to the hospital billing charges provided in the SID (state estimates) and NIS (national estimates) and are presented in 2016 US$; these costs exclude professional (physician) fees. The mean cost per hospitalization was calculated and standardized by age and § Million Hearts-preventable ED visits and hospitalizations were defined as events where patients had one of the specified ICD-10-CM codes listed as their firstlisted/principal diagnosis (national implementation of ICD-10-CM codes occurred in October 2015); Million Hearts-related deaths were events with a specified ICD-10 code listed as the underlying cause of death (national implementation of ICD-10 mortality codes occurred in 1999). Precursor cardiovascular conditions include stable angina pectoris, transient ischemic attack, and other acute and subacute ischemic heart diseases. Other cardiovascular conditions include heart failure, abdominal aortic aneurysm, atheroembolism (hospitalizations only), atherosclerosis and peripheral artery disease (deaths only), hypertension without heart failure, and cardiac arrest that had another Million Hearts-preventable event type coded as a secondary diagnosis or contributing cause of death. ¶  national event type distribution.**** Age-standardized per capita hospitalization costs, representing the overall cost per adult aged ≥18 years living in the jurisdiction, are presented at the national and state levels. State-level estimates for the number of Million Heartspreventable events and hospitalization costs (in 2016 US$) expected to occur during 2017-2021 were calculated in two ways. For states with complete 2016 data (ED, hospitalization, and mortality estimates), the overall age-specific mutually exclusive rates for 2016 were applied to the projected state population estimates † † † † during 2017-2021 and summed to determine the expected event totals; the 2016 mean stateand age-specific cost per hospitalization was applied to the expected hospitalization event total to estimate expected costs. For states with incomplete 2016 data, it was assumed that the proportional relationship across their ED, hospitalization and mortality rates were the same as the average calculated among states with complete data. If a state was missing 2016 hospitalization data, the national age-specific average cost per hospitalization event was applied to their expected age-specific hospitalization event totals and summed. Expected overall U.S. event totals and hospitalization costs during 2017-2021 equal the sum of the state-level estimates. § § § §

Results
Nationally, in 2016, over 2.2 million Million Hearts-preventable hospitalizations and 415,480 deaths occurred ( Table 1). The hospitalizations resulted in an estimated $32.7 billion in costs. For both event types, the burden was higher among men than among women (age-standardized hospitalization rates of 989.6 and 725.1 per 100,000 population, respectively, and mortality rates of 172.3 and 143.0 per 100,000, respectively) and increased with age. However, an estimated 805,000 hospitalizations and 75,245 deaths occurred among adults aged 18-64 years. Among all racial/ ethnic groups, the highest mortality rates were in non-Hispanic blacks (211.6 per 100,000). Acute myocardial infarctions and strokes accounted for approximately half (47%) of hospitalizations (rates of 204.5 and 199.1 per 100,000, respectively) and approximately two thirds (61%) of deaths (42.2 and 53.7 per 100,000, respectively). "Other" cardiovascular events, which include those related to heart failure, contributed to 46% of hospitalizations and 38% of deaths (rates = 394.6 and 59.8 per 100,000, respectively).
Age-standardized event rates per 100,000 population varied considerably across states with available data, including for treat-and-release ED visits ( Figure 1, https://stacks.cdc.gov/view/cdc/58168). Among the 35 jurisdictions with complete overall data, the three with the lowest overall mutually exclusive event rates were Utah (805.7), Wyoming (828.9), and Vermont (840.6) and those with the highest rates were DC (2,048.2), Tennessee (1,551.6), and Kentucky (1,510.3) (Figure).
In 2016, the age-standardized per-capita hospitalization cost was $125 in the United States, and ranged across states with available data from $76, in New Mexico and Wyoming, to $294, in DC ( Table 2). The age-and event type-standardized mean cost per hospitalization was $16,274 nationally and ranged from $11,307 in Arkansas to $24,017 in Alaska. If the 2016 overall mutually exclusive event rates were to remain constant during 2017-2021, an estimated 16.3 million events are expected to occur, including 2.2 million ED visits, 11.8 million hospitalizations, and 2.2 million deaths (Table 3) (Supplementary Table 2, https://stacks.cdc.gov/ view/cdc/58171); if mean hospitalization costs per event remained constant, an estimated $173.7 billion in costs would be expected to occur. Preventing one million events during 2017-2021 would result in approximately a 6.1% reduction in expected event totals and associated costs.

Conclusion and Comment
The subset of cardiovascular events ¶ ¶ ¶ ¶ targeted for prevention by Million Hearts 2022 places a considerable burden on the health and economic well-being of Americans (2,7). Despite these events being highly preventable (3), they accounted for approximately 2 million hospitalizations and 400,000 deaths in 2016. Furthermore, without a more concerted effort to improve CVD risk factors, an estimated 16  Considerable disparities in event rates were evident. Nationally, non-Hispanic blacks continue to experience the highest CVD mortality rates (32% higher than those in non-Hispanic whites). This disparity is due, in part, to the high prevalence of uncontrolled blood pressure among blacks (6), placing them at higher risk than other racial/ethnic groups for acute myocardial infarction, stroke, and other CVD conditions, including heart failure (2,8). Additionally, despite the considerable increase in risk for a cardiovascular event with increasing age, in 2016, over 800,000 combined hospitalizations and deaths occurred among adults aged <65 years (approximately one in three events). Other studies have shown that decades-long improvement in heart disease and stroke mortality have stalled (9,10) and that younger populations, especially those aged 35-64 years, are experiencing worse outcomes across the country (4,9). In 2016, Million Hearts-preventable event rates among persons aged 35-64 years varied considerably by demographic characteristics and U.S. state (Supplementary Table 3

Per-capita costs (US$) §, **
Alabama  Figure 2, https://stacks.cdc.gov/view/ cdc/58169). Therefore, implementation of strategies that focus on the prevention, early diagnosis, and effective management of CVD risk factors among younger adults is needed to prevent events in both the short-and long-term. This is one of the first studies to demonstrate striking state-level variation in nonfatal cardiovascular event rates and hospitalization costs using data collected among adults of all ages and across all payer types, including the uninsured. Whereas the burden of state-level mortality was higher in the southeastern United States, which aligns with the findings from previous studies (2), rates for ED visits and hospitalizations were higher in both this region and elsewhere, including many Midwestern states. The overall state variation in nonfatal event rates and associated hospitalization costs is likely driven by both geographic differences in disease prevalence and severity, and differences in care delivery and public health quality (2,11). Additional focus on improving the environments in which persons live, work, and play (e.g., built environment modifications to promote increased physical activity) (12), leveraging community resources to aid in CVD risk factor management (e.g., referral to nutritional and fitness counseling groups) (13), providing effective outpatient care (e.g., use of team-based care for hypertension and cholesterol management) (14), and improving the care received after a cardiovascular event (e.g., systematic referral to cardiac rehabilitation services for those with eligible diagnoses) (15) might reduce the need for and the expense of many of these acute care services.
This study uses the best available data to describe the burden of Million Hearts-preventable events at the national and state levels. However, the findings in this report are subject to at least six limitations. First, not all jurisdictions provided ED and/or hospitalization data; efforts to impute missing values might produce inaccurate estimates. Second, nonfatal events are attributed to treatment location and not patient residence, therefore jurisdictions in close proximity to large population centers in neighboring states (e.g., DC) could have overestimated or underestimated rates. Third, whereas the methodology used attempts to identify mutually exclusive events, there is potential for over-or undercounting events. Fourth, cardiovascular § Described by applying HCUP cost-to-charge ratios to the charges the hospitals billed for the entire hospital stay in each state; these costs exclude professional (physician) fees. ¶ Standardized by age to the 2010 U.S. Census population and by event type to the national distribution of events (acute myocardial infarctions, strokes, precursor events, and other cardiovascular events) observed during 2016. ** Represents the overall cost per adult aged ≥18 years living in the jurisdiction, standardized by age to the 2010 U.S. Census population.
† † Represents the sum of number of treat-and-release ED visits, acute, nonfatal hospitalizations, and deaths that occurred in that jurisdiction. § § Data were not collected or were not available for analysis (treat-and-release ED data for Mississippi and Oregon are regularly collected, but 2016 data were not available at time of this report). ¶ ¶ The ED and hospitalization event and cost values are likely overestimates for this location because these events are attributed to where the event was treated and not the residence of the patient. *** Transfers to other acute care hospitals are not identified in discharge disposition codes, so they could not be excluded from the analysis and the rate may be slightly inflated. events that do not result in ED or hospital use or death are not counted. Fifth, because administrative data are being used, differences in use of health care (e.g., changes in how events are medically managed) or coding practices (e.g., changes in how nonfatal events are billed) might affect the event rates and costs presented in this study rather than changes in disease burden. However, this study attempts to address differences in practice patterns by excluding elective hospitalizations and including certain ED events (e.g., heart failure-related visits). Finally, the hospitalization cost estimates are likely conservative, as they do not include professional (physician) fees, and costs were not available for treat-and-release ED visits. Each state would need to realize an approximate 6% decrease in its expected event totals during 2017-2021 to collectively prevent one million events at the national level. This is feasible if clinical and public health partners in every state mobilize and strengthen their focus on implementing the prevention strategies outlined by Million Hearts 2022 (https://millionhearts.hhs.gov/files/MH-Framework.pdf ) to achieve 80% or greater performance on the ABCS and at least a 20% reduction in physical inactivity, tobacco use prevalence, and sodium consumption (16). What are the implications for public health practice?
Achieving the Million Hearts 2022 goal likely requires states to focus on using prevention strategies that best meet the cardiovascular health needs of the persons they serve. Abbreviation: ED = emergency department. * State-level estimates for the number of Million Hearts-preventable events (those targeted for prevention by the initiative) and hospitalization costs expected to occur during 2017-2021 were calculated in two ways. For states with complete 2016 data (ED, hospitalization, and mortality estimates), the overall age-specific mutually exclusive rates for 2016 were applied to the projected state population estimates during 2017-2021 and summed to determine the expected event totals; the 2016 mean state-and age-specific cost per hospitalization was applied to the expected hospitalization event total to estimate expected costs. For states with incomplete 2016 data, it was assumed that the proportional relationship across their ED, hospitalization, and mortality rates was the same as the average calculated among states with complete data. If a state was missing 2016 hospitalization data, the national age-specific average cost per hospitalization event was applied to their expected age-specific hospitalization event totals and summed. Expected overall U.S. event totals and costs during 2017-2021 equals the sum of the state-level estimates; this method differs from the method being used to officially track these estimates at the national level. † ED rate was missing, therefore, an estimate was used. § Hospitalization rate and cost per hospitalization information were missing, therefore, estimates were used. ¶ Calculated by using estimated ED visit rates and, where applicable, estimated hospitalization rates and summing those resultant event counts with the known death counts.